Technical Report 571

HUMAN FACTORS REPRESENTATIONS

*; FOR COMBAT MODELS

George J. Miller and Seth Bonder

Vector Research, Incorporated

%
: I

-,.or

0... U. S. Army D
Research Institute for the Behavioral and Social Sciences

July 1982
C . Approved for Public release, distribution unlimited.

83
 UNCLASSIFIED
SECURITY CLASSIFICATION Of THIS PAGE (When Date Ent-red)
REPORT DOCUMENTATION PAGE READ INSTRUCTIONS
RE R DBEFORE COMPLETING FORM
1. REPORT NUMBER 2. GOVT ACCESSION NO. 3. RECIPIENT'S CATALOG NUMBER

Technical Report 571

4. TITLE (nd Subtitle) S. TYPE OF REPORT & PERIOD COVERED

HUMAN FACTORS REPRESENTATIONS FOR COMBAT MODELS Final May 1981 - July 1982
6. PERFORMING ORG. REPORT NUMBER
___VR-ARI-6 FR 82-1
7. AUTHOR(e) 6. CONTRACT OR GRANT NUMBER(-)

George J. Miller, Seth Bonder MDA 903-81-C-0303

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT. PROJECT. TASK
Vector Research, Incorporated AREA 6 WORK UNIT NUMBERS

P.O. Box 1506 2Q161102B74F

Ann Arbor, MI 48106
It. CONTROLLING OFFICE NAME AND ADDRESS 12. REPORT DATE
ARI Field Unit July 1982
P.O. Box 3122 IS. NUMBER OF PAGES
* .,Fort Leavenworth, KS 66027 195
14. MONITORING AGENCY NAME & ADORESS(II different from Controlling Office) 15. SECURITY CLASS. (of this report)

U.S. Army Research Institute for the Behavioral

and Social Sciences
b 5001
.•, Eisenhower Avenue 15. DECLASSIFICATION/DOWNGRADING
SCHEDULE•
Alexandria, VA 22312 SD
16. DISTRIBUTION STATEMENT (of this Report)

Approved for public release; distribution unlimited.

17. DISTRIBUTION STATEMENT (of the abstract entered In Block 2C, It different from Report)

S..,

10. SUPPLEMENTARY NOTES

13. KEY WORDS (Continue on reverse aide If necessary nd Identify by block number)

Computerized simulations Combat effectiveness

Simulation Human factors engineering
Operations research Human performance
Combat
1 21. A9SrRAC? (Cowthno M rvereeefd it nceeey and identify by block number)

.4Because human factors are not explicitly represented in the combat simu-
lations employed by the Army, a detailed review was made of how 112 combat
processes are represented in the following simulations: CARMONETTE, BLDM,
DIVLEV, FOURCE, JIFFY, VECTOR-2, CEM, McCLINTIC, and CASTFOREM. From the re-
view it was possible to enumerate the assumptions about human factors implicit
in these simulations and describe a structure within which the connections
between human characteristics and combat processes could be related. This
structure
-sal serves as a criterion for evaluating proposed human,, (Continued)
DO FM 1473 EDITION OFINOV 6S IS OSOLETE UNCLASSIFIED

SECURITY CLASSIFICATION OF THIS PAGE (When Date Entered)

 UNCLASSiFIED
SIECURITY CLASSIFICATION OF THIS PAGE(flU. Date Entered)

* Item 20 (Continued)
Ifactors research projects related to combat simulations. The highest priority
combat processes and human factors for research are also enumerated.

7.

pie

Accesion
0 For
NTIS GRA& I
DTIC TABo
unannoun~ced f
Justiticatto

Distribution/_______
A1vaii and/or~I

D Spec0ial
UNCASSFIE Dt Elrd
CL SIIC TOoF HS PAe1l.n
Availbilit

. .
. .
. -.
o
c.. . . . . . . . . . . . . . .
- - .
 Technical Report 571

HUMAN FACTORS REPRESENTATIONS

FOR COMBAT MODELS

George J. Miller and Seth Bonder

Vector Research, Incorporated

Contract monitored by Robert McOuie

Approved as technically adequate

and submitted for publication by
Robert M. Sasmor, Director
BASIC RESEARCH

U.S. ARMY RESEARCH INSTITUTE FOR THE BEHAVIORAL AND SOCIAL SCIENCES
5001 Eisenhower Avenue, Alexandria, Virginia 22333

Office, Deputy Chief of Staff for Personnel

Department of the Army

July 1982

Army Project Number Basic Research

20161102874F

Approved for public rIleise; distribution unlimited.

iii

° . + . ° +.. ° . . ° .. . .o- .. , . . . . . . .. o
 71 .7

ARI Research Reports and Technical Reports are intended for sponsors of
R&D tasks and for other research and military agencies. Any findings ready
for implementation at the time of publication are presented in the last part
of the Brief. Upon completion of a major phase of the task, formal recom-
mendations for official action normally are conveyed to appropriate military
agencies by briefing or Disposition Fou in.

It..'

,-Z

{ ""-

- . .
v ,

. . . . . . .
J r - -

FOREWORD

Although human beings fight wars, human factors are not well represented
in simulations of combined arms combat. This is due to a lack of knowledge
of the specific effects of human factors on battle processes. This report
identifies high-priority behavioral science areas for research on such simu-
lations. The report makes explicit the presently implicit assumptions re-
e •lated to human factors and provides a structure within which to relate char-
acteristics of human behavior to characteristics of combat modeling. Such a
structure provides a framework that individual research projects should build
upon to improve combat simulations used by the Army.

The research on which this report is based was performed by Vector Re-
search, Inc. (VRI). Besides the authors, other individuals who contributed
to the results were Robert Blum, Robert Farrell, and General William DePuy
(USA, ret.) of the VRI staff, who furnished useful suggestions and insights;
Stanley Spaulding and Jeffrey Alden, who made sensitivity runs with the BLDM
combat model; and Stanley Halpin, Irving Alderman, and Robert Sasmor of
the Army Research Institute (ARI), who also contributed useful suggestions.

EDGAR M. JOHNSON
Technical Director

. . . . . . . . .................................

v
 HUMAN FACTORS REPRESENTATIONS FOR COMBAT MODELS

EXECUTIVE SUMMARY

Requirement:

Human factors are not explicitly represented in Army simulations of

combined arms combat. This is because of a lack of knowledge of the effect
of human behavior on the various battlefield processes. Accordingly, the
Army Research Institute is developing a program to identify factors, algo-
rithms, and models that could represent the effect of human performance on
weapons and tactics.

Procedure:

The approach employed was to analyze nine existing combat simulations:

CARMONETTE FOURCE CEM

BLDM JIFFY McCLINTIC
DIVLEV VECTOR-2 CASTFOREM

The methods by which these simulations represent 112 combat processes are
described in detail. This information is then used to identify human per-
formance interactions in the simulations and the simulations' sensitivity
to these interactions. This information was then used to develop programs
and priorities for human factors research related to combined arms
simulations.

Findings:

In terms of combat processes, highest priority human factors research

topics are:
(1) employment of maneuver-units;
(2) engineering operations; and
(3) fire suppression.

In terms of human abilities, highest priority research topics are:

2 (1) performance of tasks under adverse conditions;

(2) decision making; and
(3) recognition of features and patterns.

vii

' . vii
a.''''*" - '' ;- '' ." " ." - -" " . '" " ." --' ." " .' '." ", . - " '
Utilization of Findings:
This report provides a structure that will enable researchers to iden-
tify the connections between human behavioral knowledge and the processes
in combat simulation. This structure provides the framework that individual
research projects should build upon in further studies. Thus, it provides a
criterion for evaluation of proposed research in this area.

vii
 --- ~
.---- - . S 4. - 7 . - .

HUMAN FACTORS REPRESENTATIONS FOR COMBAT MODELS

CONTENTS

Section Page
1.0 OVERVIEW. .. .... ...... ....... ...... ..... 1 -

2.0 COMBAT PROCESSES AND HUMAN FACTORS AREAS. .. .... ........ 10

2.1 Model Processes .. .. ...... ....... ........ 10

2.1.1 Approach .. .. ...... ....... ........ 10
2.1.2 Results .. ..... ...... ....... ..... 14
2.2 A Typology of Human Factors Areas .. .. ...... ...... 19
2.2.1 Approach .. .. ...... ....... ........ 19
2.2.2 Results .. ..... ...... ....... ..... 21
2.3 Areas of Model Sensitivity .. .... ....... ...... 23
*2.3.1 Approach .. .. ...... ....... ........ 24
2.3.2 Results .. .... ....... ....... ..... 25
3.0 STRUCTURE AND PRIORITIES FOR HUMAN FACTORS RESEARCH . ... .. . .28

3.1 Development of the Research Structure. .. .... ....... 30

3.2 Development of Research Priorities. .. ....... .... 33
3.3 Research Priorities and Observations .. ..... ....... 39
REFERENCES .. .... ....... ...... ....... ...... 57
APPENDIX A. COMBAT PROCESSES IN SPECIFIC SIMULATIONS. .. .... ... A-1
Maneuver Unit Combat. .. .... ...... ....... .... A-4
Artillery Fire. .. .... ....... ...... ....... A-11
Air-to-Ground Attacks. .. ....... ...... ....... A-15
j Air-to-Air Combat......................A-18
Air Defense. .. ....... ...... ....... ..... A-21
Mobility, Countermobility, and Survivability. .. .... ..... A-24
Intelligence and Electronic Warfare. .. ....... ...... A-29
Communications ....... .. ....... ...... ....... A-34
Mission Assignment, Organization for Combat, and
Resource Allocation .. .. .. .. .. .. .. .. .. .. . .A-37
Maneuver Control. .. ..... ...... ....... ..... A-42
Fire Support Allocation. .. ....... ...... ...... A-46 -

Resupply and Replacement .. .. ...... ....... ..... A-5O

Movement .. .. ....................... A-53
Maintenance and Repair; Medical'Support.............A-56
Construction, Maintenance, and Operation of Airfied.. ..... A-59
APPENDIX B. A TYPOLOGY OF HUMAN FACTORS AREAS. .. ....... ... B-1

ix
CONTENTS (Continued)

Page

APPENDIX C. SENSITIVITY ANALYSES OF SIMULATIONS ..... ........... C-,

APPENDIX D. NOTES ON PRIORITIES FOR HUMAN FACTORS RESEARCH ..... D-1

LIST OF EXHIBITS
Exhibit 1-1. Impact on Army Decisions of Human Factors Research
for Combat Models ..... ... ................... 3

1-2. Conceptual Approach to the Project ..... ........... 4

2-i. Fleishman Ability Taxonomy .... ............... ... 22

3-I. Typology for Human Factors Representations in Combat
Models ..... ....... ... .... ... ... 29

3-2. Impact on Army Decisions of Human Factors Research

for Combat Models ...... ................... .... 35
3-3. Relative Ranking of Model Process Phenomena .... .. 40

3-4. Highest Priority Phenomena .... .................. 45

3-5. Relative Ranking of Task Clusters ..... ... ... 46

3-6. Highest Priority Task Clusters ..... .... .... 47

3-7. Relative Ranking of Abilities ...... ............. 48

3-8. Highest Priority Abilities ..... ....... ... 49 -

A-1. Models Considered .... .... ................... A-2 ,.

B-1. Definitions of Generic Task Clusters ... ......... ... B-2

B-2. Task Clusters Associated with Each Model Process

Phenomenon ..... ....... ... ........ B-4

B-3. Task Analysis--Sumary Statistics .... ......... ... B-9

B-4. Fleishman Ability Taxonomy .... ............... .. B-10

B-5. Analysis of Abilities Required for Each Task Cluster . . B-14

B-6. Abilities Analysis--Summary Statistics ...........

... B-16

C-i. Summary of BLDM E).ursions ..... .............. ... C-3

-."
,,.,:,-L-,-.'.,
•;, . ,-.'.o. ,.:, '.'..-.'.-...............'.-.............-.-...-..................-.'...........-............-....-.-.-.-.....-..'
 CONTENTS (Continued)

Page

Exhibit C-2. Effect of Blue's Time to Fire ...... ............. C-6

C-3. Effect of Blue's Priorities ...... .............. C-7

C-4. Effect of Blue's Open Fire Decision ..... .......... C-S

C-5. Effect of Number of Blue Participants ........... .... C-9

C-6. Effect of Number of Blue and Red Participants..... C-10

C-7. Effect of Blue Round Dispersions ........... C-li

C-8. Effect of Blue First Round Dispersions .. ........ .. C-12

C-9. Effect of Blue Subsequent Round Dispersions ... ...... C-13

C-lO. Effect of Blue Round Dispersion Following a Miss C-14

C-li. Effect of Blue Round Dispersion Following a Hit .... C-15

C-12. Sumary of VECTOR-2 Excursions ............ C-18

C-13. Sensitivity Analyses in Terrain Line-of-Sight Study C-22

C-14. Hand-Off Time and Reaction Time Variations ... ...... C-23

C-15. VECTOR-O Sensitivity Analyses ... ............. ... C-25

C-16. Comparison of Effectiveness of Blue Battalion

Supported by M9 ACE Versus D7 Dozer ............ .... C-29

C-17. Relative Sensitivity Associated with Each Model

Process Phenomenon .................. C-33

D-1. Determination of Relative Ranking of Model Process

Phenomena ........ .......... ..... D-2

D-2. Determination of Relative Ranking of Task Clusters . . D-9

D-3. Determination of Relative Ranking of Abilities . . . . D-lO

S.X.

ix

-
F ,:',
;- -,,"'"'Z,,
. -/ '.i-' ,:' .:,.":"-:',"
:',',,"."'," , '.i--".' '. . '"- i-" •i . -' . . - T' i '- .' .- .."..'" -' "
/ .,.*.-.. . . . . . . . . . il . a . i . . . .
 1.0 OVERVIEW

Human participation and influence are pervasive in land and air com-

bat. Although this is well recognized, human behavior and human factors
.- 1
generally are considered only implicitly in force-on-force models

employed by the Army analysis community. The implicit nature of the

modeling, and the assumptions underlying the models, are due in large

part to the absence of knowledge regarding the effect of human factors

and behavior on battlefield processes. Accordingly, the Army Research

Institute (ARI) is structuring a basic research program to identify

appropriate assumptions and models for representing human factors and

behavior in combat models. This report describes the work and results of

a project performed by Vector Research, Inc. (VRI) for ARI to identify

high priority behavioral science efforts for the research program.

The report attempts to make explicit the presently implicit models

and assumptions in the human factors area and to provide a structure

within which the connections between human behavioral knowledge and com-

bat modelin) can be clearly understood. This structure provides a frame-

work within which individual research projects can be conducted with some
assurance that they will contribute to improvements in model representa-

tions of human factors. Such improvements should have a twofold benefit

to the Army. First, they should improve the credibility of Army opera-

tional analyses and expand the scope of analysis issues that can be

addressed by the operational analysis community. Second, they will re-

sult in models which are useful to the Army behavioral science community

for analysis of the benefits of training and analysis of recruiting

requi rements.

. . . . .-

o,. e° . , ., a • ° , t,, tL ., • . . ... . . .. . . •,. .- . ,•°

. •,•.. • •,°°
,.-

The ultimate impact on Army decisions of human factors research to

improve combat models is illustrated in exhibit 1-1: The research will

result in an improved capability to predict human behavior in combat.

*. Use of this knowledge within combat models will result in analysis

results which are more accurate and more sensitive to important human

characteristics and interactions. Proper use of these results in Army

decision making should effect better decisions in such areas as system

procurement, force structuring, training, and recruiting.

The overall approach taken in the project was to analyze existing

combat models to identify the explicit and implicit human factors inter-

actions in the models and the sensitivity of simulated battle results to

assumptions concerning these interactions. Conceptually, the approach is

illustrated in exhibit 1-2. The exhibit indicates that simulated battle

*- results from a combat model can be traced to the representation of combat

*:i processes and related phenomena within the model. In the real world,

these phenomena are partially based on associated human tasks which, in

turn, consist of "behavioral components" describing task performance.

" Thus, human factors research for improving combat models can be viewed as

research to predict task performance as a function of the behavioral com-

ponents and to relate the behavioral components to measurable human

characteristics, aspects of the combat environment, and measures of the

training and experience of individuals performing the tasks. 1 For

*'I example, a commonly used model output is the loss exchange ratio -- the

As will be noted later, actual research might not explicitly consider

the behavioral components, but might rather attempt to predict task per-
formance directly as a function of human characteristics, the combat
environment, and training and experience.

Se * *...... . . - . -. .... .. . . . . .

J - -
. . . . ". . .. .
""""" "" " " - : .. .: " ' -i _ , ,- -- I- - - ,b, --- , - .
-. EXHIBIT 1-1: IMPACT ON ARMIY DECISIONS OF HUM~AN FACTORS RESEARCH FOR
COMBAT MODELS

Human factors research

Improved prediction
of humian behavior
in combat

Revisions to
combat model s

More "accurate"
model output

Use of model output

in Army decisions

3
 7I'-

w CU
z or- (3 (A
4 x c C - (3 0
w K ( I~U.

x u
jU .C
cm cc w
0
~ -
- -
K

*~~~ L& K L& M

S M
.L
~ ee . Z 0
~z
A
AU . - -U

C A M - Z K SM0 (

U ( U. 66
L
SMj

in
U CL

= w 0. A. 0. 'I
U.
U
usLU U 0 C -1 U.
GM cc u uiS

L)
0
I-K

u z 0

o 0
w SM
9L.
(A-
0 0 (
~ 4
0 -
C.
M C
SM U
A.9L
- -
-

). K A. M 1- K U
Ui 0 SM . 4 0 A
U W o. A. u
.4

ratio of enemy losses to friendly losses. The value of this output mea-

sure depends in part on the model's representation of the accuracy of

fire of the weapons involved in the simulated engagement. Accuracy of

fire depends on performance of the task "aim and track;" this performance

is a function of certain basic abilities of the weapon crew members.

Improvement in the capability to predict such task performance can result

.* in improved representation of the accuracy of fire in combat models and

thereby provide improved estimates of the loss exchange ratio.

The project tasks, which were designed around this conceptual struc-

ture, involved the use of heuristic and judgmental techniques to develop

a logical basis for setting research priorities to develop the needed

human factors relationships. The project consisted of four substantive

tasks:

(1) Identify model processes and phenomena which involve human

factors interactions. Fifteen model processes were identified,

each comprising three to 12 phenomena involving human factors.
A representative set of nine combat models was studied to

determine the manner in which each phenomenon is represented

in each model.

(2) Organize these processes and phenomena into a structured typol-

ogy of human factors areas (i.e., identify related tasks and

their behavioral components). Eighteen generic groupings of

tasks (referred to as "task clusters") were defined, and the

task clusters associated with each of the model process phenom-

ena were identified. An analysis was performed of the degree

to which each of 37 abilities is required to perform the tasks

included in each task cluster.

. -..
. ..
. ;:::; :,::--:I"6:::::::::::::::: ::: ::::: ::...;-.-.
i",i 2::...- .:. . - ..:..: :- ;.
 (3) Identify those human factors areas for which changes in assump-

tions might lead to significant changes in simulated combat

results. A total of 63 runs of two combat models was made and

the results of ten previous studies were analyzed to determine

the potential sensitivities of simulated battle results to

changes in human factors assumptions.

(4) Use the information generated in tasks 1, 2, and 3 to structure

a framework for conducting human factors research designed to

improve human factors representations for combat models and to

identify relevant high priority research efforts.

The sensitivity analysis of task 3 investigated the potential impact

of separate changes to each of many assumptions about human factors

effects in combat models. Some of these individual effects were found to

be highly significant -- a number could cause a change in the predicted

winner of an engagement. No analysis was conducted of the potential

l effect of the simultaneous improvement of human factors representations

in a whole spectrum of model phenomena. However, the observed impact of

some of the individual changes in assumptions suggests that a general

*improvement in the treatment of human factors in combat models would be

,. likely to have a very large and unpredictable effect on simulated battle

]* results.

The project thus identified a significant need for research to de-

velop relationships for use in representing human factors in combat models.

Two general approaches to human factors research can be taken to provide

the needed relationships. The first approach would involve four general

types of research:

6
 (1) research to describe, as functions of measurable human abili-

ties, 2 the effect of human factors on the processes and

*phenomena represented in combat models;

(2) research to develop methods to predict human abilities as func-

tions of demographic descriptors of an individual (such as

ASVAB scores, age, sex, and education);

(3) research to predict the effect of the combat environment on

abilities; and

(4) research to establish the effect of training and experience on

abilities.

The second research approach would omit explicit consideration of human

abilities. This approach would have less of a phenomenological structure

roll than the first but would also require less long-range basic research.

It thus might produce shorter-term interim results for use with combat

models. For this second approach, three general types of research would

be needed:

(1) research to describe, as functions of demographic descriptors

of an individual, the effects of human factors on model pro-

cesses and phenomena;

'- (2) research to predict the relationships between conditions in the

combat environment and the human factors effects on model pro-

cesses and phenomena; and

(3) research to establish relationships between training (and

experience) and the human factors effects on model processes

and phenomena.

2precise definitions of "abilities" and other terms used in this

research appear later in this report.

i' 7
In the area of model processes and phenomena, high priority research

topics (representing the top 10 to 15 percent of all phenomena

identified) involve:

(1) deployment, employment, movement, and maneuver of maneuver unit

forces;

(2) engineer operations (primarily because of their potential

effect on (1)); and

(3) fire suppression.

Alternatively, priorities for research can be established for sets of

generic human tasks, where each set of tasks is associated with a number

of model processes and phenomena. High priority research topics for such

task clusters (representing the top 33 percent of all task clusters

considered) involve:

(1) command, control, communications, and intelligence; and

(2) utilization of weapons under combat conditions (aiming, track-

ing, loading, and firing).

In the area of human abilities, high priority research topics (repre-

senting roughly the top 30 percent of all abilities considered) involve:

(1) abilities necessary to perform tasks under adverse conditions;

(2) abilities required in decision making; and

(3) abilities involving recognition and visualization of features

and patterns.

This report consists of three chapters and four appendixes. Follow-

ing this executive summary, chapter 2.0 describes the work performed on

project tasks 1,2, and 3 (which might be viewed as the information-

gathering phase of this research). For each task, the chapter describes

........ .. . ...
the approach taken and summarizes the results of the task. Chapter 3.0

presents a structure for conducting the needed human factors research and

suggests priorities for such research, based on the information assembled

in the first three tasks. Detailed information developed in each project

task is presented in an appendix -- appendix A for task 1, B for task 2,

C for task 3, and D for task 4.

-Am

A°-

.€. . .. . .. .. . . . . . . . . . . . . . . .'

* 9
 2.0 COMBAT PROCESSES AND HUMAN FACTORS AREAS

A large amount of information was collected and analyzed to develop

an understanding of current'human factors representations in combat

models and the potential impacts of changes in these representations.

This chapter describes the work performed in this information gathering

and analysis portion of the study (tasks 1, 2, and 3) and summarizes

results and insights gained in the process of assembling and organizing

the information. Section 2.1 discusses the identification of model pro-

cesses which explicitly or implicitly represent human factors effects

(task 1). Section 2.2 describes the methods and results of task 2, which

involved restructuring the information assembled in task 1 in terms of

human factors considerations. Finally, section 2.3 summarizes the infor-

mation gathered to assess the sensitivity of models to human factors

assumptions and representations.

2.1 MODEL PROCESSES

The first task in this project involved determining the phenomena

now represented in combat models which describe, explicitly or implic-

itly, human factors effects. The approach taken in this task is

described in section 2.1.1; results of the task are summarized in section

2.1.2.

2.1.1 APPROACH

In order to develop a description of the ways in which human factors

interact with combat models, a representative set of Army models was

analyzed to determine those model phenomena which are influenced by human

1
10
. . . . . . .1.. .0 - .

.
. . . -' ....-.
.% . . . . -'-..-
T" ,... - " " -
 -- -- -- -
factors and the ways in which those interactions are represented (either

explicitly or implicitly) in contemporary models. An attempt was made to

identify d broad spectrum of force-on-force models which include repre-

sentative samples of battalion-level, division-level, and theater-level

models; deterministic models and Monte-Carlo simulations; and computer-

ized war games as well as fully automated models. Only dynamic models
(models which represent a time trajectory of changing combat conditions)

and emulatory models (models which are separable into identifiable enti-

ties and processes) were considered.3 Further requirements for inclu-

sion in this analysis were that a model be actively used or anticipated

to be used by the Army, and that the model be reasonably well documented 47

(so that the model's content could be analyzed easily).

Based on these criteria, nine force-on-force Army models were iden-

tified for use in this study. Each of these models is described briefly

below:

(1) CARMONETTE is a Monte-Carlo simulation of ground combat, gener-

ally used to represent combat at company to battalion level.

It was one of the first force-on-force models developed for the

Army and (with various modifications ard improvements) is still

in active use by several Army agencies.

(2) BLDM is a deterministic model of ground combat at company to

battal ion level. It is representative of a set of battalion-

level models (including Bonder-IUA, AIDM, TRACOM, etc.) which

3Actually,
all models contain some non-emulatory features, but for each
of the models considered here there is at least some degree to which the
processes and entities being represented can be identified in the model
structure. It is easier to identify and incorporate human factors
representations in an emulatory model than in a non-emulatory one.

+ 11

a.- ..
. .... ".. ... "....
" -..... " .. ...- " " ' " . . .- . . ~ .
 -. o

use a differential equation or "Lanchester-type" equation

structure to represent attrition processes.

(3) CASTFOREM is the lowest-level, highest-resolution model in the

hierarchy of models being developed as a part of the Army

Models Improvement Program (AMIP). 4 It is a detailed Monte-

Carlo simulation of battalion-level combat which employs user-

provided decision tables to represent decision processes.

(4) DIVLEV is a division-level combined arms war 'game (i.e., a

player-controlled model). It uses deterministic procedures

(including differential equation structures) to assess weapon

effects.

(5) FOURCE is a division-level deterministic model which uses

Lanchester-type equations to assess maneuver unit attrition.

The model emphasizes representation of staff performance and

information flow for use in assessing command and control and

intelligence alternatives.

(6) JIFFY is a corps-level war game which uses simple, equations to

assess attrition deterministically.

(7) VECTOR-2 is a corps-to-theater level deterministic model of

combined arms combat. It employs differential equation struc-

tures to assess maneuver force and some other types of attri-

tion and user-input tactical decision rules to represent deci-

sion making.

(8) CEM is a theater-level deterministic model of combined arms

combat which uses "firepower potential" (or "firepower score")

4 Documentation for the other AMIP models CORDIVEM and FORCEM

-- -- was
not available in time for inclusion in this analysis.
12.

12-'

" ,'." - .i . -- -. . ' . . . .. , ."

 constructs to assess attrition and built-in decision rules to

* control the action.

(9) McCLINTIC is a highly aggregated quick-running theater-level

war game whose attrition computations are based on firepower

.. scores.

References to docunentation for each of these models can be found in

appendix A.

Each of these nine models was evaluated in order to determine the

major processes which the models represent. The following processes

(which are defined in detail in appendix A) were identified:

(1) maneuver unit combat;

.- (2) artillery fire;

(3) air-to-ground attacks;

(4) air-to-air combat;

(5) air defense;

.(6) mobility, countermobility, and survivability;

(7) intelligence and electronic warfare;

(8) communications;
(9) mission assignment, organization for combat, and resource

allocation;

(10) maneuver control;

(11) fire support allocation;

(12) resupply and replacement;

(13) movement;
(14) maintenance and repair, and medical support; and

(15) construction, maintenance, and operation of airfields.

13

kP*4 .. .... ' I. I " I ... . .. * " - .* ' - - ,. . / -. . .; . - ,. ..

 Each process was then subdivided into a set of phenomena which are influ-

enced by human factors. This was accomplished through consideration by

the project team (which included individuals knowledgeable in both combat

and combat models) of the ways in which humans participate in or other-

wise influence each of the processes. For example, for the process of

maneuver unit combat, the following phenomena were identified: force

participation, initial deployments of weapons, maneuver patterns of indi-

vidual weapons, target acquisition, target identification, target selec-

tion (and target switching), decision to fire, ordnance selection, time

to fire or firing rate, accuracy of fire, fire suppression (direct and

indirect), and employment of obscuration. Appendix A lists the phenomena

associated with each of the processes.

Finally, each model was studied to determine the way in which each

of the process phenomena is represented. Sources of this analysis

included available model documentation, discussions with model users and

developers, and experience of the project staff in developing and using

some of the models involved. It should be noted that available documen-

tation for many of the models is old, incomplete, or imprecise. There-

fore, the results of this assessment are not a detailed comparison of

model capabilities; rather, the results are intended to provide a general

indication of the treatment of human factors in existing models.

Detailed results of this analysis appear in appendix A; the next section

of this chapter summarizes these results.

2.1.2 RESULTS
Two general types of phenomena influenced by human factors were

identified -- "human performance phenomena" (those involving the conduct

14

.. .. . . . . . . . . . .
 of one or more physical tasks) and "decision making phenomena." The

phenomena are generally assumed to be interconnected by "organizational

coordination" and to be represented in the context of "battlefield

conditions." These general types of phenomena and related model features

are discussed separately in the paragraphs below.

Within the human performance phenomena, human factors are sometimes

represented explicitly. For example, some of the detailed models of com-

pany to battalion level combat include inputs specifying the time to fire

a round under various conditions (given the target is moving or is sta-

tionary, subsequent to a hit or to a miss on the previous round, etc.).

More frequently, however, these phenomena represent human factors implic-

itly. Some implicit representations are directly relatable to human

capabilities. For example, the kill rates which are required as input to

some of the differential-equation-based models can be computed externally

as functions of measurable performance parameters such as firing times

and hit probabilities. Other implicit representations are more abstract

(i.e., they are non-emulatory). An example is the firepower score which

is used to determine attrition and movement in some aggregated models.

It is extremely difficult to predict how such a score would change as a

result of a specific change in an assumption about human factors in

maneuver unit combat.

The data used to represent human performance phenomena frequently

are drawn from non-combat situations. For example, firing accuracy data

for maneuver unit weapons are generally collected on firing ranges. It is

not known to what extent these data are representative of human perfor-

mance in combat situations. Furthermore, the individuals used to gener-

ate such data may not be typical of those who would perform similar tasks

15

...,V. - -..... •. , ,. •. . . • ., ," .

 in combat. (They are sometimes preselected for their ability to perform

the task for which the data are being collected.) Finally, the data used

in current representations of human performance phenomena in combat

models generally do not take into account differences among individuals

(or even between opposing forces) due to training, combat experience,

- cultural background, length of service, etc.

Decision making phenomena in combat models are represented in vari-

ous ways. Some models require decision rules or decision tables as

input; others use built-in decision rules, possibly in conjunction with

some input parameters which allow the user some control of decision

thresholds. In some cases, decisions are represented implicitly by their

effects. (Representation of maneuver unit combat results using firepower

scores is an example of this -- the effects of many low-level decisions

are subsumed into a simple equation involving a single index of firepower

potential for each side.) Some of the models are games in which many

decisions are made by humans interacting with other parts of the model.

Most frequently, a given model will use some combination of these tech-

niques to represent decision making. It should be noted that in all

models, regardless of type or level, decision making is never aggregated

away completely -- some decision making is always represented explicitly.

Decision rules in models are most frequently based on doctrine. The

extent to which doctrine will be followed in actual combat is, of course,

unknown. Some decision rules are based on heuristic formulations which

may or may not seem reasonable to a tactician or a behavioral scientist.

In general, decision rules do not include effects of the characteristics

of individual commanders on the decisions or on the ability to implement

them. They also tend to ignore the impact of incomplete information and

16
 uncertainty on decisions and tend to take very limited account of the
impact of external factors (such as anticipated weather conditions) on

decisions.

Some people would argue that the limitations of decision rules in a

model can be overcome by replacing them with a person who makes the

necessary decisions. However, there are several limitations of existing

games (and garners) which tend to reduce the realism of decision making

phenomena in models, even when humans are making the decisions. First,

in many games the information provided to a gamer differs from that

available to a commander in combat. This is particularly true of open

games, in which the players representing both Red and Blue have access to

the same information and to the decisions made by each other. (Even in

some closed games, the garners know in advance the area in which the major

penetration attempt will be made since the scenario threat operational

concept is continually replayed.) Secondly, the decisions required of a

gamer may differ from those required of a commander in combat. Some

games, for example, require a single gamer to make decisions which would

be made by several commanders in a real combat situation. Furthermore,

the sequence and relative timing of game decisions may differ from that

found in combat. Some games, for instance, allow the decision maker to

intervene only at prespecified times during the simulated battle (e.g.,

once every six hours). The gamers themselves may not be representative

of commanders in combat, and US garners certainly do not have the perspec-

tive, training, and cultural background of the commanders of Red and Blue

ally forces. It is impossible to subject garners to all the real stresses

of combat (such as fear, fatigue, and knowledge of the "real" effects of

decisions), although some recent games do attempt to subject the players

17

' " . . . .

•. , . . , , -, ., • , - . . . . .- " .
 L77 T T t_ T

to realistic time pressures. Finally, unrealistic game rules or assess-

ment logic can result in garners "playing the game" rather than making

realistic decisions. (For example, a gamer's knowledge of a fixed delay

imposed on the opposing force when a barrier or minefield is encountered
allows that gamer to plan around the delay whose duration he knows in
advance.)
The effects of organizational coordination and battlefield condi-
tions are generally not well represented in combat models. Organiza-
tional coordination is generally assumed to be perfect, and the effects
of fear, fatigue, and confusion caused by battlefield conditions are
usually not represented. One exception is that most models include some
simple representat-ion of the suppressive effects of fire. However, there
exists no experimental basis for these representations -- the suppression
process is not sufficiently well understood for there to be any assurance
that it has been modeled properly.5
In summary, human factors pervade combat, and combat models tend to
reflect this pervasiveness, at least implicitly. However, because of a
lack of knowledge concerning the details of human interactions in a com-
bat environment, models (and their data) tend to rely on idealized per-
formance to represent behavior, simplistic models of behavior, implicit
representation of behavior within more abstract model constructs, doc-
trine as a proxy for behavior, and the use of garners to simulate behav-
ior. The ease with which these human factors representations can be
Improved in a particular model depends in part on the overall model

5 Fora detailed discussion of the suppression process, including a

description of the current state of knowledge about that process and a
proposed research program to improve the state of knowledge, see
[ODCSRDA, 1975].

18

~ . ..
. .j
structure. For example, models with detailed time resolution can more

easily reflect human factors effects on the relative timing of events

than can models with a more highly aggregated treatment of time. This

suggests that the nature of the changes required to improve human factors

representations in models will vary in part as a function of model struc-

ture. In some cases, new submodels can be relatively easily added using

existing model structures; in other cases some model restructuring may be

required to achieve proper representation of human factors.

2.2 A TYPOLOGY OF HUMAN FACTORS AREAS

The summary of the previous section makes it clear that human fac-

tors are not well represented in combat models. In order to develop an

understanding of the nature of the human factors research required to

improve human factors representations, task 2 of the project involved

restructuring the information assembled in task 1 in a form that is mean-

ingful to behavioral science researchers. Section 2.2.1 outlines the way

in which this was done, while section 2.2.2 summarizes what was learned

from this exercise.

2.2.1 APPROACH

The first step in generating the typology of human factors areas was

to develop a set of generic tasks or groupings of related tasks encompass- -

ing the major human functions included in combat models. These tasks and

groupings of tasks will be referred to as task clusters. The following

18 task clusters were identified:

(1) vehicle maneuver,

(2) dismounted weapon maneuver,

19
 ,j
-, + ,' _6
I, _ _ - . .w ; . ~ - . - . + . +.
-. . - , + , . _ - . • .. ... . .. . . .. °.____

(3) visual acquisition,

(4) aiming and tracking,

(5) weapon operation,

(6) individual decision making (threat),

(7) individual decision making (non-threat),

(8) command decision making,

(9) command planning,

(10) radar acquisition,

(11) target and intelligence development,

(12) communication,

(13) driving,

(14) aircraft operation,

(15) construction,
(16) maintenance and repair,

(17) medical care, and

(18) resupply.

*. Detailed definitions of these task clusters can be found in appendix B.

Next, those task clusters involved in each of the model process phe-

nomena were identified. For example, the phenomenon "initial deployments

of weapons" within the model process "maneuver unit combat" involves task

clusters 1, 2, and 9 from the above list. The complete set of task

cluster-phenomenon associations is listed in appendix B. Identification

of the task clusters and association of them with the phenomena were

based on the knowledge of project staff members with backgrounds in mili-

tary operations and combat modeling.

For each of the 18 task clusters, an abilities analysis was con-

ducted to determine the types and degrees of abilities required to

20

+°.,' . •+ --. • . - . • .
+
-, ' .+, + "
,'•" ".'. °'+° °+" "'. +'. ,i •.' '. . • +
 perform the associated tasks. The approach used involved a taxonomy of

abilities and an assessment scheme developed by Fleishman [1975] (see

also [Theologus, Romashko, and Fleishman, 1970]). A list of the 37 abil-

ities included in the Fleishman taxonomy is given in exhibit 2-1. The

assessment scheme required that, for each task cluster, an estimate be

made of the degree to which each ability is required for errorless per-

formance of the associated tasks. Definitions of the abilities and a

more detailed description of the assessment scheme can be found in

appendix B.
• The abilities analysis was performed by the project staff using

information found in Army field manuals, literature describing previous

abilities analyses (for example, [ARI, 1979]), and knowledge and exper-

ience of the project team. The results of this relatively cursory

assessment do not constitute a detailed abilities analysis of all aspects

of human factors in combat. However, the results do provide a general

indication of the ability levels which must be represented explicitly or

implicitly in combat models. This indication provides a means of deter-

mining which abilities appear most significant in combat model represen-

tations and a way to relite improved knowledge about abilities to

improved representations of combat model processes. Details of the abil-

ities analysis can be found in appendix B; a summary of the results of

L

the analysis appears in the next section.

2.2.2 RESULTS

While the primary purpose of this task was to generate information

useful in structuring a research program to improve human factors repre-

, sentations in combat models, some information was gleaned from the task

which provides an indication of the most prevalent human factors areas in

21

*.• , . ,.--.% -, ...........

,- -... - - .... . .. ., . ..
 EXHIBIT 2-1: FLEISHMAN ABILITY TAXONOMY

1. VERBAL COMPREHENSION 20. STATIC STRENGTH

2. VERBAL EXPRESSION 21. EXPLOSIVE STRENGTH

3. IDEATIONAL FLUENCY 22. DYNAMIC STRENGTH

4. ORIGINALITY 23. STAMINA

5. MEMORIZATION 24. EXTENT FLEXIBILITY

r.
6- PROBLEM SENSITIVITY 25. DYNAMIC FLEXIBILITY
7. MATHEMATICAL REASONING 26. GROSS BODY EQUILIBRIUM

8. NUMBER FACILITY 27. CHOICE REACTION TIME

9. DEDUCTIVE REASONING 28- REACTION TIME

10- INDUCTIVE REASONING 29. SPEED OF LIMB MOVEMENT

11- INFORMATION ORDERING 30. WRIST-FINGER SPEED

12. CATEGORY FLEXIBILITY 31. GROSS BODY COORDINATION

13. SPATIAL ORIENTATION 32. MULTILIMB COORDINATION

14. VISUALIZATION 33- FINGER DEXTERITY

15- SPEED OF CLOSURE 34. MANUAL DEXTERITY

16- FLEXIBILITY OF CLOSURE 35- ARM-HAND STEADINESS

17. SELECTIVE ATTENTION 36- RATE CONTROL

18. TIME SHARING 37. CONTROL PRECISION

19- PERCEPTUAL SPEED

-1
22

,......"....'-... .. . . . - .- - . . .. , . ° ° ,...--,
combat models. Appendix B includes tables presenting summary statistics

collected during the abilities and task cluster analyses described in the

previous section. From these tables, it is clear that task clusters 6

(individual decision making [threat]), 8 (command decision making) and 9

(command planning) are the most prevalent of the 18 task clusters; and

that abilities 6 (problem sensitivity), 17 (selective attention), 19

(perceptual speed), and 27 (choice reaction time) are the most prevalent
of the 37 abilities. These statistics tend to indicate that the most

prevalent human factors areas in combat models involve:

(1) decision making; and

(2) performing tasks under adverse conditions (in particular, con-

ditions in which problems are likely, there is distracting

stimulation, or there are significant time pressures).

2.3 ARSAS OF MODEL SENSITIVITY

The first two tasks of the project provided information describing

how human factors representations interact with combat model processes.

This information was used to assess what research would be required to

improve the human factors representations. However, in order to priori-

tize such research, it would be useful to have information concerning the

relative sensitivity of combat models to potential changes in the human

factors assumptions. If simulated battle results are relatively insen- -

sitive to potential changes to human factors assumptions in a given area,

research in that area might be of lower priority than in an area in which -

the potential sensitivity appears to be great.

The purpose of the third task of this project was to identify those

model process phenomena for which changing human factors assumptions

23
 could have a significant impact on model output. Section 2.3.1 below

describes the approach used to assess this sensitivity, while section

2.3.2 summarizes the results of the sensitivity analysis.

2.3.1 APPROACH

Two types of sources were used to generate information about the

potential impact of human factors assumptions on simulated combat

results. First, sensitivity runs of combat models were made to address

specific questions concerning model sensitivity. Two of the nine models

described in section 2.1.1 were used for this purpose: BLDM (with which

a total of 42 runs was made) and VECTOR-2 (with which 21 runs were made).
Results of the excursions are summarized in appendix C. Secondly, model

runs performed for previous studies were analyzed to develop additional

insights into the sensitivity of models to human factors assumptions.

Summaries of these runs also can be found in appendix C.

It should be noted that, because of the large number of human fac-

tors areas investigated, the results of this task are necessarily based

on a limited number of runs in any one area. The results are also sub-

ject to the effects of the specific scenario assumptions employed

(although an attempt was made to use scenarios typical of those used in

current Army studies). Furthermore, it was not feasible to investigate

to any significant extent the interactions resulting from changing mul-

tiple human factors assumptions simultaneously. Also, it was not always

clear to what extent a parameter or assumption should be varied to en-

compass the entire range of potential effects of changing a human fac-

tors assumption. For these reasons, the results of this task are not

presented as a comprehensive analysis of the potential impacts of human

24

.. . . . .. . .* . . . .. .. . . ,
 factors in all situations. However, the results do give a general indi-

cation of the relative potential sensitivities in scenarios typically

used in studies and in models which represent the current state of the

art of representing human factors effects.

2.3.2 RESULTS

The material in appendix C presents the relative sensitivity of com-

bat model output to changes in human factors assumptions related to the

various model process phenomena identified in task 1. Those areas which

6
appeared to be associated with the highest sensitivities are:

(1) deployment and maneuver of maneuver unit weapons, including

phenomena involving terrain area selection, defender position

selection, and attacker route selection;

(2) reserve commitment timing, including decisions to commit

reserve or second echelon forces, movement rates of committed

reserve units, and delays in arrival of committed forces due to

barriers, minefields, etc.;

(3) firing times in maneuver unit combat; 7 and

(4) suppression of maneuver unit forces (both microscopic

". "pop-down" effects and more macroscopic effects, which might

cause a portion of a force to avoid participation in an engage-

ment altogether).

6 Thesefour areas are groupings of the 13 phenomena to which combat

models have high sensitivity, as indicated in exhibit C-17 of appendix C.
7This refers to the time required to fire a round at an available tar-
. get. One can view the maneuver unit combat process as an attempt by
each side to generate the highest pqssible rate of attriting weapons on
the opposing side. Other things being equal, a side's attrition rate is
approximately inversely proportional to the averaging firing time.

25

...... . ..4. .. . .. ...... . .. ... ....

°o . .. .... . . . .. " i
. ,L " .-... .- . . . . . .- - -Z ' -' - . .- - . - - -- " . -' ' - -'-" - - . ° -'" ' " -' - - .- -

Note that all of these areas are closely associated with the maneuver

unit combat process, leading to the conclusion that representation of

human factors effects in maneuver unit combat is a highly significant

contributor to simulated battle results.

In several areas there appeared to be very little sensitivity of the

model results to changes in human factors assumptions. Two of the more

interesting of these areas are those involving delays in communication or

in decision making (unless the delays become very long, i.e., several

hours or more), and those involving target acquisition capabilities

(whether by maneuver unit weapons, attack helicopters, counterbattery

radars, Copperhead designators, or air defense radars). 8 This latter

observation probably is related to the fact that current scenarios tend

to result in a target-rich environment. In such an environment, moderate

changes in the ability to acquire a target appear to have little impact,

since there are plenty of targets available.

Several general observations can be made concerning the impact on

model output of human factors assumptions. First, higher-resolution,

lower-level models tend to be more sensitive to a particular human fac-

tors assumption than more aggregated, higher-level models. This is

simply because a given phenomenon is a greater portion of the overall

model content in a lower-level model than in a higher-level one. Thus,

the details of the representation of a particular phenomenon are of vary-

ing importance, depending on the highest echelon represented by the model.

Secondly, human factors assumptions are likely to have a greater effect

if the opposing forces are evenly balanced. In appendix C, two examples

8 See exhibit C-17 of appendix C.

26
 are given in which unit breakpoints (the criteria used in a model to

determine when a maneuver unit will disengage from combat) were varied.

In one of the examples, in which the forces were evenly balanced, moder-

ate changes in the breakpoints had an extreme effect on the results,

sometimes reversing the course of the battle. In the other example, a

moderate change in the defender's breakpoint had little crfect, but this

was in a situation in which the defender was already losing badly.

Thirdly, model results tend to be less sensitive to symetric changes in

human factors assumptions (i.e., when the same change is applied to both

opposing forces rather than just to one side). Compare, for example,

exhibits C-5 and C-6 of appendix C, which display the effects of one-

sided and symmetric changes (respectively) to the number of participants

in an engagement. However, in some situations model results can be very

sensitive to symmetric changes in human factors representations. For

example, the first of the above-mentioned studies in which breakpoints

were varied involved symmetric changes in the breakpoint criteria, and

the effect was highly significant.

Finally, it should be stressed that this sensitivity analysis inves-

.4.
tigated the potential impact of separate changes to each of many assump-

tions about human factors effects in combat models. Some of these indi-

vidual effects were found to be highly significant -- a number could

cause a reversal in the predicted winner of a simulated engagement. No

analysis was conducted of the potential effect of the simultaneous

improvement of human factors representations in a whole spectrum of model

phenomena. However, the observed impact of some of the in.A.:dual

changes in assumptions suggests that a general improvement in the treat-

ment of human factors in combat models would be likely to have a very

large and unpredictable effect on simulated battle results.

27

r"
- % . ' ° , •. ... . . " - , , . . - . - . • ..
. • •. - . • • . .-
 3.0 STRUCTURE AND PRIORITIES FOR HUMAN FACTORS RESEARCH

The previous chapter described the information gathered to under-

stand current human factors representations in combat models. This

information is used to structure a framework for conducting research that

will provide knowledge to improve those representations. Exhibit 3-i

summarizes the nature of this information: Models consist of processes

which, in turn, include phenomena that are influenced by human factors.

A list of such processes and phenomena has been developed, a description

of the manner in which each phenomenon is represented in selected models

has been generated, and information has been assembled concerning the

sensitivity of models to assumptions about human factors interactions

with the phenomena. A phenomenon involves one or more task clusters com-

prising tasks performed by humans. A list of such task clusters has

been produced for each phenomenon. Tasks require abilities for their

successful performance. As assessment has been conducted of the degree

to which various abilities are required for each task cluster.

The purpose of the fourth task of this project was to develop a

logical structure for this information and to use the information to

identify high priority research areas. The structure is used to organize

a research program to acquire the knowledge needed to improve the repre-

sentation of human factors in combat models. This chapter describes this

structure (in section 3.1), describes the development of research prior- -

ities (in section 3.2), and presents and discusses the priorities (in

section 3.3).

28
* EXHIBIT 3-1: TYPOLOGY FOR HUMAN FACTORS REPRESENTATIONS IN COMBAT MODELS

COMBAT
MODEL

MODEL MODEL
PROCESS * PROCESS

RCS

ROESPOES
J RCS

HE4INNPENMNNPEOENNPEOEO

'MODE

2TUCUR
1-- -- - - - - -- - - - - - - - -

4.
.HUMAN 4.-. * . . . **

FACTORS. -.
 3.1 DEVELOPMENT OF THE RESEARCH STRUCTURE

i! Research to improve human factors representations in combat models

can be viewed as providing functional relationships for predicting task

performance and abilities levels. Conceptually, the research would be

designed to develop sets of relationships involving the following

parameters:

""= level of ability i for an individual (i=-, 37);

"ai = change in level of ability i (i=1, ... 37);

Pj : level of performance by an individual (or crew or other small

group of individuals) of tasks in the context of model

phenomenon j;

Apj = change in level of task performance;

d = set of demographic descriptors of an individual (e.g., ASVAB

scores, age, sex, education);

e = set of descriptors of the combat environment (e.g., weather

conditions, intensity of combat, length of time a force has

been engaged); and

t = characterizations of an individual's training and experience

(e.g., time spent in training for a certain task, amount of

previous combat experience).

Either of two general approaches to human factors research can be

taken to provide the needed relationships. The first approach would

allow for more of a scientific basis for the research, but would also
require a longer-range research program having more of a basic research

content. This approach would involve four general k~ids of human factors

research:

30

S.
 (1) research to determine functions to predict task performance in

the context of a given model phenomenon (e.g., to predict

performance of the tasks in the task cluster "aiming and

tracking" in terms of the resultant accuracy of fire in

maneuver unit combat) as a function of ability levels, i.e.,

development of the functions gj, where

Pj = gj(a,, ... , a37 );

(2) research to relate abilities to the demographic descriptors,

i.e., development of the functions hi, where

a i = hi(d);
(3) research to assess the effect of the combat environment on

abilities, i.e., development of the functions ri , where

= rie);

(4) research to establish the effect of training and experience on

abilities, i.e., development of the functions si , where

Aai si(t).

The first of these types of research would develop relationships

either for direct use in combat models or to provide means of generating

model inputs. The second type of research would involve

(1) developing quantitative measures for the abilities where these

do not already exist;

(2) developing methods to predict ability levels from measurable

human characteristics;

31

............................... .... ,L" ::

I "-.'
%,' ...
-,'
/ ."._
."-. ...' .. "....- -" ..-.
'..' """ " -". - - " - " " . .".. '- " " "
. . . -t
 (3) determining the distribution in the Army of ability levels

among individuals (so that average values of P. can

9
eventually be assessed for use in combat models); and
(4) identifying subjects for use in the first type of research (to

develop the gj functions).

The third and fourth types of research will allow adjusting the basic

ability levels to account for actual combat conditions and for the train-

ing and experience of individuals likely to be involved in combat. In

reality, there is probably some interaction among demoqraphic descrip-

tors, combat environment, and training and experience in the determina-

tion of ability levels, but the effects are represented here as though

they are independent in order to assure that the research program will be
manageable.

The second general research approach would eliminate explicit treat-

ment of abilities from the research program and would instead develop

methods to predict task performance directly from demographic descrip-

tors. This approach, while having less of a phenomenological structure

than the first approach, would be useful in providing shorter-term

results for use with combat models. The resultant relationships might

later be improved via lonqer-term research structured around the first

approach. For example, to assess the impact of commander training on the

9 f course, for some applications of combat models, such as investiga-

tions of the effects of human abilities on combat effectiveness, use of
average values of Pj would not be appropriate.
10 The
Army might wish to conduct this type of research in-house in order
to assure that a proper mix of Army personnel with appropriate demo-
graphic characteristics is used as subjects in the first type of
research. It might be efficient and effective to use the same subjects
for development of both the gi functions and the hi functions.

32
* . -. ,.....,.....- ...... . .. . . .- L L-- ..- . , -...-. . - -. - - -. ' ---..

quality of command decision making, it might be impractical to relate a

commander's abilities to his training (type 4 research) and then to

determine decision making effectiveness as a function of ability levels

(type 1 research). It might be easier to analyze the impact of commander

and unit training directly on unit performance in an experimental setting

such as the National Training Center, although some information will be

lost as a result of taking this more efficient approach. For this second

research approach, three general kinds of human factors research would be

needed:

(1) research to determine functions to predict task performance in

the context of a given model phenomenon as a function of demo-

graphic descriptors of the individuals performing the tasks,

i.e., development of the functions Gj, where

Pj Gj(d) •

(2) research to assess the effect of the combat environment on task

performance, i.e., development of the functions Rj, where

"Pj = Rj(e) ;
(3) research to establish the effect of training and experience on

task performance, i.e., development of the functions Sj,

where

Pj = Sj(t)

3.2 DEVELOPMENT OF RESEARCH PRIORITIES

From the point of view of improving combat models, research to

predict task performance from demographic descriptors is considered

essential, while research to assess the effect of the combat environment

is of slightly higher priority than research on the effects of training

33

. . . . . .
 and experience. (For other purposes, e.g., determining training require-

ments, research on training effects may be more important than research

on environmental effects.) In order to prioritize research efforts

within type of research (for either of the two research approaches), it

would be useful to relate the research to the ultimate potential impact

of conducting it. Exhibit 3-2 provides a conceptual representation of

the impact on Army decision making of human factors research for improv-

ing combat models. As the exhibit illustrates, human factors research

will result in an improved capability to predict human behavior in

combat. Incorporation of the resulting prediction methods into combat

models will result in model output which is more accurate and more sensi-

tive to important human characteristics and interactions. Proper use of

this output in Army decision making should result in better Army deci-

sions concerning system procurement, force structuring, training,

recruiting, etc.

The discipline of decision analysis provides a framework within

which to relate the items in this chain of activities so as to determine

(at least conceptually) the potential benefit from improved Army deci-

sions resulting from human factors research in a given area. Consider

an oversimplified situation in which the Army wishes to select from among

several equal-cost alternative new weapon systems. Suppose the Army

desires to select that system which maximizes the probability that a US

battalion conducts a successful defense. A combat model (in this case, a

battalion-level Monte-Carlo simulation) can be used to estimate this

llFor a description of the way in which decision analysis can be used to

structure problems of this general type see, for example, [Raiffa, 1970]
or (Holloway, 19791.

34

.l" ° m N°. . ...,. * . ..... . . .-. ° ° . . . o .. '

 EXHIBIT 3-2: IMPACT ON ARMY DECISIONS OF HUMAN FACTORS RESEARCH FOR
COMBAT MODELS

Human factors research

Improved prediction
of human behavior
in combat

Revisions to
combat models

More "accurate"
model output

Use of model output

in Army decisions

Improved decisions
-ii

a...

o4'

35

S... ..... .... ..

 probability with each of the alternative systems. However, the capabil-

ity of the model to predict this probability is limited by the degree to

which human factors interactions are properly represented in the model.

If Pj is the level at which the model represents the performance of a

single human task, and selection of alternative k results in a weapon

system whose performance can be characterized by the parameter Wk, the

model result (in this example, the probability of a defender win) can be

expressed as a function of Pj and Wk:

o(Pj,Wk) - predicted probability the defender wins given human

performance level Pj and selection of alternative k.

Current uncertainty concerning the true value of Pj can be represented

by a probability distribution:
f(pj) probability density function for human performance

parameter Pj.
Human factors research concerning Pj can be viewed as reducing this

uncertainty. Without such research, the existing uncertainty degrades

the capability to select a good alternative system, so that (in

principle) the Army must select that k which maximizes the expected

benefit of the system:

max f o(pj,Wk) f(pj) dpj.

Pj
The maximum potential effect of human factors research is to eliminate

- all uncertainty about Pj, thus allowing the Army to select the best

system without degradation in the decision due to uncertainty about Pj.

Prior to conducting this human factors research, the estimate of the -7

potential expected benefit from first performing research on Pj and

then selecting the best system is

f max O(Pj,Wk) f(pj) dpi.

Pj k

36
 Thus, the potential benefit to the Army's decision concerning this weapon

system resulting from research on Pj is12

f max P(Pj,Wk) f(pj) dpj - max f P(jk)fp)d.
pj k k Pj
While this decision-theoretic view is a useful way to conceptualize

the potential benefit of human factors research for combat models, it is

not practical to use this formal approach to set research priorities.

The approach requires the computation of the benefit of research in the

context of a specific Army decision problem or a specific set of decision

problems. The large number of decision problems to which combat models

may be applied in the future would make the approach unfeasible, particu-

larly since most of these problems have not yet even been identified.

Because of this problem (and other only slightly less formidable

problems, such as the difficulty of assessing the p function for all

values of all ai and all Pj), a more heuristic approach has been
taken to set priorities within each of the types of research, based on

information developed in tasks 1, 2, and 3 of this project. In developing

priorities for investigating task perfor ance to predict Pj or APj

(i.e., to determine the gj, Gj, Rj, or Sj functions), two approaches were

taken. The priority scheme which is more appropriate will depend on the

way specific human factors research projects are to be structured. The

first approach involved developing priorities for the model process

phenomena. Such priorities should be useful in selecting research

,. projects which emphasize the performance of tasks in the context of one

or a few related phenomena. The second approach involved developing

1 2 Indecision theory parlance, this expression gives the "expected value

of perfect information" about P.
P

37
priorities for the 18 task clusters. These priorities should assist in

selecting research projects which emphasize the performance of one or

more related tasks generally. In such projects individual research

efforts would still involve investigation of task performance in the con-

text of individual process phenomena, but with some hope of obtaining

insights more generally applicable to phenomena not specifically investi-

gated. Examples of both types of research projects are given in the next

section.
In developing priorities for the phenomena, the following criteria

(listed in order of importance) were used:

(1) sensitivity of combat model results to changes in human factors

assumptions;
(2) current model fidelity in representing the phenomenon (used as

a proxy for existing knowledge about human factors interactions

with the phenomenon); and

(3) the degree to which the various abilities are required for

successful performance of the tasks associated with the

phenomenon (used as a measure of the degree to which human

factors influence the phenomenon).

The following criterion was used as the basis for prioritizing task clus-

ters: the number of phenomena that require the task and which signifi-

cantly influence model output. The three types of research regarding

prediction of abilities (i.e., to determine the hi , ri , and si func-

tions) were prioritized using the following criterion: the relative

degree to which each ability is required for tasks associated with phe-

nomena that significantly influence model output. All three sets of pri-
orities were developed using quantitative (though heuristic) techniques

38
. .....- . v_. . . . . .. "" -

based on the information developed in tasks 1, 2, and 3 of the project.

Details concerning the development of all priorities can be found in

appendix D; the priorities themselves are described in the next section.

3.3 RESEARCH PRIORITIES AND OBSERVATIONS

The previous two sections described a structure for organizing and

prioritizing research to improve the information base for representing

human factors in combat models. This section presents and discusses the

priorities derived from use of the structure.

Exhibit 3-3 indicates the relative ranking of all the model process

phenomena in terms of the need for research on task performance in the

context of individual phenomena (with low numbers indicating a compara-

tively greater need for such research). The highest priority phenomena

are listed in priority order in exhibit 3-4. Exhibit 3-5 provides the
relative ranking of the 18 task clusters in terms of the need for

research on task performance generally. The highest priority task clus-

ters are listed in priority order in exhibit 3-6. Exhibit 3-7 gives the

relative ranking of the 37 abilities in terms of the need for research to

develop the hi , ri, and si functions. The highest priority abilities

are listed in priority order in exhibit '-8. It should be noted that,

for research to develop ri and si , the prioritization of abilities

could be further refined by adding consideration of a priori estimates of

the effect of the combat environment on the abilities (for ri) and the

effect of training and experience on the abilities (for si).

39
L -

EXHIBIT 3-3: RELATIVE RANKING OF M4ODEL PROCESS PHENOMENA

OVERALL
PROCESS PHENOMENON RANK

Maneuver Unit Combat:

1. Force participation 5 (tie)
2. Initial deployments of weapons 5 (tie)
3. Maneuver patterns of individual I
weapons
4. Target acquisition capabilities 56
5. Target identification capabilities 17
6. Target selection (and target
switching) 27 (tie)
7. Decision to fire 27 (tie)
8. Ordnance selection 46
9. Time to fire or firing rate 11
10. Accuracy of fire 24
11. Fire suppression 4
12. Employment of obscuration 37 (tie)

Artillery Fire:
1. Ordnance selection 16
2. Target element acquisition and
selection (if point fire) 48
3. Decision to fire 51 (tie)
4. Delivery pattern 28
5. Delivery accuracy 25
6. Target posture 19
7. Suppressive effects of fire 18 (tie)
8. Effects of fire on communications
and decision-making ability 12 (tie)
9. Time to fire or firing rate 26
Air-To-Ground Attacks:
1. Ability to acquire (or reacquire)
the target 52
2. Engagement decision (whether to
attack the target) 51 (tie)
3. Ordnance selection 42
4. Target element acquisition and
selection (if point fire) 34
5. Aircraft standoff distance and/or
maneuver pattern over target 49 (tie)
6. Weapon delivery accuracy 39
7. Target posture 30
8. Suppressive effects of the attack 18 (tie)
9. Effects of the attack on communica-
tions and decision-making ability 12 (tie)

-- Continued --

40

%,,~~............'.'.".-.-.-.....-...-...-..."......................
. -". ._. ......
" a A
 EXHIBIT 3-3: RELATIVE RANKING OF MODEL PROCESS PHENOMENA
(Continued)

PROCESS PHENOMENON OVERALL

RANK

Air-To-Ground Attacks (cont.):

10. Effects of the attack on air-
field sortie generation capability
(if target is an airfield)
11. Aircraft disengagement decision 51 (tie)
Air-To-Air Combat:
1. Engagement decisions (whether to
engage and type of engagement)
2. Aircraft maneuver patterns
3. Target acquisition capabilities
4. Target selection
5. Ordnance selection -
6. Time to fire -
7. Accuracy of fire -
8. Abort decisions -

Air Defense:
1. Coordination of fire 33
2. Target acquisition capabilities 57
3. Target selection 41 (tie)
4. Decision to fire 51 (tie)
5. Degraded capability of aircraft to
destroy targets on the ground 35 (tie)
6. Aircraft mission abort decisions 41 (tie)
Mobility, Countermobility, and Survivability:
1. Location, density, and extent of a
minefield or obstacle 7
2. Attrition suffered when encounter-
ing a minefield 43
3. Delay in movement when encountering
a minefield or obstacle 2
4. Reduction in movement rates due to
minefields and obstacles 6
5. Increase in exposure due to
encountering a minefield or obstacle 32
6. Delays in traveling on damaged roads 3
7. Delays in crossing rivers or gaps 8
8. Degree of protection afforded by a
prepared defensive position 20
9. Location and extent of a prepared
defensive position 22
4L

-- Continued --

41

"~~~~~.'...'..
...... '...-"'."". ...... ,. - .......- "" 1i.
,. v-l. - fl . -. *. %r L,-.-...
. .T. .........

EXHIBIT 3-3: RELATIVE RANKING OF MODEL PROCESS PHENOMENA

(Continued)

PROCESS PHENOMENON OVERALL

RANK
Intelligence and Electronic Warfare:
1. Collection management decisions -
what information to pursue 14 (tie)
2. Collection management decisions -
how to collect desired information 37 (tie)
3. Target acquisition capabilities 29
4. Accuracy of target identification
(including false targets) 36
5. Target processing time 13 (tie)
6. Target location error 13 (tie)
7. Timeliness of intelligence
processing 44 (tie)
8. Accuracy of intelligence processing 44 (tie)
9. Completeness of intelligence
10:processing 44 (tie)
10. Decisions to jam 14 (tie)
11. Effectiveness of jamming 15
12. Collection management decisions -
location of sensors 50 (tie)

Communications:
1. Time to transmit message 55
2. Message content 40 (tie)
3. Transmission accuracy 40 (tie)
4. Reception accuracy 40 (tie)
Mission Assignment, Organization for Combat,
and Resource Allocation:
1. Initial deployments of maneuver
units 9 (tie)
2. Changes in these deployments 23 (tie)
3. Cross attachments among maneuver
units 45 (tie)
4. Allocation and/or positioning of
attack helicopters 45 (tie)
5. Allocation and/or positioning of
field artillery forces 45 (tie)
6. Allocation and/or positioning of
CAS/BI aircraft 45 (tie)
7. Allocation and/or positioning of
air defense forces 45 (tie)
8. Allocation and/or positioninq of
target acquisition resources 37 (tie)
9. Allocation and/or positioning of
supply points 45 (tie)

-- Continued --

42

.:° . • ". • . - . . . •. . . . . . . .
 EXHIBIT 3-3: RELATIVE RANKING OF MODEL PROCESS PHENOMENA
(Continued)

PROCESS PHENOMENON OVERALL

RANK

Mission Assignment, Organization for Combat,

and Resource Allocation (cont.):
10. Allocation of aircraft to missions 45 (tie)
11. Delays in making and implementing
decisions 53

Maneuver Control:
1. Decisions to change the activity
and/or location of a unit 23 (tie)
2. Decisions to request and commit
reserves 9 (tie)
3. Decisions for unit retirement 4 (tie)
4. Decisions to request fire support 54 (tie)
5. Decisions to request engineer
support 23 (tie)
6. Decisions for the tactical
relocation of supporting elements 37 (tie)
7. Delays in making and implementing
decisions 50 (tie)

Fire Support Allocation:

1. Target and mission priorities 54 (tie)
2. Preferences for type and amount of
support against a given target in
given situation 23 (tie)
3. Preferences for types and amounts
of fire to be allocated to pre-
planned missions 23 (tie)
4. Criteria for ignoring tarqets or
dropping them from list of acquired
targets 50 (tie)
5. Delays in making and implementinq
decisions 23 (tie)

Resupply and Replacement:

1. Locations of supply points 45 (tie)
2. Stockpiling decisions at supply
points 45 (tie)
3. Quantities of supplies and replace-
ments allocated to units on the
battlefield 23 (tie)
4. Delays in decisions to provide
sup . es or replacements 45 (tie)

-- Continued --

43

"- . " ,~ .- - . ..- - -.- . . " :- . .. .- - - -

 EXHIBIT 3-3: RELATIVE RANKING OF MODEL PROCESS PHENOMENA
(Concluded)

PROCESS PHENOMENON OVERALL

RANK

Resupply and Replacement (cont.):

5. Time required to load and unload
allocated supplies 38

Movement:
1. Aircraft speeds and flight paths 35 (tie)
2. Movement rates of ground forces 10
3. Route selection of forces moving
on the ground 45 (tie)
4. Degree of exposure of moving forces
(air and ground) 31

Maintenance and Repair; Medical Support:

1. Rates at which weapons and
personnel are returned to duty
given availability of support 21
2. Availability of repair capability
and medical support 14 (tie)
3. Rate at which damaqed capability of
a facility is restored 47

Construction, Maintenance, and Operation of

Airfields:
1. Sortie generatinq capability of an
airfield
2. Rate of recovery of sortie
generating capability followinq an
attack on the airfield
3. Construction and repair rates for
runways, aircraft shelters, and
entire airfields
4. Delays in scrambling aircraft
5. Degree to which available shelters
are in use when an attack of an
airfield occurs
6. Aircraft basing rules

,.

"44

-44

• . ~ * . " . .' - -
, " ' . ., , I ,s La
, .
.,
- '' - ?
b
- '
..
'
.. ]..-' -
...... .. *...-., V . . .....
 EXHIBIT 3-4: HIGHEST PRIORITY PHENOMENA

RANK PROCESS AND PHENOMENON

1i Maneuver Unit Combat: Maneuver patterns of individual

weapons

2 Mobility, Countermobility, and Survivability: Delay in

movement when encountering a minefield or obstacle
3 Mobility, Countermobility, and Survivability: Delays in
traveling on damaged roads
4 Maneuver Unit Combat: Fire suppression

5 (tie) Maneuver Unit Combat: Force participation

5 (tie) Maneuver Unit Combat: Initial deployment of weapons

6 Mobility, Countermobility, and Survivability: Reduction in

movement rates due to minefields or obstacles

7 Mobility, Countermobility, and Survivability: Location,

density, and extent of a minefield or obstacle

8 Mobility, Countermobility, and Survivability: Delays in

crossing rivers or gaps
9 (tie) Mission Assignment, Organization for Combat, and Resource
Allocation: Initial deployments of maneuver units

9 (tie) Maneuver Control: Decisions to chanqe the activity and/or

location of a unit
10 Movement: Movement rates of ground forces
11 Maneuver Unit Combat: Time to fire or firinq rate

45

~~~~~~~~~~~~~~~~~~~.. ............-..
. .. ...... ..... -, .. :.;..-.... ... ; -i "
 EXHIBIT 3-5: RELATIVE RANKING OF TASK CLUSTERS

TASK CLUSTER RANK

1. Vehicle maneuver 3 (tie)
2. Dismounted weapon maneuver 5 (tie)
3. Visual acquisition 7

4. Aiming and tracking 6 (tie)

5. Weapon operation 5 (tie)

6. Individual decision making (threat) 3 (tie)

7. Individual decision making (non-threat) 8
8. Command decision making 2
9. Command planning I
10. Radar acquisition 11 (tie)
11. Target and intelligence development 6 (tie)
12. Communication 6 (tie)

13. Driving 4
14. Aircraft operation 9
15. Construction 6 (tie)
16. Maintenance and repair 10
17. Medical care 11 (tie)
18. Resupply 11 (tie)

46

----------------------------------
 EXHIBIT 3-6: HIGHEST PRIORITY TASK CLUSTERS '-

RANK TASK CLUSTER

1 9. Command planning

2 8. Command decision making

3 (tie) 1. Vehicle maneuver

3 (tie) 6. Individual decision making (threat)

4 13. Driving

5 (tie) 2. Dismounted weapon maneuver

5 (tie) 5. Weapon operation

6 (tie) 4. Aiming and tracking

6 (tie) 11. Target and intelligence development

6 (tie) 12. Communication

6 (tie) 15. Construction

44

477
* W -. - .. .. . , .- o, -- N . - . - ; ,, , -7..---- - .i -, ; - ,- - . ., ,~- . ,* .

EXHIBIT 3-7: RELATIVE RANKING OF ABILITIES

ABILITY RANK
1. Verbal Comprehension 12
2. Verbal Expression 18
' 3. Ideational Fluency 14
4. Originality 7
" 5. Memorization 9
6. Problem Sensitivity 2
7. Mathematical Reasoning 11
8. Number Facility 15
- 9. Deductive Reasoning 10
10. Inductive Reasoning 13
11. Information Ordering 16
12. Category Felxibility 19
13. Spatial Orientation 17
14. Visualization 8
15. Speed of Closure 4
16. Flexibility of Closure 6
17. Selective Attention 1 (tie)
18. Time Sharing 5
19. Perceptual Speed 3
20. Static Strenqth 26
21. Explosive Strength 28
22. Dynamic Strength 33
23. Stamina 24
24. Extent Flexibility 30
25. Dynamic Flexibility 29
26. Gross Body Equilibrium 34
27. Choice Reaction Time 1 (tie)
28. Reaction Time 21
29. Speed of Limb Movement 23
30. Wrist-Finger Speed 27
31. Gross Body Coordination 36
32. Miltilimb Coordination 25
33. Finger Dexterity 35
34. Manual Dexterity 31
35. Arm-Hand Steadiness 32
36. Rate Control 22
37. Control Precision 20

-4i 48

'"
' " '' * "" "' '' " " "' " " "' ' " " ' " " -" " . .' ' " " " " " "" " " - ." " ~ " -i "
 EXHIBIT 3-8: HIGHEST PRIORITY ABILITIES

RANK ABILITY

1 (tie) 17. Selective Attention

1 (tie) 27. Choice Reaction Time

2 6. Problem Sensitivity

3 19. Perceptual Speed

4 15. Speed of Closure

5 18. Time Sharing

6 16. Flexibility of Closure

7 4. Oriqinality

8 14. Visualization
9 5. Memorization

10 9. Deductive Reasoning
11 7. Mathematical Reasoning

49

'2t.
 For various reasons, these priorities and the previously described

research structure into which they fit are intended to provide general

guidelines rather than a rigid organization for research efforts. Listed

below are some mediating factors which should also be considered in

developing a research plan:

* In reality, there are probably overlapping priorities for work

among the types of research so that, for example, higher priority

research in the training and experience area is more important

than lower-priority research concerning representation of model

phenomena.

* Within a given type of research, budget constraints may make the

funding of several relatively inexpensive lower-priority items

more cost effective than partial funding of a high priority but

very expensive item.

• Consideration should be given to possible mutual benefits to

several items resulting from the funding of a single research

effort.

* There is the possibility of additional benefit from a particu-

lar research effort besides the acquisition of information for

improving combat models. (For example, research concerning the

fire suppression phenomenon might provide knowledge useful in

training soldiers to use suppression to their advantage.) Such

additional benefit might enhance the overall attractiveness of a

; particular research effort.

In spite of these caveats, the structure and priorities presented

earlier provide an overall framework for organizing human factors

research for improving combat models.. As an aggregated summary of these

50

. . . . . .• .
- •. o - , . .
. 0' '.o, -,- ''' " - -. " " .'
" .,
". '. ' " -- " ' . . " "•' . . '. - .. -" . ' " '" . -, . ". . '
 priorities, several general topics can be identified as high priority

research areas. Concerning the model phenomena, the highest priority

area involves research on the deployment, employment, movement, and

maneuver of maneuver unit forces (i.e., phenomena with ranks 1, 5, 9, and
13
10, as listed in exhibit 3-4). Because of its potential effect on
this most important area, research concerning engineer operations (mobil-

ity, countermobility, and survivability) also is a high priority area

(phenomena with ranks 2, 3, 6, 7, and 8). A third topic of high priority

involves research on fire suppression and related phenomena (phenomena

with ranks 4 and 5).14 Concerning task clusters, by far the highest

priority research involves the broad area of command, control, communica-

tions and intelligence or C31 (corresponding to task clusters with

ranks 1, 2, and 6, as listed in exhibit 3-6). Another important area

involves the use of weapons under combat conditions, including aiming,

13Two independent sources tend to confirm this result. ARI recently

conducted a survey of combat arms officers with previous experience as
. battalion commanders and currently assigned as faculty members at the
Command and General Staff College. When presented with the list of
model processes and phenomena developed in this project, the officers
identified mission assignment (and related command and control func-
tions) and maneuver unit combat as the two most significant combat pro-
cesses in terms of the quantity and quality of human involvement re-
quired. Within the mission assignment process, the two most signifi-
cant phenomena were identified to be (1) initial deployments of ma-
. neuver units and (2) changes to the activity and/or location of a unit.
Within the maneuver unit combat process, the three most significant
phenomena were identified to be (1) force participation, (2) maneuver
_ patterns during an engagement, and (3) initial deployments of weapons.
In separate discussions with General William DePuy (US Army, Retired),
tactical deployment and maneuver of forces were identified as the
overriding areas affecting the outcome of actual combat.
14 Fordetailed suggestions for a research program in this area, see
[ODCSRDA, 1975].

51

..,,
:.-.:,.-..-.....
;...-.- ....-
.. . . .> i :.* .. ; . ......... ...
** -.-.. ;..,.-.,.-....-.......................... ......... .....
.

15
tracking, loading, and firing (task clusters with ranks 5 and 6).
Examination of high priority task clusters also tends to confirm the

above-mentioned need for research in the area of weapon deployment and

maneuver (task clusters with ranks 3, 4, and 5). Concerning abilities,

those involving the performance of tasks under adverse conditions (e.g.,

with distracting stimuli or under time pressure) appear to be most

important (i.e., abilities with ranks 1, 2, 3, and 4, as indicated in

exhibit 3-8). Also highly important are abilities required in decision

making (abilities with ranks 1, 2, 7, 10, and 11) and those involving

recognition and visualization of specific features and patterns (such as

would be required in identifying targets or performing templating tasks

-- abilities with ranks 3, 4, 6, and 8). The following paragraphs

describe illustrative examples of research programs in some of these

areas.
As an example of research related to a specific model process phe-

nomenon, consider the phenomenon indicated in exhibit 3-4 to be of high-

est priority: "maneuver patterns of individual weapons." An associated

research problem would be: Assess performance of the task cluster "vehi-

cle maneuver" in the context of its effect on maneuver patterns of indi-

vidual weapons in maneuver unit combat; determine how such performance

varies as a function of basic abilities of the crew members operating

16
the vehicles. The required research would consist of the following

steps:

15ARI has conducted some research in this area for artillery weapons --

see (Coke, Crumley, and Schwalm, 1980] and [Schwalm, Crumley, Coke, and
Sachs, 1980].
16 Note that this research problem involves development of one of the gj
functions; i.e., it is an example of the first approach to research
described earlier.

52

. . . . . . . . . . . . .... . . . . . - idm~ i.. mmlm m~mliLM5.kwm lw b ' 'm

 '1) Determine appropriate measures of task output. For this
phenomenon such measures might describe aspects of the

weapon-to-weapon line-of-sight processes that result from the

maneuver. See [Farrell, 1977] for a discussion of possible

measures.

(2) Design and conduct field experiments to measure this output as

a function of ability levels. 17 The most important abilities

are believed to be spatial orientation, choice reaction time,

flexibility of closure, selective attention, and problem sensi-

tivity (see exhibit B-5 of appendix B).

(3) Use the results of the experiments to establish a functional

relationship between ability levels and task output.

As an example of a research program structured around task clusters,

ARI has already initiated research in the C31 area (the highest prior-

ity set of tasks) which is illustrative of the type of research needed to

predict the performance of related tasks across a broad spectrum of model

process phenomena [Defense Supply Service, 1982]. The research effort

has three major objectives:

(1) identification and categorization of tactical command control

decision tasks and component processes in terms of state-of-the

art feasibility for representation in combat models;

(2) where currently feasible, development of realistic representa-

tions of tactical decision making logic and processes for

utilization in combat models; and

1 7Alternatively,if the second research approach were being taken, the

experiments could measure the output as a function of demographic
descriptors of crew members.

53

[ . .h
 (3) design of a research program to acquire knowledge needed to
represent in combat models the decision logic and processes not

currently feasible.
The following steps are involved in the design of the research program

associated with this third objective:

(1) Select key kinds of decisions on which to focus the research.

In terms of the research structure described in the previous

section, this step involves identification of specific process

phenomena to provide a context for conducting research on a set

of related tasks with the hope that the results of the research

will. be generalizable to other phenomena (i.e., other kinds of

decisions).
(2) Use expert judgment to identify "ground truth" solutions to

these decision problems. That is, identify "good" decisions

given perfect knowledge of any information bearing on the

decisions. (This will provide a standard for decision quality

for later use in evaluating experimental results.)

(3) Develop a method for providing systematic deviations from

ground truth of the information available for each decision.

(4) Identify a combat simulation for use in generating a decision

making environment.

(5) Design experiments which use this simulation to explore deci-

sion making in the presence of uncertainties and other devia-

tions from ground truth. These experiments will provide the

means to develop the Gj functions descr4bing decision making

*, performance.

V, 54

,.......,.,.,,,............ ... ,......,............ ,.. .......-... ;. .. ... """-

' ' ' l ' ' i i l i l '' i t
71.
_'i.: .:% , "-'.
, .T''," ,' ,''_ i.:. ;''' " "'-T " : " -:--i"'l i . . . " . .
 -2,
Finally, consider an example of basic research concerning prediction

of ability levels: Determine the relationship between demographic

descriptors and the ability "selective attention," using the range of

demographic descriptors of the US Army population. The following steps

would constitute the research:

(1) Develop a quantitative measure for selective attention.

(2) Identify demographic descriptors which, a priori, are believed

potentially useful in predicting an individual's capability for

selective attention.
(3) Design and conduct experiments to measure selective attention

as a function of the demographic descriptors.

(4) Use the results of the experiments to establish a functional

relationship between demographic descriptors and ability

level.

(5) Use this relationship and the demographic distribution of the

Army population to establish the distribution of levels of the

ability in the population.

(6) For later research, identify subjects having a representative

mix of the important demographic descriptors.

55
 REFERENCES

[AMSAA, 1977]

US Army Materiel Systems Analysis Activity, DIVLEV War Game Model

Computer Program, William R. Kennedy, et al., Aberdeen Proving
Ground, Maryland, January 1977.

[AMSAA, undated]

US Army Materiel Systems Analysis Activity, Suppression Estimates

in DIVLEV, 10 pages, undated.

[ARI, 1979]

US Army Research Institute for the Behavioral and Social Sciences,

Human Performance in Continuous Operations: Volume I: Human
Performance Guidelines, Research Product 80-4a, Alexandria,
Virginia, December 1979.

* [Bonder and Proegler, et al., 1978]

Bonder, Seth and Proegler, Lolt, et al., Effectiveness Analysis of

Raytheon Dragon PIP Concepts, Vector Research, Incorporated, VRI-
RAYTHEON-2 FR78-1, Ann Arbor, Michigan, 18 July 1978, CONFIDENTIAL.

[CAA, 1980]
US Army Concepts Analysis Agency, Concepts Evaluation Model V (CEM V)
Part I -Technical Description, CAA-D-80-3, Bethesda, Maryland,
January 1980.

[CACDA, 1980]

US Army Combined Arms Center, CACDA JIFFY III War Game, Volume T:
Executive Summary (Technical Report 5-80), Volume II: Methodology
(Technical Report 6-80), Volume IV: Users Manual (Technical Report
8-80), Volume V: Programmers Manual (Technical Report 9-80), AC,'
24826, Fort Leavenworth, Kansas, September-October 1980.

[CCTC, 1978]

Vector Research, Incorporated, VECTOR-2 System for Simulation of

Theater-Level Combat, CSM UM 244-78, prepared for the Command and
Control Technical Center, Washington, DC, October 1978.

57
 7: 7. 77 77.

[CCTC, 1979a]
Vector Research, Incorporated, VECTOR-2 System for Simulation of
Theater-Level Combat, Technical Memorandum TM 201-79, prepared for
the Command and Control Technical Center, Washington, DC, January
1979.
[CCTC, 1979b]
Vector Research, Incorporated, VECTOR-2 System for Simulation of
Theater-Level Combat, Computer System Manual, Volumes I and II,
CSM MM 244-79, prepared for the Command and Control Technical
Center, Washington, DC, January 1979.

[CCTC, 1979c]
Vector Research, Incorporated, VECTOR-2 System for Simulation of
Theater-Level Combat, Tactical Decision Rules and Tactical Data
Base, Technical Memorandum TM 214-79, prepared for the Command and
Control Technical Center, Washinciton, DC, October 1979.
[Coke, Crumley, and Schwalm, 19801
Coke, Jay S., Crumley, Lloyd M., and Schwalm, Robert C., Emplacing,
Firing, and March Ordering an M1O9AI Howitzer: Tasks and Task Times
(Preliminary), Working Paper FSFU 80-5, US Army Research Institute
for the Behavioral and Social Sciences, November 1980.
[Defense Supply Service, 1982]
Defense Supply Service, Research on Cognitive Functioning in C2 in
Support of Combat Modeling Efforts, Solicitation Number MDA903-82-
R-0044, 1 April 1982.

[Farrell, 1977]

Farrell, R.L., Investigation of the Tactical Control of Cover and

Exposure and Its Relation to Predicted Combat Results, VRI-DUSA-1
FR77-1, Vector Research, Incorporated, Ann Arbor, Michigan, February
1977.

[Farrell and Freedman, 1975]

Farrell, R.L. and Freedman, R., Investigations of the Variation of
- Combat Model Predictions with Terrain Line of Siaht, VRI-AMSAA-l
FR75-1, Vector Research, Incorporated, Ann Arbor, Michigan, January
1975.
[Field Artillery School, 1980]

US Army Field Artillery School, Fire Support Mission Area Analysis

Phase I! (Level II) Report Volume I, Fort Sill, Oklahoma, 15 December
1980 (SECRET-NOFORN).

58

................... .. ... ....

........ . 1.il l fl iilid11 .n J n J ml u mn.m
. I,, _-,.;,
L. - 'o ; . - . - . -- - .. . .~- - -'-- -- -- .~--
." --" . '- - ' ' v
".-'.- r - ' -' - - ." ." ." ' . ' -- - " "- • .-

[Fleishman, 1975]

Fleishman, E.A., "Taxonomic Problems in Human Performance Research,"

in Singleton, W.T., and Spurgeon, P. (eds.), Measurement of Human
Resources, Taylor and Francis, London, 1975.

[GRC, 1974]

General Research Corporation, CARMONETTE. Volume I: General Descrip-

tion, Volume II: Data Preparation and Output Guide, and Volume III:
Technical Documentation, prepared for the US Army Concepts Analysis
Agency, AD-AO07 843, 844, 845, Washington, DC, November 1974.

[Holloway, 1979]
Holloway, Charles A., Decision Making Under Uncertainty: Models and
Choices, Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1979.

[ODCSRDA, 1975]

Office of the Deputy Chief of Staff for Research, Development, and

Acquisition, Report of the Army Scientific Advisory Panel Ad Hoc
Group on fire Suppression, Washington, DC, July 1975.
[Pickett and AuBuchon, 1976]

US Army Combined Arms Center, Sensitivity Analysis of BLDM Acquisi-

tion and Firepower Allocation, TP 11-76, Directorate of Combat
Operations Analysis, US Army Combined Arms Combat Developments
Activity, Fort Leavenworth, Kansas, November 1976.

[Raiffa, 1970]

Raiffa, Howard, Decision Analysis: Introductory Lectures on Choices

Under Uncertainty, Addison-Wesley, Reading, Massachusetts, 1970.

[Schwalm, Crumley, Coke, and Sachs, 1980]

Schwalm, Robert C., Crumley, Lloyd M., Coke, Jay S., and Sachs,
Sidney A., A Description of the ARI Crew Performance Model: A Pre-
liminary Report, Working Paper FSFU 80-6, US Army Research Institute
for the Behavioral and Social Sciences, 1980.

[Theologus, Romashko, and Fleishman, 1970]

Theologus, G.C., Romashko, T., and Fleishman, E.A., Development of
a Taxonomy of Human Performance: A Feasibility Study of Ability
Dimensions for Classifying Hun-in Tasks, American Institutes for
Research, Technical Report 5, Washinqton, DC, January 1970.

59
 [Thompson and White, et al., 1981]

Thompson, D., White, W., et al., Verification of VECTOR-2 with the

1973 Arab-Israeli War and Pilot Analysis of Related Military Balance
Issues, VRI-NFAC-1 FR81-1, Vector Research, Incorporated, Ann Arbor,
i1c'-igan, April 1981 (SECRET NOFORN).
[TRASANA, 1978]

US Army TRADOC Systems Analysis Activity, Command, Control, Communi-

cations, and Combat Effectiveness Model Documentation, Volume I:
Design Report, Volume II: Technical Report, Volume III: Implementa-
tion Report, White Sands Missile Range, New Mexico, Technical Memoran-
dum 3-78, October 1978.
[TRASANA, undated]

US Army TRADOC Systems Analysis Activity, CASTFOREM III, undated.

[US Army War College, 1981]

US Army War College, McClintic Theater Model, Volume I: War Game

Director's Manual, Volume II: User's Manual, and Volume III:
Controller's Manual, Department of War Gaming, Carlisle Barracks,
Pennsylvania, July 1981.

[VRI, 1973]

Vector Research, Incorporated, VECTOR-O: The Battle Model Proto-

type, Volume IV: Test Results, VRI-WSEG-2 FR73-1 (WSEG report
number 222). Ann Arbor, Michigan, December 1973.

[VRI, 1974]
Vector Research, Incorporated, Application of the ARAFCAS Model to
a European Scenario, VRI-BDM-I-ARAFCAS-I, Ann Arbor, Michigan, 24
June 1974 (SECRET NOFORN).

[VRI, 1975]

Vector Research, Incorporated, The Battalion Level Differential

Model (BLDM): Planner-User's Manual, VRI-BDM-CLGP-l TR75-8,
Ann Arbor, Michigan, December 1975.
[VRI, 1977]
Vector Research, Incorporated, Analysis of Tactical Nuclear Force
Capability, VRI-DNA-l FR77-1, Ann Arbor, Michigan, 30 March 1977
(SECRET-RESTRICTED DATA).

60

.. . .. .. . . .. ... . . .
 [VRI, 1981a]

Vector Research, Incorporated, An Exploratory Study of Maneuver

Unit Reinforcement Versus Second Echelon Interdiction Packages,
VRI-TRADOC-9 FR81-1, Ann Arbor, Michigan, 31 October 1981.

[VRI, 1981b]
3.. Vector Research, Incorporated, Hughes Helicopters Near Term Scout
Helicopter with Mast Mounted Sight, VRI-HUGHES-1 FR81-1, Ann Arbor,
Michigan, 5 March 1981.

61-

,°61

.................. .-
 APPENDIX A
COMBAT PROCESSES IN SPECIFIC SIMULATIONS

This appendix consists of detailed notes describing combat model

processes which are influenced by human factors. As described in section

2.1 of the body of this report, a representative set of nine force-on-

force combat models was studied in order to determine

(1) major processes represented in the models,

(2) phenomena associated with each of these processes which are

influenced by human factors, and

(3) the manner in which each phenomenon is represented in each of

the models.
Exhibit A-i identifies the nine models and indicates references to model

documentation used in the analysis of each of them. The pages which

follow the exhibit contain 15 sets of notes regarding each of the

following 15 model processes identified in the project:

(1) maneuver unit combat;

(2) artillery fire;

(3) air-to-ground attacks;

(4) air-to-air combat;

(5) air defense;

(6) mobility, countermobility, and survivability;

(7) intelligence and electronic warfare;

(8) communications;

(9) mission assignment, organization for combat, and resource

allocation;
(10) maneuver control;

1A-1

...................................................................... . ..
 EXHIBIT A-i: MODELS CONSIDERED :
MODELS REFERENCES

CARMONETTE [GRC, 1974]

BLOM [VRI, 1975]

P CASTFOREM

DIVLEV

~FOURCE
[TRASANA, undated]

[AMSAA, 1977] [AMSAA, undated]

[TRASANA, 1978] 9b,17c

GEM [CAA, 1980]

McCLINTIC [US Army War College, 1981]

A-2
 (11) fire support allocation;

(12) resupply and replacement;

'- (13) movement;

(14) maintenance and repair, medical support; and

(15) construction, maintenance, and repair of airfields.

Each set of notes includes the following information:

. a page defining the model process and listing the phenomena

within the process which are influenced by human factors;

* a table containing symbols which indicate the method by which

each phenomenon is represented within each of the nine models;

and
* a set-of definitions of the symbols which appear in the

table.1

Among the symbols used is an 'T' which appears in the corners of many of

the model/phenomena cells of the tables. This indicates that the model

contains an explicit representation of at least some aspects of the

phenomenon in question, or is based on input which can-be generated via

an explicit representation of the phenomenon. Where no "X" appears, the

phenomenon is either not represented, is represented only implicitly, or

is assumed always to occur perfectly or be correctly done.

.. 9

IFollowing the notes for the first model process (maneuver unit combat)
is an additional set of definitions of general symbols which appear
repeatedly in the tables for all of the model processes.

,* A-3

..- .C- - ..' . • . . . ° .- ..

* . . . . ... • . .
- -. . , ., *.4_*-.. .
C' *C,"- ,-C
*. ,",-. , -.- ., , •. , .-- . . . .. . ..
 .o..

MODEL PROCESS
- MANEUVER UNIT COMBAT -

DEFINITION:

REPRESENTATION OF FIRE AND MANEUVER OF FRONT-LINE

FORCES, RESULTING IN ATTRITION, SUPPLY CONSUMPTION,

AND CHANGES IN WEAPON LOCATIONS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* FORCE PARTICIPATION

" INITIAL DEPLOYMENTS OF WEAPONS

" MANEUVER PATTERNS OF INDIVIDUAL WEAPONS

* TARGET ACQUISITION CAPABILITIES

• TARGET IDENTIFICATION CAPABILITIES

0 TARGET SELECTION (AND TARGET SWITCHING)

0 DECISION TO FIRE

0 ORDNANCE SELECTION

- TIME TO FIRE OR FIRING RATE

0 ACCURACY OF FIRE

- FIRE SUPPRESSION (DIRECT, INDIRECT)

- EMPLOYMENT OF OBSCURATION

A-4
Ido

.410

IN

11
cc;

CD z

z w(

I-.~~' Ado
w-C
~
if
~ V
Il
7T7
* I~cc
-y -C cc~

* A-5

LJ7J7_
 Maneuver Unit Combat

I -A. Nonparticipants (by weapon type) are a function of the amount of

exposed flank and the type of engagement and type of unit involved
(the latter two considerations are controlled by an input
fraction)

1-8. Nonparticipants are a function of the amount of exposed flank. Any

other nonparticipation is represented implicitly in input firepower
scores.

1-C. All weapons participate, although degree of participation of any

weapon ca- be controlled by pre-specified "orders."

1-0. Based on styl ized unit deployments which are built into the model.

I-E. Governed by user-specified decision tables.

1-F. Weapons in a position in which they are out of range of opposing

forces (see 2-C) do not participate.

2-A. Input for each weapon type as a function of the type of engagement
and type of unit involved.

2-B. Based on stylized unit deployments which are built into the model.

2-C. Weapons in a given unit are deployed uniformly over a rectangular

-. area whose dimensions are input (and which may differ for opposing
forces in a given engagement).

3-A. Standard patterns for each weapon type are governed by inputs as a
function of the type of engagement and type of unit involved.

3-B. Governed by pre-specified "orders."

3-C. Governed by user-specified decision tables.

4-A. Use of serial or parallel acquisition is input and depends on

weapon type. Acquisition may be either visual or pinpoint. No
false or dead targets are acquired. Acquisition capabilities are
input as functions of firer, target, range, visibility conditions,
and movement status of firer and target. Given line of sight,
acquisition time is assumed to be exponentially distributed.

4-B. Acquisition represented by a Markov chain model with four states:

(1) no information about target; (2) target location known to
nearest grid square; (3) target erroneously pinpointed; (4)
target accurately pinpointed. State transition probabilities (some
input and some computed in the model) depend on existnnce or
absence of line of sight, suppression state of observer, whether
the target is firing, movement status of target and observer,
target solid angle subtended .at the observer, and sensor type. No
false targets are acquired, but dead targets which are not "known
dead" by the observer may be acquired.

A-6
 4-C. Implicitly included in input kill rates, which vary with firer,
target, range, and target exposure (hull defilade on fully
exposed).

6-A. Highest priority target acquired is selected; priorities are input

for each firer. (In BLDM, priority is increased if firer and
target are on the same "axis.") Parallel acauirers switch to
higher priority targets as they are acquired. Otherwise, target is
engaged until killed or line of sight is lost.

6-B. Targets are selected from alternative priority lists based on

either value or danger of the target. Selection of the list to be
used by any firer is prespecified by the user in input "orders."
Orders also specify whether suppressive fire into an area is to be
employed. Target switching is allowed after each firing. Input
probabilities specify the chance of observing that a dead target
has been killed; if a target is not known to be dead, additional
rounds can be fired at it.

6-C. Fire at any point in time is assumed to be allocated to targets

using the following proportions:

Fi j : Ui jj/ Ui jj,

where Fi- is the fraction of weapon type i allocated to target

type j,lii is the kill rate for a weapon of type i firing at a
target of Iype j, and Nj is the number of j type targets
present.

6-0. Target selection is based on a priority scheme; priority is a

function of range, estimated threat, exposure, detection status,
time since last fired or moved, number of other weapons firing at
the target, fire sector, and turret orientation of the target.
Target switching can occur after the desired number of bursts have
been fired or the desired (perceived) kill level has been
achieved.

7-A. Firing always. occurs if ordnance is available and if an acquired

target is in range; ranges are input.
7-B. Governed by pre-specified "orders."

7-C. Governed by user-specified decision tables.

8-A. Preferences for ordnance type to be used against a target of

specified vulnerability are input. Up to two ordnance types are
available for a given weapon; preferred type is used if available.

8-B. Governed by user-specified decision tables.

8-C. Uses rules built into model code, based in part on range to the
target.

A-7

... '.....-..-.'.-.'.-....'.".-...-".-
..-...-. ".. .... .................
 9-A. Implicitly included in kill rates, which are input as a function of
firer, target, range, target exposure (hull defilade or fully
exposed), and (for VECTOR-2 and FOURCE) movement status of firer
and target.

9-B. Computed in the model from inputs specifying means and standard
deviations of times to aim, reaim, and reload the weapon, and
average round velocities.;

10-A. Implicitly included in kill rates, which are input as a function of

firer, target, range, target exposure (hull defilade or fully
exposed), and (for VECTOR-2 and FOURCE) movement status of firer
and target.

10-B. Computed in the model from average miss distances for each weapon
and ordnance type (input as a function of suppression state of the
firer, whether the firer and target are moving, range to the
target, and whether this is a first or subsequent round) and from
target area data.

11-A. If fractional attrition rate of any weapon type exceeds input

threshold, all weapons are assumed to increase their use of cover
to lower the rate to the threshold. Weapons suppressed in this way
cannot move, fire, or be attrited.

11-B. "Neutralization weights" (relative suppressing cpability) of all

round types are input; total weight landing in a given area during
a user-specified "neutralization interval" is compared with input
thresholds to determine if weapons in the area are either partially
suppressed or (for dismounted infantry and open vehicles only)
pinned down. Partially suppressed ground weapons conduct
surveillance and fire at reduced accuracy, require twice the usual
time to aim, and move at reduced rates. Partially suppressed
helicopters finish guiding any missiles in flight, then drop to
treetop level.
11-C. Fraction of defending force in suppressed state due to direct fire
equals

1 - exp (- CP0 X/M),

where M is the size of the defending force, X is the size of the

portion of the attacking force in overwatch, Po is the rate at
which an overwatch weapon kills a defending weapon, and C is a
constant whose value is determined via detailed consideration of a
specific attack scenario. The fraction of a target suppressed by
artillery is given by an expression of the form

1 - exp (- KF),
where F is the fraction of the target area covered by artillery
fire and K is a constant set so that, when F equals 1, the fraction
suppressed is .5 for moving targets and .75 for static targets.

A-8

.. . . . . ..
. . . . . ..... . . . . . • . . . . . . ..
 11-D. Suppression is caused either by the rate of received fire in the
near vicinity of a target or by the rate of non-lethal hits,
depending on the version of the model used. Suppressed weapons are
unable to fire; in some versions of the model they are also
unavailable as targets and previous acquisitions of and by them are
lost. Suppression has an input duration dependent on firer and
target type.

11-E. Will be represented in CASTFOREM, but existing specifications do

not describe how.

11-F. Firepower force ratios are used to determine (via input tables) the
fraction of weapons suppressed. Suppressed weapons are not
available to fire.

11-G. If an engaged unit is under fire from artillery, the number of

weapons that can fire is reduced by an input fraction. The
fraction is a function of whether the unit is moving or
stationary. The speed of a moving unit is reduced when under fire
from artillery or air.

12-A. Either input directly by games or governed by gamer-supplied

contingency orders, which can be activated as a function of time,
unit location, casualty level, or movement status of a specified
unit.

12-B. Will be represented in CASTFOREM, but existing specifications do

not describe how.

-o-

N°

.4
4'

A-9
.7 n.

F * -. .. . . . . A.-.-- . .
 7• 7 ,

General Symbols (used in conjunction with all model processes)

Ext. Represented implicitly by model input which can be generated by an
explicit external representation of the phenomenon.
I
Gam. Represented via gaming (i.e., phenomenon is represented through an
external player, who inputs its effects into the model
interac ti vel y).

In. Directly input by the user.

Nom. Represented implicitly at some nominal, constant level based on

model input.

Perf. Assumed to be always perfect or correctly done.

NR. Phenomenon is not represented.

-. The process with which this phenomenon is associated is not

represented."-.

X. Represented explicitly (see further description of this symbol in

text).

..........
.. .. . .. .. . . . . . . . _.

.. .. . . . . . . . . . . . .

A--0
 ' . L. , -' J % . -. .- _. :.---,--:~-
." . -. .. . . . . . . . . .- .. . - . ° . -.. . . .

MODEL PROCESS
- ARTILLERY FIRE -

DEFINITION:

REPRESENTATION OF ARTILLERY FIRE AGAINST A TARGET,

CAUSING ATTRITION AND DEGRADED FUNCTIONING OF THE

TARGET, AND RESULTING IN ORDNANCE CONSUMPTION AND

INCREASED DETECTABILITY OF THE FIRING WEAPONS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* ORDNANCE SELECTION

* TARGET ELEMENT ACQUISITION AND SELECTION (IF POINT

FIRE)

* DECISION TO FIRE

* DELIVERY PATTERN

* DELIVERY ACCURACY

• TARGET POSTURE

* SUPPRESSIVE EFFECTS OF FIRE

. EFFECTS OF FIRE ON COMMUNICATIONS AND

DECISION-MAKING ABILITY

0 TIME TO FIRE OR FIRING RATE

A-i1
1-4-

[ . . . . . . . ".

-' 4'- " '- " " ". . . .. . . . .. " " ,. - ,,-. . . - . - - -
 .1

Luz M

Q-~ 02
-

- =
CM=

c c

cd~~A- - 2
 Artillery Fire

1-A. Preference of ordnance type to be used against targets of a

specified vulnerability are input. Up to two ordnance types are
availanle for a given weapon; preferred type is used if available.

1-8. Uses rules built into model.

* 2-A. Highest priority target el e.ment acquired is engaged. Acquisition

probabilities and priorities are input.

2-8. Forward observers acquire and select targets using a representation

essentially the same as for target acquisition and selection by
maneuver unit weapons.

3-A. Firing always occurs if requested.

4-A. Included in area of effects, for which values are built into the
model.

6-A. Input as a function of target activity.

6-8. Model determines one of two postures for a target: responding to

hostile fire and not responding to hostile fire.

6-C. If target is stationary, all elements are in defilade. Otherwise

fraction of exposed elements is a function of current loss rate.

6-0. Based on maneuver patterns of individual target weapons, which are

input for all groups of weapons on the battlefield.

7-A. Counterbattery or air defense suppression fire causes an input

fraction of target weapons to be suppressed per volley or round
fired against them; suppression has an input duration.
7-B. Counterbattery fire can partially neutralize opposing DS artillery. L
Fraction neutralized for current engagement is

S- eK(C/D)

where K = input coefficient,

C = amount of antipersonnel counterbattery firepower in a
sub sector,
0 = number of DS battalions receiving counterbattery fire.

7-C. Speed of a moving target is reduced by a constant factor when under

attack by artillery fire. If artillery is attacked, it cannot fire
for 15 minutes (to represent relocation of battery).

7-0. The first attack of an artillery battery on an opposing artillery

battery causes total suppression (no firing) for the length of the
mission plus 15 minutes. Subsequent attacks within six hours cause
total suppression for the length of the mission plus 25 minutes.
These times are assumed necessary for battery relocation.

A-13
.9'
 7-E. Firepower score force ratios are used to determine (via input
tables) the fraction of weapons suppressed. Suppressed weapons
cannot fire.

8-A. Decision time at command posts is increased as a function of

fractional damage to command post; amounts of increase are input.

9-A. A three minute delay between volleys is assumed; it takes six hours
to reload missile and rocket launchers.

9-B. Built-in firing rates are degraded for personnel casualties by

multiplying by a term
-
+Ie-13 .73 (.5 - X)]

where X is the fractional amount of personnel lost.

... - "'
. . . .. . . . . . . . . . . .

- . - . . . .. . . . .. .

SA-14
r ., . ,- _. '. , ~ _J ' . " . " ' ." - . ".. " r " " ' -. .' . -. " _"- - " . ".-- - . -.- .

MODEL PROCESS
- AIR-TO-GROUND ATTACKS -

DEFINITION:

REPRESENTATION OF THE EFFECTS OF AIRCRAFT ATTACKING A

TARGET ON THE GROUND, CAUSING ATTRITION AND DEGRADED

FUNCTIONING OF THE TARGET, AND RESULTING IN ORDNANCE

AND POL CONSUMPTION AND INCREASED EXPOSURE OF THE

ATTACKING AIRCRAFT

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* ABILITY TO ACQUIRE (OR REACQUIRE) THE TARGET

* ENGAGEMENT DECISION (WHETHER TO ATTACK THE TARGET)

* ORDNANCE SELECTION

* TARGET ELEMENT ACQUISITION AND SELECTION (IF POINT

FIRE)

* AIRCRAFT STANDOFF DISTANCE AND/OR MANEUVER PATTERN

OVER TARGET

* WEAPON DELIVERY ACCURACY

TARGET POSTURE

* SUPPRESSIVE EFFECTS OF THE ATTACK

* EFFECTS OF THE ATTACK ON COMMUNICATIONS AND

DECISION-MAKING ABILITY

* EFFECTS OF THE ATTACK ON AIRFIELD SORTIE GENERATION

CAPABILITY (IF TARGET IS AN AIRFIELD)

* AIRCRAFT DISENGAGEMENT DECISION

A-i5
 -777=77 77-

- 00"; v C =

z z

0, , 1
"4/ 9- 1 1~4
,
1
ic
4/ 4/

o) 'o
0
O'
"I.~~~ 4,Co~ 1

-
LW 41o-

-Ofj.o

o,.,<
4"o ~
041:,

-I 3;. ~ 4/

I~~7 I A
 Air-to-Ground Attacks

I-A. Input probability as a function of type of target, class of sensor

making the original acquisition, weather, and time elapsed since
the oriqinal acquisition.

3-A. Ordnance load selection for a mission is governed by user-specified

decision rules, as part of the mission assignment process.
Priorities for load selection are input for use in these rules.
Ordnance type and quantity selection for a given load is input for
each pass over the target.

4-A. Highest priority target element acquired is engaged. Acquisition

probabilities and priorities are input.

7-A. Input as a function of target activity.

8-A. Counterbattery or air defense suppression fire causes an input

fraction of target weapons to be suppressed per round fired aaainst
them; suppression has an input duration.

8-B. Suppressive effects of interdiction strikes represented through

increases in time for resupply and higher-level reserve commitment
if friendly air superiority does not exist.

8-C. Aircraft attacking a moving target cause a fixed time delay before
movement can resume.

8-0. An engaged unit's speed is reduced to an input fraction of its

otherwise attainable speed.

9-A. Decision time at command posts is increased as a function of

fractional damage to command post; amounts of increase are input.

11-A. Governed by user-specified tactical decision rules and by input

limits on the total number of passes to be flown against a given
target. Existing rules in most versions allow the input number of
passes always to be flown.

11-B. Aircraft are assumed to spend a fixed time over the target, based
on an input limit on time of flight.
A.17

A-17 "

~~...... . ........ ... •. . °. ... . ....... o . . ... . . . . °

 MODEL PROCESS
- AIR-TO-AIR COMBAT - j
DEFINITION:

REPRESENTATION OF ENGAGEMENTS BETWEEN OPPOSING GROUPS

OF AIRCRAFT, RESULTING IN ATTRITION TO AIRCRAFT,

. CONSUMPTION OF ORDNANCE AND POL, AND ABORTS OF OR

DELAYS IN ACCOMPLISHMENT OF AIR MISSIONS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

- ENGAGEMENT DECISIONS (WHETHER TO ENGAGE AND TYPE OF

ENGAGEMENT)

' AIRCRAFT MANEUVER PATTERNS

_ TARGET ACQUISITION CAPABILITIES

. TARGET SELECTION

" ORDNANCE SELECTION

- TIME TO FIRE

*ACCURACY OF FIRE
. ABORT DECISIONS

,e

A-18

"""' " .' ..............................

"'.. '........, " "--.
,.*,,,.,..,,-..w,,..
" " " '""...,,,.--'
"*"" ....
. .,. .... .,... . . ,-.'
 C)C

CDC

c~r-

C) II-

I i

IA.1
* Air-to-Air Combat

2-A. Details of maneuver patterns represented only implicitly by effects

data; determination of whether standoff or duel-type combat occurs
for a given part of the battle is determined dynamically during the
engagement by types of ordnance available (i.e., whether standoff
ordnance is carried) and by either user-specified decision rules or
by an input specifying a maximum number of standoff exchanges to be
followed by a single dual phase. Existing rules default to the
input maximum number of standoff exchanges.

- 3-A. A single input probability represents entire intercept process

against penetrating aircraft: detection by ground control system,
availability of interceptor for the engagement, and probability of
successful intercept.

4-A. Specified by either user-input decision rules or by a model

algorithm which allocates aircraft randomly to targets. Existing
rules default to the random allocation algorithm.

, 4-8. Allocation is random and assumes one-on-one engagements only.

7-A. Implicitly included in air-to-air exchange rates, which are input

as a function of firer type, target type, and whether aircraft are
using standoff or duel type ordnance.

8-A. Governed by user-provided tactical decision rules, or by a

model-specified rule which bases abort decisions on a
penetrator-to-interceptor strength ratio (after each standoff
exchange) or on surviving friendly strength (after the duel
phase).

* 8-B. All intercepted penetrators abort their air-to-ground missions.

Aborts of air-to-air engagements are represented implicitly only
through input mission kill probabilities.

8-C. Unescorted penetrators jettison ordnance and return home if

attacked. Other types of aborts not explicitly represented.

S.
A.. . .
 MODEL PROCESS
- AIR DEFENSE -

DEFINITION:

REPRESENTATION OF FIRE BY AIR DEFENSE ARTILLERY,

RESULTING IN ATTRITION OR DEGRADED FUNCTIONING OF

TARGET AIRCRAFT, AND ORDNANCE CONSUMPTION AND

INCREASED DETECTABILITY OF THE FIRING WEAPONS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* COORDINATION OF FIRE

* TARGET ACQUISITION CAPABILITIES

* TARGET SELECTION

* DECISION TO FIRE

* DEGRADED CAPABILITY OF AIRCRAFT TO DESTROY TARGETS

ON THE GROUND

AIRCRAFT MISSION ABORT DECISIONS

.- . .

* . . *..** ' . - .

A-21
 MODEL PROCESS: AIR DEFENSE1

12 34 5 6

.70

6Z6

MOEL~

CARMONETTE --- --

DIVLEV --- --

FOURCE ------

JIFFy 2 Nomi Nom Nom, Nom, Nom, Nom

VECTOR-? A In A Nomn NR A

CEM 11o11 in Nom Nom, NR NR

McCLINTIC Nown Nom, 14o1 Nomi H4R NR

CASTFUREM ---- -

INot Includinig air defense fire against attack helicopters durinq marieuver
inilt combat. See "Maneuver Unit Combat" for details.
2
jIFFY represents only local air defense of maneuver units under attack.
althouqh It is usually used in conjunction with a separate air effects model.

A-22

%4

~hq~ .......... 4%.

 Air Defense
1-A. Assumed that no more than one air defense site on the average will
fire for every two aircraft in an overflying group.

3-A. Air defense fire is distributed randomly among target aircraft.

6-A. Not represented on overflight; governed by user-input tactical

decision rules in target area. Current rules do not abort
missions.

;.A.

,,..

.°

a. A-23

" ~ ~ ~ * * .... . . . I - I . . . . .
 MODEL PROCESS
- MOBILITY, COUNTERMOBILITY, AND SURVIVABILITY -

DEFINITION:

REPRESENTATION OF ENGINEERING OPERATIONS AND THEIR

EFFECTS ON OTHER PROCESSES, INCLUDING ATTRITION,

DELAYS, AND INCREASED EXPOSURE DUE TO MINEFIELDS AND

OBSTACLESj THE VULNERABILITY OF FORCES DEFENDING AT

PREPARED POSITIONS; AND THE ABILITY TO MOVE ON ROADS

AND ACROSS RIVERS AND GAPS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

• LOCATION, DENSITY, AND EXTENT OF A MINEFIELD OR

OBSTACLE

O ATTRITION SUFFERED WHEN ENCOUNTERING A MINEFIELD

, DELAY IN MOVEMENT WHEN ENCOUNTERING A MINEFIELD

OR OBSTACLE

, REDUCTION IN MOVEMENT RATES DUE TO MINEFIELDS AND

OBSTACLES
" INCREASE IN EXPOSURE DUE TO ENCOUNTERING A

MINEFIELD OR OBSTACLE

* DELAYS IN TRAVELING ON DAMAGED ROADS

DELAYS IN CROSSING RIVERS OR GAPS

- DEGREE OF PROTECTION AFFORDED BY A PREPARED

DEFENSIVE POSITION

* LOCATION AND EXTENT OF A PREPARED DEFENSIVE

POSITION

A-24

~.
..--,
 I-S
00.20'
p

'0 . -, . G = . =

o do do

cc
-04=

=
Lad
>.
6 ~0- . = I

o __________________________________

Ja-
.
>! *d'a.-I

a0 fu r ;
Pa L~ , C

cc2

CD ..
 Mobi. ity, Countermobility, and Survivability

1-A. Could be represented via inputs describing characteristics of

terrain and ability to move an it.

1-B. Locations are governed by ganers. Density and extent are not
explicitly represented.

I-C. Locations of minefields are controlled by user-specified decision

tables governing minefield emplacement. Density and extent are
assumed constant based on user inputs.

1-D. Decisions concerning minefield location and density are made by

garners. Extent of a minefield is computed internally as a function
of resources available to emplace the minefield, rate of
emplacement using these resources, time available to emplace the
minefield, and a "work degradation factor" which degrades
emplacement rates as a function of level of the hostilities.

2-A. Both discovery losses and crossing losses are input; the latter
depend on whether the minefield is in an urban area and the method
used to cross or clear the minefield.

2-B. Percentage attrition is built into the model.

2-C. If a minefield has a breachpath, all vehicles are assumed to know

its location; otherwise input probabilities determine whether a
minefield is detected when encountered. If undetected, minefield
will be crossed and input probabilities determine whether vehicle
survives the crossing.

2-D. Attrition is assessed only if the crossing force is using a hasty

breach technique. Attrition is then computed as a function of
fraction of trafficable terrain frontage covered by the portion of
the minefield not bypassed, the fraction of the force used to clear
the minefield, and input casualty data.
3-A. Assumed to require one division period (12 hours).

3-B. Delay is a random time between two and four hours.

3-C. If a minefield is detected (see 2-C) user-specified decision tables

specify the response; the response results in an input delay.

4-A. Input factor is a function of whether the minefield is in an urban

area, whether it is covered by fire, and the method used to cross
;* or clear it.

4-B. Could be represented via inputs describing characteristics of

terrain and ability to move on it.

4-C. When crossing a breachpath, vehicles travel at reduced speeds.

A2

A-26.,

.1F
4-0. Rates of advance are reduced to 75 percent of their normal values
when minefields or barriers are being crossed.
5-A. Combat strength of a unit in an overwatched minefield is halved.

5-8. "Percent weapon enhancement" for ambush areas is input for each
barrier.

5-C. Exposure corresponds to the terrain on which the minefield is

located.

6-A. Additional (input) delay is imposed in delivery of supplies and

replacements and in commitment of army and corps reserves if
friendly air environment does not exist. Friendly air environment
is present if enemy air sortie losses exceed an input threshold.

6-B. Delays governed by input; repair is specified by user-provided

decision tables.

7-A. Directly input for front-line forces; represented implicitly for

forces in the rear. .7

7-B. Could be represented via inputs describing characteristics of

terrain and ability to move on it.

7-C. Delays governed by input; bridging is specified by user-provided

decision tables.

8-A. Defenders at defensible positions (located via input) are assumed

to be in hull defilade (or to have some other degree of added
protection, as specified by kill rate data).

8-B. Represented implicity via firepower scores, which vary depending on

whether position is prepared or hasty.

8-C. Could be represented via inputs describing characteristics of

terrain.

8-0. An input fraction of stationary vehicles in hull defilade (and

therefore with decreased vulnerability to direct fire weapons)
depends on whether the force is in a "fortified position," a
"prepared position," or "conducting a hasty defense."

9-A. Whether defending force has had time for current position to be
prepared is a function of recent FEBA movement -- position is
prepared if average movement is below an input threshold.

9-8. Could be represented via inputs describing characteristics of

terrain.

9-C. Construction of prepared positions is governed by user-specified

decision tables. They have a standard size based on user input.

A-27

4 .
 9-0. Blue battalions are assumed to be in "fortified positions"
initially, and in "prepared positions" at each successive
destination.

I-'

-. 4

A -,

-e

.1

.4

.4;

-A

*1
.4
.4

.4.'
.4.,
.4.'
*1
4

'.4,
.4,
.4'

'4

A A-28

* . -... 4..4..-. ~ *~ *** .- . .. . -

................ . . . . .4 . .4
. .
................. ................................................
.4.4
 MODEL PROCESS
- INTELLIGENCE AND ELECTRONIC WARFARE -

DEFINITION:

REPFSENTATION OF THE ACQUISITION AND DEVELOPMENT OF

TARGETS FOR FIRE SUPPORT; THE ACQUISITION AND

DEVELOPMENT OF INTELLIGENCE CONCERNING ENEMY ORDER OF

BATTLE, CAPABILITIES, ACTIVITIES, ETC.; AND THE USE OF

COUNTERMEASURES AND COUNTER-COUNTERMEASURES TO CONTROL

THE USE OF THE ELECTROMAGNETIC SPECTRUM

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* COLLECTION MANAGEMENT DECISIONS - WHAT INFORMATION

TO PURSUE

* COLLECTION MANAGEMENT DECISIONS - HOW TO COLLECT

DESIRED INFORMATION
* TARGET ACQUISITION CAPABILITIES

* ACCURACY OF TARGET IDENTIFICATION (INCLUDING FALSE

TARGETS)

* TARGET PROCESSING TIME

• TARGET LOCATION ERROR

" TIMELINESS OF INTELLIGENCE PROCESSING

* ACCURACY OF INTELLIGENCE PROCESSING

" COMPLETENESS OF INTELLIGENCE PROCESSING

• DECISIONS TO JAM

* EFFECTIVENESS OF JAMMING

• COLLECTION MANAGEMENT DECISIONS - LOCATION OF

SENSORS

A-29

............. "...'..-.....,-..........-." - . --- - . ,. .

.4

WINN

.1
.
1 I t
jo
Ia

I- -

U.1J

"o -,

C -

-j0 II I

4', Q.' =
C., -2,06.,

dt2*4 4

(101
* ~A f
: 0

V) d~ot, 4f

-.. M6, ,

Cmh9P'
L...Z .

.'o~&00 02
&0'aC
It2

4a

-~ ~ 1' _____A__30
 Intelligence and Electronic Warfare

1-A. HUMINT missions on enemy activities in a specified area are called

for by garners. Other types of collection occur automatically
within the model.

1-B. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

2-A. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

3-A. Input as a function of sensor type, target type, target activity,

range to the target, target strength, and weather visibility.

3-B. Input for acquisition of units by overflying aircraft. Built into

the model for national sources. Most other capabilities
represented implicitly.

3-C. Represented as overall acquisition probabilities or "percent of

knowledge" values input as a function of target type.

4-A. Fixed probability of identification of targets by national sources

is built into the model. Other sources identify targets
accurately.

4-B. Target is identified as maneuver unit unless it is known to be

artillery.

5-A. Staff element processing times are modeled explicitly using a

priority queueing structure for handling incoming messages.
Minimum processing times and maximum acceptable throughput times
are input for each message type. Messages are lost if the system
is overloaded.

6-A. Locations of sightings are reported; actual target locations can

differ because of target movement.

7-A. Previously acquired targets are dropped from lists as they be-
come stale. Time by which a target of a given type is stale is
input. Other targets not under direct observation may remain on
the list even if they are no longer present.

7-B. Intelligence concerning enemy strengths is weighted average of two

recent actual strength levels -- timeliness is thus determined by
values of weights.

7-C. HUMINT and national intelligence information is delayed a fixed

time before being made available to garners.

7-0. Governed by controller inputs to gaiters.. 7

A-31

'"+"' " ". -°"+°"

" . . "" .+ . " . " . °" . "
. ~.
.
." ' ".
. '. .
."
. . . . .
.-.. "-
""'
.
+".
..-
"+ .
.
" .-.
" "°
. .
. ..."..
" ""w
.
" "
.
....""
.
".
.
-."•. "
.
.
.
. .
. . . . .
" ..
-
,+.
.V' .e, . ,-t+
',P. . . . .
,,.,.. . . . . . . . . . . . . . . . . . . . + . .- . -. -. .
-. . - ..-. -. . - ...
- . ..- ' .. +
. m++m.+
it+++. .++ii+
. . +++
s.i. . .
+. .
++++... . . . . . . . . . . . . . . . . ,. . . ++ . "-. '". . .
.'","o ". , . + .+" . .
°" .."" .. " " .
", . , ..-" " '
4 p 6. . . . . . - . . . . . . ..
0V

7-E. Staff element processing times are modeled explicitly using a

priority queueing structure for handling incoming messages.
Minimum processing times and maximum acceptable throughput times
are input for each message type. Messages are lost if the system
is overloaded.
j
8-A. OB intelligence consists of estimates (derived from sensor reports)
of opposing strengths, updated periodically via a Bayesian
processing scheme. Thus, inaccuracies of initial estimates and of
sensor reports result in inaccurate intelligence. Weather,
terrain, and friendly situation are assumed to be known accurately.
Weather forecasts are input (and can be incorrect).

8-B. Except for time lag (see 7-B) intelligence is assumed to be

accurate.

8-C. Fixed probability of identification of a unit's activity by

national sources is built into the model. Other intelligence
information (but see 4-A above) is reported accurately as of the
time of collection.

8-0. Governed by controller inputs to gamers.

8-E. Processing is based on built-in rules which determine how to merge

existing information with new sensor-provided data. Because both
new data and existing data can be incorrect (e.g., stale or
.4. incomplete), erroneous inferences can be drawn.

9-A. OB intelligence consists of estimates (derived from sensor reports)

of opposing strengths, updated periodically via a Bayesian
processing scheme. Thus, incompleteness in initial estimates and
in sensor reports result in incomplete intelligence. Weather,
terrain, and friendly situation are assumed to be known. Weather
forecasts are input.

9-B. 08 intelligence is complete in the sense that no relevant

information is assumed to be missed by the process, except as
affected by time lags (see 7-B).

9-C. Intelligence information includes unit type, location, activity,

and strength. A fixed percentage of enemy orders are also
intercepted.

9-D. Governed by controller inputs to gamers.

9-E. If processing system is overloaded, some messages will be lost.

System produces estimates of opposing force strengths and location
of Red main thrust.

10-A. A fixed percentage of orders are jammed (not based on any explicit
decision to jam). Other types of jamming are represented only
implicitly.

A-32

i
~~~~~~.
. . . . ........
. ,_... ..... ............ . -..... ,
: !: i'======= ===========X " "': : :: {:' : : :li i.n .*' a.. '~i:i : " ..I--
. l ) i . . . -- I. ::::. M ::! :i : i:: :
i
 -
4' ' . . *, - , -. .. ,-. .- ,-- i . *.* .I. . - - - -o- - . , .- ._ _ _ . . . . .

10-8. Will be represented in CASTFOREM, but existing specifications do

not describe how.

11-A. Jammed orders are not received, and sender is not informed of the -
jamming.

11-B. Probability of reception of a message is computed based on carrier

propagation theory and whether the receiver is jammed.

11-C. Degradation factors (which are drawn at random from sets of input
factors) reduce firepower .scores for use in force movement
computations and decrease the number of artillery battery missions
avaii able.

12-A. General position on battlefield is directly input. Detailed siting

decisions represented implicitly via input line-of-sight
probabilities, which depend on sensor type, terrain type, and range
to the target.

12-B. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

.............................
..............

A-33
 MODEL PROCESS
,.COMMUNICATIONS -

-. DEFINITION:

REPRESENTATION OF THE TRANSMISSION OF MESSAGES ON THE

q" BATTLEFIELD, WITH INHERENT ERRORS AND DELAYS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

- TIME TO TRANSMIT A MESSAGE

* MESSAGE CONTENT
.9
* TRANSMISSION ACCURACY

' RECEPTION ACCURACY

-I:-

'a'

A-3

!.
I

5,

A-34

Se. . . . . . . . ...-. . .

,.......... -... . . . . . . . . . . ..... . . . . . . ...

.S. . . . .. .. .-.. .....° ' '" ,...... .. ...... ..,- ,..-. - .
' '%
°''"
°%°."",'°- '- '. '- °- °. " " -" ° % "" '" " " " " ° ." ." ." .-" ."".-".
"" '- '-.".
-' .-.. " -"-S.""-
.,.- "" ..-.. • "."S
 MODEL PROCESS: COMMUNICATIONS

1 2 3 4

10.

MODE a,

x.

CAMNET A Pef Prf Pr

x.

BLOR- ine Perf Per? Pert

OIYEY Now Nlow Noue Noul

McCLINTIC MR Gan Gain A

x x
CAWrFREM B Pert Perf U

A-35
 Communications
1-A. Frequency with which accumulated intelligence information is passed
is input.
1-B. Queueinq of messages in the communication net (with resultant
delays) is simulated to generate a time of reception.
4-A. A fixed percentaQe of messages (orders) are not received, to
represent the effects of jamming.

4-B. Messages are accurate if received, but probability of reception is

based on carrier propogation theory and on whether the receiver is
jammed.

4%.

.4

...

A-36

A................ ........ .-. . . .- . ...

 y i*7~
_q

MODEL PROCESS
MISS-ION ASSIGNMENT, ORGANIZATION FOR COMBAT, AND
RESOURCE ALLOCATION -

DEFINITION:

REPRESENTATION OF MISSION ASSIGNMENT, FRONTING, AND

ORGANIZATION FOR COMBAT OF MANEUVER UNITS; MISSION

ASSIGNMENT, ALLOCATION, AND POSITIONING OF COMBAT

SUPPORT AND SERVICE SUPPORT RESOURCES; AND THE

ASSIGNMENT OF MISSIONS TO AIRCRAFT

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* INITIAL DEPLOYMENTS OF MANEUVER UNITS

* CHANGES IN THESE DEPLOYMENTS

* CROSS ATTACHMENTS AMONG MANEUVER UNITS

* ALLOCATION AND/OR POSITIONING OF

* ATTACK HELICOPTERS

* FIELD ARTILLERY FORCES

* CAS/BI AIRCRAFT

* AIR DEFENSE FORCES.

* TARGET ACQUISITION RESOURCES

* SUPPLY POINTS

* ALLOCATION OF AIRCRAFT TO MISSIONS

* DELAYS IN MAKING AND IMPLEMENTING DECISIONS

A- 37

..* ..-. ,..'. .. ,, ,,.,.- , -.-.. ..-. ,-.......' _... .. *. -,. . . *..- ,- -. . ' - -.. . . - . . . ,, ... -.
-a'7

,41

0
0U~"01 P=

z -

co '0~

=
I
Os U , 0
o~ ~ 41 00'

U, I,
0vii 0I
-40,5 suA

I 5

-lo/ 411(4
soII

CD 0 p
a,-~~~. - oa Oa - a*

CDA-30

M L.a,
a .5'1" /)
i.4.
a..0c .
., Mission Assignment, Organization for Combat, and Resource Allocation

I-A. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

2-A. Built-in rules proced through command heirarchy, (1) allocating

reinforcing divisions, (2) replacing weak front-line divisions or
brigades, (3) deciding whether to conTnit or hold a reserve corps in
an army, a reserve division in a corps, and (for Blue) a reserve
brigade in a division, (4) adjusting frontages, based primarily on
unit missions, strength estimates (and corresponding force ratios),
frontage restrictions, and penetration depths. Input thresholds
govern many decisions.

2-B. Either input directly by garners or governed by gamer-suoolied

contingency orders, which can be activated as a function of time,
unit location, casualty level, or movement status of a specified
unit.

2-C. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

3-A. Input for battalions and brigades. For higher-level units, see
2-A.

3-B. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

4-A. Governed by user-provided tactical decision rules. Existing rules

in most versions allocate forces not organic to a division as a
function of friendly to enemy force ratios and the number of
maneuver units being supported (model then assumes most forces
"float" with the FEBA).

4-B. For Blue only, model rules allocate corps helicopters tc divisions
in proportion to artillery allocation; allocate to brigades equally
with a later adjustment possible (if one brigade is in reserve)
based on estimated outcome of the current period's combat.

4-C. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

5-A. Governed by user-provided tactical decision rules. Existing rules

in most versions allocate forces not organic to a division as a
function of friendly to enemy force ratios and the number of
maneuver units being supported (model then assumes most forces
"float" with the FEBA).

5-B. Model rules allocate artillery battalions through the command

heirarchy. Divisional artillery for front-line divisions is used
as DS for the division. A fraction of corps artillery (and Blue
reserve division' artillery) may also be used in this role if
estimated by division to be needed for a favorable outcome.
Artillery is allocated down to division generally in support of
strength on offense and weakness on defense, where strength is
measured by forc- ratios.

A-39

-li . .
 5-C. Gainers set priorities for allocation of artillery to support
specific maneuver units. Movement of artillery is automatically
initiated in the model to assure all units are between 30 percent
and 70 percent of their maximum range from the FEBA. Such movement
can be precluded by ganer input.

5-0. Governed by built-in decision rules designed to represent Blue and

.4..
Red doctrine in Blue division-level defensive scenario.

6-A. Governed by user-provided tactical decision rules. Existing rules

- in most versions allocate forces not organic to a division as a
function of friendly to enemy force ratios and the number of
maneuver units being supported (model then assumes most forces
"float" with the FEBA)

6-B. CAS aircraft are allocated to divisions in proportion to artillery

allocations.

6-C. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

7-A. Governed by user-provided tactical decision rules. Existing rules

in most versions allocate forces not organic to a division as a
function of friendly to enemy force ratios and the number of
maneuver units being supported (model then assur-- most forces
"float" with the FEBA).

8-A. Governed by user-provided tactical decision rules. Existing rules

in most versions allocate forces not organic to a division as a
function of friendly to enemy force ratios and the number of
maneuver units being supported (model then assumes most forces
"float" with the FEBA).

- 8-8. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.
9-A. Governed by user-provided tactical decision rules. Existing rules
in most versions do not reallocate or relocate supply points
(although model assumes they "float" with the FEBA).

9-8. Model automatically updates positions to keep them within the

bounds of their parent units.
" 10-A. Governed by user-provided tactical decision rules. Existing rules
base these allocations on acquired targets, suitability of the
aircraft for the possible missions, and current "air posture" -- a
function of aircraft inventory, recent aircraft loss exchange
ratio, and current mission of ground forces.

10-8. Percentage of aircraft on each mission are input for first period
and adjusted thereafter automatically as. a function of aircraft
loss rates in the air and on the ground. If recent FEBA movement
is rearward and exceeds input thresholds, aircraft otherwise
assigned to other missions are diverted to CAS.

". A-40

., . . . . . . . . .

. -
*.*.*~~~ . . . . . . . .~ . . . .. ..
*
'
11-A. Commitment of army and corps reserves occurs after input del ays
(which include
decisions implementation
are implemented without as well as travel time). Other
time delay. Ii .

11-B. Staff element processing times are modeled explicitly using a

priority queueing structure for handling incoming messages.
Minimum processing times and maximum acceptable throughput times
are input for each message type. (Messages are lost if the system
is overloaded.)

.4°

4 ..

-I

N.N

. . . * .

A-41
.4

MODEL PROCESS
- MANEUVER CONTROL -

* DEFINITION:

REPRESENTATION OF THE CONTROL OF ACTIVITIES AND

SUPPORT OF FRONT-LINE MANEUVER UNITS 4L

PHENOMENA INFLUENCED BY HUMAN FACTORS:

* DECISIONS TO CHANGE THE ACTIVITY AND/OR LOCATION OF

A UNIT (INCLUDING CRITERIA FOR INITIATING AND

BREAKING OFF FROM COMBAT)

* DECISIONS TO REQUEST AND COMMIT RESERVES

* DECISIONS FOR UNIT RETIREMENT

0 DECISIONS TO REQUEST FIRE SUPPORT

* DECISIONS TO REQUEST ENGINEERING SUPPORT

DECISIONS FOR THE TACTICAL RELOCATION OF SUPPORTING

ELEMENTS (SUCH AS FIELD ARTILLERY BATTERIES)

* DELAYS IN MAKING AND IMPLEMENTING DECISIONS

A-4

Jm

~~A-42"'

*! . . . . . . . . . "

:.. .:....
.,.
....
.....
.. ...
....
.,. .-....-.- .-. . .- . • . , .- , . .. : - - ..: . . . . .. . -. . . - . , - . ...-
 7777 77 - ---- - f

MODEL PROCESS: MANEUVER CONTROL1

12 3 4 5 67

*a A
(11 'b 'b'

0 F-Q
!t0
LJ £

BLI)m NR NR NR E NR NR KR

x x x x K
IJIVLt:V 0 H B 0 NR Ext NR

FOURCE F 1) 0 G NR AB

x K
x K
JIFFY Galai Gaul Gaul Nomn Gala Galia NR

xx x x x x
VEC rUl?-2 A A .A A A Noni A

x x
CEM B Noli 14og11 B Nom 1401i NH

MCCL INT IC Galai Gala Gala Galai Gala Gamn Gaul

LASIFORNl E L. C F B IRI

rbis cateqory refers to detailed control of units in contact. For periodic reallo~cation
of forces (which may itself include soule reserve couimitmnent and retiremtent of forces) Sep
"Mission Assigrunent, Orqanization for Combhat, and Resource Allocation."

A-43
 Maneuver Control
1-A. Governed by user-provided tactical decision rules. Existing rules
inmost versions base decisions on mission of friendly force,
front-line strengths and attrition, command post strengths,
availability of artillery and air defense weapons, availability of
POL, and proximity of opposing forces and of neighboring friendly
forces.
1-B. Activity changes are represented implicitly in input firepower
scores. Location changes are based on firepower score force
ratios, and limited by rules restricting amount of exposed flank
allowed.
I-C. Engagement termination criteria include an input number of
surviving weapons on a side being within an input distance of a
user-specified location, personnel casualties exceeding an input
threshold, or vehicle casualties exceeding an input threshold.
1-0. Either input directly by gainers or governed by garner-supplied
contingency orders, which can be activated as a function of time,
unit location, casualty level, or movement status of a specified
unit.
I-E. Based on user-specified decision tables.

1-F. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.
2-A. Governed by user-provided tactical decision rules (and by the
availability of reserve forces). Existing rules in most versions
base decisions on strengths and activities of the forces involved
and, for Red, the time at the front and the amount or rate of
advance of the front-line force.
2-8. Either input directly by gainers or governed by ganer-supplied
contingency orders, which can be activated as a function of time,
unit location, casualty level, or movement status of a specified
unit.
2-C. Based on user-specified decision tables.
2-0. Governed by built-in decision rules designed to represent Blue and
Red doctrine in a Blue division-level defensive scenario.
3-A. Governed by user-provided tactical decision rules. Existing rules
in most versions base decisions on unit strengths, time at the
front, and amount of advance of the front-line force.
3-B. Either input directly by gainers or governed by ganer-supplied
contingency orders, which can be activated as a function of time,
unit location, casualty level, or movement status of a specified
unit.

A-44

I .. ~~~~* -. ... *.*.*.u

... ~ ~ i ~ |H w e ~ m
"-y - -' .4-7
- - , - .'4' ' - - L . .--
'
. .-7-7777
-" U . b. -. . .: . .- ,.- . .. .. . . . . . ...

3-C. Based on user-specified decision tables.

3-0. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

4-A. Governed by user-provided tactical decision rules. Existing rules

in most versions base decisions primarily on current or scheduled
maneuver unit activities.

4-B. Primarily represented implicitly in input firepower scores. However,

DS artillery can be employed at two ordnance expenditure levels --
division commander requests higher level for next division cycle
if estimate of situation indicates it is necessary to
accomplishment of mission.

4-C. Command units base calls for artillery and attack helicopter
support on target information supplied them by their subordinate.
Target priorities for such calls are input.

4-0. Calls for artillery and helicopter support occur automatically in

the model as a function of loss rates. (Artillery is requested if
current loss rate exceeds ten percent of the player-input
acceptable loss rate.)

4-E. Forward observers who call for fire in areas containing ac-
quired targets are simulated using input priorities. In some
versions, preplanned fire is delivered to specified positions
on the battlefield at specified times, based on input.

4-F. Employment of helicopters is governed by user-specified decision

tables.

4-G. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

5-A. Requests for laying minefields are governed by user-provided

tactical decicion rules. Existing rules in most versions do not
make such requests.

5-B. Based on user-specified decision tables.

6-A. Relocation of command posts above battalion and Blue artillery

batteries is governed by built-in decision rules. Relocation of
other supporting elements is represented implicitly by maintaining
a fixed relationship with the unit they support.

* 7-A. Durations of delays are input as a function of type of decision,

whether the force is engaged, which headquarters (battalion,
brigade, etc.) participate in the decision and (if the decision is
a change in activity) the type of activity to be adopted.

7-B. Staff element processing times are modeled explicitly using a

priority queueing structure for handling incoming messages.
Minimum processing times and maximum acceptable throughput times
are input for each message type. Messages dre lost if the system;
is overloaded.
A-45

.................................
. o

. . . . . . . . . . . . . . . . . . . . . .. i
 MODEL PROCESS
- FIRE SUPPORT ALLOCATION -

DEFINITION:

REPRESENTATION OF THE ALLOCATION OF AVAILABLE FIRE

SUPPORT RESOURCES (ARTILLERY, FIXED-WING AIRCRAFT,

ATTACK HELICOPTERS, AIR DEFENSE WEAPONS) FOR USE ON

PREPLANNED MISSIONS AND AGAINST TARGETS OF

" OPPORTUNITY

%.

PHENOMENA INFLUENCED BY HUMAN FACTORS:

0 TARGET AND MISSION PRIORITIES

* PREFERENCES FOR TYPE AND AMOUNT OF SUPPORT AGAINST

A GIVEN TARGET IN A GIVEN SITUATION

, PREFERENCES FOR TYPES AND AMOUNTS OF FIRE ALLOCATED

TO PREPLANNED MISSIONS (PREPARATION, FINAL PROTEC-

TIVE FIRES, ETC.)

. CRITERIA FOR IGNORING TARGETS OR DROPPING THEM FROM

THE LIST OF ACQUIRED TARGETS

DELAYS IN MAKING AND IMPLEMENTING DECISIONS

'p.p

.*-°
A4

':ii A-46

:... . . . . . . . ,- . . . . . . . .. - -

"° ' " ° - "' ° "' ''

r ". %* .-.
"W"%"''''
,,' " % .-..
" "-'. "".".. ",
,, ',,°'"•.i,,, ."".". " """''".".. -" . .'-... . . "..
. . . ."..
.". . .".. ."."." "."-.". ".
 MODEL PROCESS: FIRE SUPPORT ALLOCATION

t12 3 4 5 . -

, o ....
q,0 4q 0 c

DV V- A.NR 0

FARMONE
TE 0 ... NR .
JIFF C
x x x C)NRN x
DIVIEV I A A HR In

x x x x

JIFFY C B B fiR NR :"

FCE A In In Hit CR
VECTOR-2 In in In A In""'

x
x x
x x
x x x .:
CEM A In In NR NR

McCLINTIC Gam Gam Gain Gain A

CASIFORFM "-."

ICARMONETTE and BLDM represent fire s,,pport only in response to calls

from engaged maneuver elements. See "Artillery Fire" and "Maneuver
Control."

A-47

0..~ . *~*~,*

* ~~' *~ *~~ .~~ . .o

Fire Support Allocation

1-A. Selection of artillery mission types (and corresponding target

types) depends on divisional commander's estimates of success at
each of two levels of artillery allocated to the direct support
role. Artillery not in the DS role is allocated implicitly (based
on effects data) between enemy artillery and reserve maneuver
units. See "Mission Assignment..." section for allocation of
aircraft to missions. Allocation to targets within a given mission
type is generally assumed to be uniform.

1-B. Priorities for support of maneuver units are input; priorities for
attack of targets in the rear are built into the model. Maneuver
unit requests have priority over deep targets.

1-C. Available artillery fire of a given type is distributed among

target types appropriate to that artillery type in proportion to
weights, each of which equals the product of the number of targets
and an input "military worth" of the target. If the gamer calls
for final protective fires, all indirect fire means are used for
this purpose for an input time period.

1-0. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

2-A. Based on built-in allocation factors which do not vary with

situation.

2-B. Available artillery fire of a given type is distributed among

target types appropriate to that artillery type in proportion to
weights, each of which equals the product of the number of targets
and an input "military worth" of the target. If the gamer calls
for final protective fires, all indirect fire means are used for
this purpose for an input time period.

2-C. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

3-A. Preparatory fires based on built-in allocation factors; other fire

plans can be prespecified by the gamers to occur at indicated times
or when a unit reaches an indicated location. Allocation of
helicopters to maneuver unit can be prespecified to occur at
indicated times.

3-8. Treated within the allocation scheme described in 1-C.

3-C. Governed by built-in decision rules designed to represent Blue and

Red doctrine in a Blue division-level defensive scenario.

4-A. Targets are ignored if support is unavailable which can reach the
target. "Stale" targets are dropped from the list after an input
period of time.

A-48
5-A. In addition to garner delays in making decisions, a 15 minute delay
between requesting and receiving fire is assessed.

5-B. Time to shift fire to a new target is input as a function of

artillery type.

5-C. Staff element processing times are modeled explicitly usinq a

priority queueing structure for handling incoming messages.
Minimum processing times and maximum acceptable throughput times
are input for each message type. Messages are lost if the systeml
is overloaded.

X.• ._

::K:~K§iK§K:
:.-. ,

A-49
 MODEL PROCESS
RESUPPLY AND REPLACEMENT -

DEFINITION:

REPRESENTATION OF THE STOCKPILING OF SUPPLIES AND

INDIVIDUAL REPLACEMENTS AND THEIR ALLOCATION TO USER

UN ITS

PHENOMENA INFLUENCED BY HUMAN FACTORS:

• LOCATIONS OF SUPPLY POINTS

* STOCKPILING DECISIONS AT SUPPLY POINTS

" QUANTITIES OF SUPPLIES AND REPLACEMENTS ALLOCATED

TO UNITS ON THE BATTLEFIELD

- DELAYS IN DECISIONS TO PROVIDE SUPPLIES OR

REPLACEMENTS

. TIME REQUIRED TO LOAD AND UNLOAD (OR OTHERWISE

HANDLE) ALLOCATED SUPPLIES

45..

r' A-50

.
+,,i,, • , .+, .. .. ." "+., ". " .. -. ,- C ..
* o. * C - °
*" ' . . -

''
I . .. . . + . " " " • + ': ".. .". .-..
'. . . . ... - + " - '
MODEL PROCESS: RESUPPLY AND REPLACEMENT

12 4

I.,Ke

j 4i

CARMONETQ

BLOMq

'4 ..
xL 'J
DILE II coN 4i

FO'RC0

JIFFY

Ii IN nI 0c a dl i'

_x x x

CSroR-2 III An ABN

A-51x
 Resupply and Replacement

2-A. Amounts to be added to stockpiles periodically are directly input

by the user. User-specified resupply rules can assure that these
stockpiles remain at acceptable levels; most existing rules do not
g do so.
2-B. Amounts to be added to stockpiles periodically are directly input
by the user.

2-C. Initial stocks are set by the model as a function of the number and
type of units for which the supply area is responsible.

3-A. Governed by user-provided tactical decision rules. Most existing

rules attempt to resupply so as to minimize fractional shortfalls
from TOE levels.

3-B. Specified by the user via input decision tables.

3-C. Available resources are distributed to units in proportion to unit

need to replace losses and consumption. Optionally, replenishment
of Red units may be limited to "decimated" divisions.

3-0. Amount of resupply requested is twice the current shortfall from

maximum supply load, or zero, if the shortfall is less than 1/8 the
maximum load. Units are resupplied in these amounts in priority
order subject to available supplies and hauling capability.

4-A. Availability of replacements is delayed an input time, which can

(in part) reflect decision delays.

5-A. See 4-A. Also, replacement personnel to a unit must be assimilaz ,'i
into the unit over a period of time. Assimilation rates are
governed by input.

.4iT.

............................................................................................

4,.. *4 . .. . . . .. . . . . . . .

... .... .... ... ....

. ..... ... .... ... .

* A-52
 . . -°

* MODEL PROCESS
MOVEMENT -

DEFINITION:

REPRESENTATION OF TRANSPORTATION AND TACTICAL MOVEMENT

OF FORCES ON THE GROUND AND IN THE AIR, RESULTING IN

NEW LOCATIONS OF UNITS, SUPPLIES, ETC., WITH

ASSOCIATED DELAYS, POL CONSUMPTION, AND CHANGES IN

VULNERABILITY

PHENOMENA INFLUENCED BY HUMAN FACTORS:

" AIRCRAFT SPEEDS AND FLIGHT PATHS

" MOVEMENT RATES OF GROUND FORCES (UNITS, SUPPLIES IN

TRANSIT, ETC.)

* ROUTE SELECTION FOR FORCES MOVING ON THE GROUND

* DEGREE OF EXPOSURE OF MOVING FORCES (AIR AND

GROUND)

off
. . . . . . . . . . . . .'

AoL

A-53
 MODEL PROCESS: MOVEMENT

12 3 4

C-, A

x x
BIVLEiI NH c d Null

.. x x x
FOURCE Noin E A B

JIFFY NRl F Nulli Now

- -xx x
VECIOH-2 A in A C

CEM Nil A A Now

McCLINTIC In aBa u

CASIFORM In 0 a A

lExludes maneuver patterns of Individual maneuver -unit

weapons -see "Maneuver Unit Comb at.

A-54

45%

0
A~~*b~* *~*.. ,4 ~ . 'A i,1 ~ .. .
 J -mr r r ~ . . - ~ .* . . ~ . -.

Movement

I-A. Average speeds are input. Simplistic, notional flight paths, built
into model logic, depend on type of mission and location of
airfields and targets.
2-A. FEBA movement is governed by firepower score force ratios.
Movement of forces in the rear is represented only through input
delays for commitment of higher-level reserves and for replacement
resources.

2-B. Specified by ganers for each move, suhject to input maximum

speeds.
2-C. Unopposed movement rates are input as a function of unit type,
terrain, on or off road, barriers, and day or night. Current
velocity is the product of unopposed movement rate and unit
posture. Posture is a linear function of current loss rate, equals
1 if loss rate is 0 and efuals 0 if loss rate equals or exceeds
input acceptable loss rate.
2-D. Maximum movement and acceleration rates are input and depend on
terrain. The ordered speed may be changed at any time using
user-specified decision tables.
2-E. Speed is based on inputs which vary with familiarity with terrain
(friendly or enemy territory) and terrain characteristics (speed
of engaged forces is further degraded -- see "Maneuver Unit
Combat," "Air-to-Ground Attacks," and "Artillery Fire").
2-F. Input as a funtion of firepower score force ratio, unit activity,
visibility conditions, type of terrain, whether weapons are mounted
or dismounted.
3-A. FEBA movement follows routes specified by an input terrain map.
Route selection in the rear is represented implicitly.

3-B. Selection from among input routes is governed by user-specified

decision tables.
4-A. Representative cover for moving forces is related to mean surface
feature height of the terrain truncated to 1 meter maximum. Other
aspects of exposure use functions of the terrain type being
traversed.
4-B. Posture of a moving Blue battalion is "hasty defense" (rather than
"fortified position" or "prepared position"). Posture affects
vulnerability to direct fire. Moving forces can also be given
different "activity factors" from stationary ones; this affects
their detectability.

4-C. A maneuver unit's movement status can affect its detectability and
posture. (Posture affects vulnerability to fire support.)

A-55

, .. -7. ....
 MODEL PROCESSES
- MAINTENANCE AND REPAIR; MEDICAL SUPPORT -

DEFINITION:

REPRESENTATION OF THE RETURN TO DUTY OF DAMAGED

FACILITIES AND EQUIPMENT AND OF INJURED PERSONNEL

, PHENOMENA INFLUENCED BY HUMAN FACTORS:

' RATE AT WHICH WEAPONS AND PERSONNEL ARE RETURNED TO

DUTY GIVEN AVAILABILITY OF REPAIR CAPABILITY AND

MEDICAL SUPPORT

- AVAILABILITY OF REPAIR CAPABILITY AND MEDICAL

SUPPORT

, RATE AT WHICH DAMAGED CAPABILITY OF A FACILITY IS

RESTORED

.. .

-- 4. ...

.. . .. .. .. .. ..
.. . ,. . .4- .. - . - . - .. . . .. . . .. . .
.. 4*=4'4 . .. . .
A-56

-J•
MODEL PROCESS: MAINTENANCE AND REPAIR; MEDICAL SUPPORT

12 3

LAb

' .fe' I

MO)L Lb

CARMONEIE - - -'-

BLOM - -

OIVLEV - - "

FOURCE - - -

x
JIFFY 0 NOin NH

x x
VECTUIR-2 A Nomii A

x
CEM B Noni NR
I.-.-.w

MCCLINIIC --

- - -- "

CASIFPURIM C A flit

A-57

~*
.,, *** **
 Maintenance and Repair; Medical Support

I-A. Temporarily damaged artillery and air defense weapons are returned
to duty at an input rate; no other weapons nor personnel are
returned to duty once damaged.

1-B. Oecimated divisions are delayed an input minumum time before they
may be returned to duty. Sick and injured personnel require an
input time in hospital before being returned to duty. Damaqed
weapons require an input repair time before being returned to duty.
They are also subject to input repair capacity constraints.

I-C. Rates to diagnose lightly, moderately, or severly wounded (damaged)

or dead personnel (ecuipment) are input. Times to aid (repair)
wounded (damaged) personnel (equipment) in the above categories are
input.

I-D. Input percentages divide weapons lost in combat into four classes:
nonrecoverable and three levels of recoverability, each level
having a different repair time. For Blue losses only, the
percentages depend on type of damaged weapon, whether the damage
was due to direct or indirect fire, and whether the damaged weapon
was attacking or defending. Operational availability probabilities
are also input and used to reflect unavailability of weapons due to
routine maintenance.

2-A. Personnel and equipment necessary for diagnosis, aid, and repair of
casualties or damaged equipment are input as a function of degree
of damage or injury and must be colocated with the damaged
equipment or injured oerson to provide the needed support. They
may be unavailable because they were not in the unit, have been
attrited, are at the wrong location, etc.

3-A. Capability of damaged command posts (reflected in magnitude of

decision delays) is restored at an input rate. Repair at other
facilities (other than airfields) is not represented.

I
A-58

• 'o,,'..o
""...-"o'..,.""o.
."-...',. . . ' . ' - '.. . -'. -" °.. ., " -. - .. . ~ ' .,. . '
 MODEL PROCESS
9 - CONSTRUCTION, MAINTENANCE, AND OPERATION OF AIRFIELDS -

I
DEFINITION:

REPRESENTATION OF AIRFIELD CONSTRUCTION, MAINTENANCE,

AND OPERATION, RESULTING IN CHANGES IN AIRFIELD

CAPACITIES, AIR SORTIE GENERATION CAPABILITIES, AND

VULNERABILITY OF PARKED AIRCRAFT

PHENOMENA INFLUENCED BY HUMAN FACTORS:

. SORTIE GENERATING CAPABILITY OF AN AIRFIELD

* RATE OF RECOVERY OF SORTIE GENERATING CAPABILITY

FOLLOWING AN ATTACK ON THE AIRFIELD

* CONSTRUCTION AND REPAIR RATES FOR RUNWAYS, AIRCRAFT

SHELTERS, AND ENTIRE AIRFIELDS

- DELAYS IN SCRAMBLING AIRCRAFT

* DEGREE TO WHICH AVAILABLE SHELTERS AND REVETMENTS

ARE IN USE WHEN AN ATTACK OF AN AIRFIELD OCCURS

*AIRCRAFT BASING RULES

A-59

limb.
IS ~ - 17- *

MODEL PROCESS: CONSTRUCTION, MAINTENANCE, AND OPERATION OF AIRFIELDS

12 3 4 5 6

0 ,

'ZFY-- -t 4 -0

CASTFONETIF---

BLOM0

01 VL V
Construction, Maintenance, and Operation of Airfields

1-A. A fixed percentage of aircraft availability is built into the model

as a function of number of days into the war. A fixed time between
sorties is imposed for a given air unit.

1-B. Tactical decision rules are structured to limit availability of

aircraft due to turnaround requirements at the airfield.

3-A. Runway repair (and damage) is not represented. Shelters and

airfields are constructed at input rates under conditions specified
by user-input tactical decision rules. Existing rules do not call
for any such construction.

4-A. Represented by input probabilities only for determining

air-to-ground effects against parked aircraft.

6-A. Governed by user-provided tactical decision rules, subject to

airfield basing constraints. Existing rules allocate aircraft to
corps or army sectors based on input TOE levels, and allocate to
bases within sectors using input preferences for basing each
aircraft type within a specific distance of the FEBA.

6-B. Aircraft are based at either primary or sanctuary bases, depending

on model inputs. (Those at sanctuary bases are invulnerable to
air-to-ground attack.)

...

A-6 1
'
.. . . . . .

APPENDIX B

A TYPOLOGY OF HUMAN FACTORS AREAS

As discussed in section 2.2, task 2 of this project involved clas-

sifying the information about model processes developed in task 1 into a

typology which allows the information to be related to potential human

factors research areas. This appendix describes details of the conduct

and results of that task.

The first step of the process involved identification of a set of

generic task clusters which (1) affect model process phenomena defined

in appendix A; and (2) encompass significant aspects of human factors.

Exhibit B-i contains a list of these task clusters and their definitions.

The task clusters associated with each of the model phenomena are indi-

cated in exhibit B-2. Exhibit B-3 gives summary statistics concerning

the frequency with which each task cluster is involved in the processes

and phenomena. As indicated in section 2.2, these statistics provide

information about the pervasiveness of certain types of tasks in combat

models.

Next, an analysis of the abilities required to perform the tasks of

each task cluster was conducted. The analysis used an ability taxonomy

and rating scheme of Fleishman (see [Fleishman, 1975] and [Theologus,

Romashko, and Fleishman, 1970]) to determine the degree to which various

abilities are required for successful performance of each task. Exhibit

B-4 lists and briefly defines the 37 abilities included in the Fleishman

taxonomy. The rating process involved evaluating the degree to which

each of the 37 abilities is required for errorless performance of each of

the 18 task clusters listed in exhibit B-I. For each task cluster, each

B-1

. . . .
 EXHIBIT B-i: DEFINITIONS OF GENERIC TASK CLUSTERS

1. Vehicle maneuver: movement of combat vehicles in contact with an

enemy force, including use of terrain for cover and concealment and
selection of firing positions.

2. Dismounted weapon maneuver: movement of dismounted combat personnel

and their weapons while in contact with an enemy force, including use
of terrain for cover and concealment and selection of firing
positions.
3. Visual acquisition: search for, sighting, and identification of
targets using (possibly aided) visual means.

4. Aiming and tracking: pointing of a direct-fire weapon (or

designator) at a stationary or moving target for the time required to
effect successful ordnance delivery.
5. Weapon operation: processes associated with loading and firing a
weapon.

6. Individual decision making (threat): selection of an alternative

course of action by an individual or crew commander where the
consequence of the decision may pose a threat to the safety of the
decision maker (e.g., a decision to engage a particular target).
Includes processing of any sensory inputs; excludes any physical
response to the decision.
7. Individual decision making (non-threat): selection of an alternative
course of action by an individual or crew commander where there is no
immediate danger associated with the decision (e.g., an ordnance
selection decision). Includes processing of any sensory inputs;
excludes any physical response to the decision.

8. Command decision making: selection of an alternative course of

action for a force where the decision must be made under time
pressure (e.g., a decision to break off an engagement). Includes
processing of any sensory inputs and initial communication of
decision to subordinates.
9. Command planning: selection of a future course of action for a force
where there is comparatively little time pressure (e.g., decisions
associated with pre-planning of air flights for the following day).
Includes processing of any sensory inputs and initial communication
of decision to subordinates.

10. Radar acquisition: detection and reporting of targets with the use
of a radar screen.
11. Target and intelligence development: processing of target
acquisitions and other information to determine the identity,
location, strength, intent, etc., of opposing forces either for
targeting or order-of-battle intelligence.

B-2

. ............ °• . . ..-.. •.°.. ..... . .

... o % ,.. •' .%...o...' ......
•
... . . . . . . . . . . . . . . .. . . . . . . . . . . . .-.. . . . . . . ., a,,- - -. , .. . .... - - . .. ..
 EXHIBIT B-i: DEFINITIONS OF GENERIC TASK CLUSTERS
(Concluded)

12. Communication: sending and receiving messages (assumed to be by

voice over radio or telephone).

13. Driving: movement of any vehicle while out of contact with enemy
maneuver forces, possibly in coordination with other vehicles.

14. Aircraft operation: operation of an aircraft in flight or in

preparation for flight.

15. Construction: operations involving the building or repair of

bridges, roads, runways, aircraft shelters, etc., as well as the
preparation of defensive positions and the emplacement of
conventional land mines.

16. Maintenance and repair: performance of routine or emergency repair ,

of mechanical or electronic equipment.

17. Medical care: provision of emergency and non-emergency care to the

sick and wounded.

18. Resupply: loading, unloading, and moving supplies (excludes driving

see3)

,. . .. ,.'.
. .. -.-...

S... . B-3
. . . . . ...-
 EXHIBIT B-2: TASK CLUSTERS ASSOCIATED WITH
EACH MODEL PROCESS PHENOMENON

PROCESS PHENOMENON ASSOCIATED TASKS

Maneuver Unit Combat:

1. Force participation 1,2,8
2. Initial deployments of weapons 1,2,9
3. Maneuver patterns of individual 1,2
weapons
4. Target acquisition capabilities 3
5. Target identification capabilities 3
6. Target selection (and target 6
switching)
7. Decision to fire 6
8. Ordnance selection 7
9. Time to fire or firing rate 5
10. Accuracy of fire 4
11. Fire suppression 1,2,3,4,5,6
12. Employment of obscuration 8

Artillery Fire:
1. Ordnance selection 7
2. Target element acquisition and 3,6
selection (if point fire)
3. Decision to fire 6
4. Delivery pattern 7
5. Delivery accuracy 4 or 5
6. Target posture 1,2,6,13
7. Suppressive effects of fire 3,4,5,6,13
8. Effects of fire on communications 8,12
and decision-making ability
9. Time to fire or firing rate 5
Air-To-Ground Attacks:
1. Ability to acquire (or reacquire) 3
the target
2. Engagement decision (whether to 6
attack the target)
3. Ordnance selection 7
4. Target element acquisition and 3,6
selection (if point fire)
5. Aircraft standoff distance and/or 14
maneuver pattern over target
6. Weapon delivery accuracy 4 or 5
7. Target posture 1,2,6,13
8. Suppressive effects of the attack 3,4,5,6,13
9. Effects of the attack on communica- 8,12
tions and decision-making ability

-- Continued --

B-4 '

"->: " ".-.--.L..--.'.!.

-i"'"'-" ."" -'i.". -'..'.i
" - . '" . .. . .. . . .- . . . . .. '

-
:.
, -.':':,..
22:,.- .-:--
-='..'.'.
--.---,.."
.'.'.i.-- '- -,- - :: - ( , . , .- . -.-,:, , , , ,. -. '
 EXHIBIT B-2: TASK CLUSTERS ASSOCIATED WITH
EACH MODEL PROCESS PHENOMENON
(Continued)

PROCESS PHENOMENON ASSOCIATED TASKS

Air-To-Ground Attacks (cont.):
10. Effects of the attack on air- 15,16
field sortie generation capability
(if target is an airfield)
11. Aircraft disengagement decision 6

Air-To-Air Combat:
1. Engagement decisions (whether to
engage and type of engagement) 6
2. Aircraft maneuver patterns 14
3. Target acquisition capabilities 3 or 10
4. Target selection 6
5. Ordnance selection 7
6. Time to fire 5
7. Accuracy of fire 4
8. Abort decisions 6
Air Defense:
1. Coordination of fire 6,9
2. Target acquisition capabilities 3 or 10
3. Target selection 6
4. Decision to fire 6
5. Degraded capability of aircraft to
destroy targets on the ground 14
6. Aircraft mission abort decisions 6
Mobility, Countermobility, and Survivability:
1. Location, density, and extent of a
minefield or obstacle 15,9
2. Attrition suffered when encounter-
ing a minefield 8,1
3. Delay in movement when encountering
a minefield or obstacle 8,1
4. Reduction in movement rates due to
minefields and obstacles 8,1
5. Increase in exposure due to
encountering a minefield or obstacl 8,1
6. Delays in traveling on damaged roads 13,15
7. Delays in crossing rivers or gaps 13,15
8. Degree of protection afforded by a
prepared defensive position 1,2,15
9. Location and extent of a prepared
defensive position 9,15

-- Continued --

B-5
 07~~~ % 7 •-

EXHIBIT B-2: TASK CLUSTERS ASSOCIATED WITH

EACH MODEL PROCESS PHENOMENON
(Continued)

PROCESS PHENOMENON ASSOCIATED TASKS

Intelligence and Electronic Warfare:
1. Collection management decisions -
what information to pursue 9
2. Collection management decisions -
how to collect desired information 9
3. Target acquisition capabilities 10
4. Accuracy of target identification
(including false targets) 11
5. Target processing time 11
6. Target location error i"
7. Timeliness of intelligence
processing I1
8. Accuracy of intelligence processing 11
9. Completeness of intelligence
processing 11
10. Decisions to jam 8
11. Effectiveness of jamming 12
12. Collection management decisions -
location of sensors 9

Communications:
1. Time to transmit message 12
2. Message content 12
3. Transmission accuracy 12
4. Reception accuracy 12

Mission Assignment, Organization for Combat,

and Resource Allocation:
1. Initial deployments of maneuver
units 9
2. Changes in these deployments 9
3. Cross attachments among maneuver
units 9
4. Allocation and/or positioning of
attack helicopters 9
5. Allocation and/or positioning of
field artillery forces 9
6. Allocation and/or positioning of
CAS/BI aircraft 9
7. Allocation and/or positioning of
air defense forces 9
8. Allocation and/or positioning of
target acquisition resources 9
9. Allocation and/or positioning of
supply points 9

-- Continued --

B-6

, . -. --- . .. . -. . . . . . - l
 EXHIBIT B-2: TASK CLUSTERS ASSOCIATED WITH
EACH MODEL PROCESS PHENOMENON
(Continued)

PROCESS PHENOMENON ASSOCIATED TASKS

Mission Assignment, Organization for Combat,
and Resource Allocation (cont.):
10. Allocation of aircraft to missions 9
11. Delays in making and implementing
decisions 9,13,14
Maneuver Control:
I. Decisions to change the activity
and/or location of a unit 8
2. Decisions to request and commit
reserves 8
3. Decisions for unit retirement 8
4. Decisions to request fire support 8
5. Decisions to request engineer
support 8
6. Decisions for the tactical
relocation of supporting elements 8
7. Delays in making and implementing
decisions 8

Fire Support Allocation:

1. Target and mission priorities 8
2. Preferences for type and amount of
support against a given target in
given situation 8
3. Preferences for types and amounts
of fire to be allocated to pre-
planned missions 9
4. Criteria for ignoring targets or
dropping them from list of acquired
targets 8
5. Delays in making and implementing
decisions
2 8 .7

Resupply and Replacement:

1. Locations of supply points 9
2. Stockpiling decisions at supply
points 9
3. Quantities of supplies and replace-
ments allocated to units on the
battlefield 9-
4. Delays in decisions to provide
supplies or replacements 9

-- Continued --

B-7

. .* . "
 EXHIBIT B-2: TASK CLUSTERS ASSOCIATED WITH
EACH MODEL PROCESS PHENOMENON
(Concluded)

PROCESS PHENOMENON ASSOCIATED TASKS

Resupply and Replacement (cont.):

5. Time required to load and unload
allocated supplies 18
Movement:
1. Aircraft speeds and flight paths 14
2. Movement rates of ground forces 13
3. Route selection of forces moving
on the ground 9
4. Degree of exposure of moving forces
(air and ground) 13,14

Maintenance and Repair; Medical Support:

1. Rates at which weapons and
personnel are returned to duty
given availability of support 16,17*
2. Availability of repair capability
and medical support 9
3. Rate at which damaged capability of
a facility is restored 15,16*

Construction, Maintenance, and Operation of

Airfields:
1. Sortie generating capability of an
airfield 16,18
2. Rate of recovery of sortie
generating capability following an
attack on the airfield 15,16,18
3. Construction and repair rates for
runways, aircraft shelters, and
entire airfields 15
4. Delays in scrambling aircraft 14,16
5. Degree to which available shelters
are in use when an attack of an
airfield occurs 9,14
6. Aircraft basing rules 9

*Also included here might be other tasks whose effective performance can
be damaged and require time to be reinstated or relearned.

B-8

, . _ . . . '" .
.'.''"'''" ."' . ' . ". . . . .
'"""'" "' ' " " " • ., , . ," . , l ,. , .
 *- *-.~ *--' * - * J --. 7 . . . . . . . . ..

EXHIBIT B-3: TASK ANALYSIS -- SUMMARY STATISTICS

NUMBER OF PROCESSES NUMBER OF PHENOMENA

TASK INVOLVING THIS INVOLVING THISI
CLUSTER TASK CLUSTER TASK CLUSTER

1 4 1
2 4 7

3 5 10
4 4 7

5 4 8

6 5 19*

7 4 5

8 7 20*

9 10* 27*

10 3 3
111 6

12 4 7

13 5 9
14 6 8

15 4 9

16 3 6

17 11
18 2 3

*Asterisks mark especially high values.

B-9

La * ~
 EXHIBIT B-4: FLEISHMAN ABILITY TAXONOMY

• 1. Verbal Comprehension: the ability to understand lanquaqe.

2. Verbal Expression: the ability to utilize lanquaqe (either oral or
written) to communicate information or ideas to another person or
persons. i:

3. Ideational Fluency: the ability to produce a number of ideas

concerning a given topic.

4. Originality: the ability to produce unusual or clever responses

related to a given topic or situation.
5. Memorization: the ability to memorize and retain new information
which occurs as a regular or routine part of the task.

6. Problem Sensitivity: the ability to recoqnize or identify the

existence of problems.

7. Mathematical Reasoning: the ability to reason abstractly usinq

quantitative concepts and symbols.

8. Number Facility: the ability to manipulate numbers in numerical

operations.

9. Deductive Reasoning: the ability to apply general concepts or rules

to specific cases or to proceed from stated premises to their loqical
conclusions.

10. Inductive Reasoning: the ability to find the most appropriate

general concepts or rules which fit sets of data or which explain how
a given series of individual items are related to each other.

11. Information Orderinq: the ability to apply rules or objectives to

given information in order to arrange that information into the best .
or most appropriate sequence.

• :12. Cateqory Flexibility: the ability to produce alternative qroupinqs

or categorizations for a set of items, based upon rules or
'* specifications produced by the individual who is carrying out the
categori zation.

13. Spatial Orientation: the ability to maintain one's orientation with

respect to objects in space or to comprehend the position of objects
in space with respect to the observer's position.
I
14. Visualization: the ability to manipulate or transform the visual
images of spatial patterns or objects into other spatial
*' arrangements.

-- Continued --

B-1 0

N. V,. . . ...
:--
 EXHIBIT B-4: FLEISHMAN ABILITY TAXONOMY
(Continued)

15. Speed of Closure: the speed with which a set of apparently disparate
sensory elements can be combined and organized into a single,
meaningful pattern or configuration.

16. Flexibility of Closure: the ability to identify or detect a

previously specified stimulus configuration which is embedded in a
more complex sensory field.

17. Selective Attention: the ability to perform a task in the presence

of distracting stimulation or under monotonous conditions without
significant loss in efficiency.

18. Time Sharing: the ability to utilize information obtained by

shifting between two or more channels of information.

19. Perceptual Speed: the speed with which sensory patterns or configur-
ations can be compared in order to determine identity or degree of
similarity.

20. Static Strength: the degree of muscular force which can be exerted
against a fairly immovable or heavy external object in order to lift,
push, or pull that object.

21. Explosive Strength: the ability to expend energy in one or a series

of explosive muscular acts.

22. Dynamic Strength: the power of arm and trunk muscles to repeatedly
or continuously support or move the body's own weight.
23. Stamina: the capacity to maintain physical activity over prolonged
periods of time.

24. Extent Flexibility: the ability to extend, flex, or stretch muscle

groups.
25. Dxnamic Flexibility: the ability to make repeated trunk and/or limb
flexing movements where both speed and flexibility of movement are
required.

26. Gross Body Equilibrium: the ability to maintain the body in an

upright position or to regain body balance especially in situations
where equilibrium is threatened or temporarily lost.

27. Choice Reaction Time: the ability to select and initiate the
appropriate response relative to a given stimulus in the situation
where two or more stimuli are possible and where the appropriate
response is selected from two or more alternatives.

-- Continued --

B-ll

. ....,-.,.,.-.... , , . -" . -'-" ', ..."."- " ..--. .,.. '' ..." ,'- ."-1" -"- " - ,,. ., ' '' " -. . .,"" " " ;,: ' , L
I''
'"'''' ', '' "," - ' -,,..
.., . -'-'..'.'... , .,' .- -,-.'..",-... . ',. -,•,"-'. . , ,'.
 EXHIBIT B-4: FLEISHMAN ABILITY TAXONOMY
(Concluded)

28. Reaction Time: the speed with which a single motor response can be
initiated after the onset of a single stimulus.

29. Speed of Limb Movement: the speed with which discrete movements of
the arms or legs can be made.

30. Wrist-Finger Speed: the speed with which discrete movements of the
fingers, hands, and wrists can be made.

31. Gross Body Coordination: the ability to coordinate movements of the

trunk and limbs.

32. Multilimb Coordination: the ability to coordinate the movements of

two or more limbs (e.g., two legs, two hands, one leg and one hand).
33. Finger Dexterity: the ability to make skillful, coordinated
movements of the fingers where manipulations of objects may or may
not be involved.

34. Manual Dexterity: the ability to make skillful, coordinated

movements of a hand, or of a hand together with its arm.

35. Arm-Hand Steadiness: the ability to make precise, steady arm-hand

positioning movements where both strength and speed are minimized.

36. Rate Control: the ability to make timed, anticipatory motor

adjustments relative to changes in the speed and/or direction of a
continuously moving object.

37. Control Precision: the ability to make controlled muscular movements

necessary to adjust or position a machine or equipment control
mechanism.

'1

i B-12
ability was assigned a value between 0 and 7, where 0 means that the

ability was not required to perform the associated tasks, 1 means that

the ability is required at a very low level of proficiency, 7 implies

that the ability is required at a very high level of proficiency, and a

rating of 2, 3, 4, 5, or 6 means that some intermediate level of profi-

ciency is required. These proficiency ratings, which were assigned by

the project team for each of the 18 task clusters, are displayed in

exhibit B-5. Exhibit B-6 presents summary statistics concerning the

ratings. As discussed in section 2.2, these statistics provide an indi-

cation of the degree to which various types of abilities are required by

the phenomena represented in combat models.

. . . . . . . . . . . . .. . . . . . . .. . . .-

... . . . .. . .. . .

B-13
 -4 1" m- -4 N~ - CN N~ N~ -4 N~ 1- CN m -

t LO
o r %DLLAM Ir MA M M LO M M -0 m LO Nl M -4

n N -4- N .- 4 N~ - -4 N~ -4 N~ CN LA 1-

V)

* LL m ci m I
m ~ C~ CI N4 N-4- L ) LO LA LA

*L L) J
CD C\NOI -4 N4 fn C - - N N4
M -\M VI
N= M M LCi
V). -4

co LAmr( LA) qr to LA, q~ q LAr M C 0D LA LO q LA LA

04 N D0- 4C 4- DC ~c LArl

-4 -4 -4 M LA
C4 1 -4 N1D-4 Ln M Ln n . U-

*l \L D -4 1-0 CD -4 -4 n M OW LA rn M

-4CQ 0-4 Mq -N MN e-4- LA ITL LA LOLAt

-4 -4 -4 -4 r-4 P" -4 -4 -4

B-14
 0 M LL)
, Vn M LO, N U) - N~ -4 eQ -4 N~ N~ -4 C' -4 CIQ M

rs U, 4 -4 -4 - 4 -44 C N -4 CN -4 LUn kC .0 0

f N (m. N mm m N~ -4~4 -4 m, :r Lo I N

Rr qr -4 1- -4 e-4 -4 - U ) U)N. M~ - J -4 m .-.I m

< m M -4CN -4 ~4 -4 4 -4 Mr Mf m
M, 'N M - NJ _q U,

LU -

r.0)

CD- CDln0 0D 0 0 0) 0 0nC 0 0D 0D 0D 0> 0 0 0 0)

C-

-4C
wA Ln )( )t 0 4 oC lC = C
LU-
-()
r- mrC C = nC Da a
-oV)0 O 0 0 L , 0 0 0 0
(A
CDa0 C Q k D0 )C- C D C
Ln -400
0 0 0 m0-4 0 0 0 00 0 0 0 00i

01r L - 4 - 4 - ) L tr 44 ~ D M C M -1 o L
cc~0 00 O L 0 00 0 0

N~
Lfl(% N .4 m m N
m . m m

B-14
 w 7.
7.-, - 07 . 77...7777-..77 • . . . .

EXHIBIT B-6: ABILITIES ANALYSIS -- SUMMARY STATISTICS

MINIMUM AVERAGE
PROFICIENCY PROFICIENCY AVERAGE PROFICIENCY
OVER ALL OVER ALL WEIGHTED BY NUMBER
ABILITY TASK CLUSTERS TASK CLUSTERS OF PHENOMENA INVOLVED
K1 2 00 2.6
1.7 2.6
1.9
3 0 2.6 2.7
4 0 2.9 3.4
5 1 2.9 3.1
6 1 3.6 4.1*
7 0 2.2 2.7
8 0 2.3 2.6
9 1 2.6 2.9
10 0 2.6 2.8
11 0 2.1 2.3
12 0 1.7 1.8
13 0 2.2 2.2
14 1 3.1 3.4
15 0 3.6 3.8
16 1 3.6 3.7
17 3* 4.9* 5.0*
18 1 3.3 3.7
19 2* 4.0* 4.0*
20 0 1.4 1.1
21 0 1.2 0.9
22 0 1.0 0.7
23 0 1.5 1.1
24 0 1.1 0.8
25 0 1.4 1.0
26 0 0.9 0.7
27 2* 4.6* 4.9*
28 0 1.8 1.4
29 0 1.4 1.1
30 0 1.1 0.9
31 0 0.7 0.5
32 0 1.6 1.2
33 0 1.0 0.6
34 0 1.4 0.9
35 0 1.3 0.8
36 0 1.9 1.5
37 0 2.0 1.7

*Asterisks mark especially high values.

o-.-..

B-.16
. .o..
. o,..--
 APPENDIX C

SENSITIVITY ANALYSES OF SIMULATIONS

This appendix describes work performed in examination of sensitiv-

ities of force-on-force combat models to obtain information on the rela-

tive need for model improvements in human factors areas. Included is

summary information concerning model sensitivity analyses performed

specifically for this study (in section C.1) and summary information con-

cerning previous analyses which were used to assess the sensitivity of

models to human factors assumptions (in section C.2). Finally, section

C.3 combines the assembled information into a rough assessment of the

relative potential sensitivity of combat models to assumptions made in

various human factors areas.

C.1 SENSITIVITY ANALYSES PERFORMED FOR THIS STUDY

Sensitivity analyses were conducted using two VRI models: BLDM, a

battalion-level model; and VECTOR-2, used as a corps-level model in this

analysis. Results from each model are described separately in the two

subsections which follow. Little attempt is made here to interpret the

results; the primary purpose of this section is to display summary

results in a useful format.

C.1.1 ANALYSIS WITH BLDM

The BLDM model was employed in a NATO scenario with a delaying US

company team opposed by two reinforced Soviet tank battalions. A 1990s

time frame was assumed. The followi, 4 runs were made:

(1) two baseline runs (designated runs I and 22) for comparison

with excursions;

c-1
 7 7 7. 7777,. -. .

(2) runs to investigate the impact of time required to fire a round

(runs 2, 5, 6, 9, 10, 13, and 14);

(3) two runs (numbers 3 and 12) to investigate the effect of target

priorities (or the effect of ability to identify targets);

(4) one run (number 4) to investigate the effect of target

acquisition rates;
(5) three runs (numbers 7, 8, and 15) to investigate the impact of

open-fire decisions;

(6) several runs (runs 11 and 16 through 21) which examined the

portion of the force participating in the engagement (this

could vary as a result of weapons being poorly deployed or

because individuals did not participate in the engagement);

and

(7) several runs to determine the impact of varying round

dispersions (runs 23 through 42).

The runs are summarized in exhibit C-I. For each run, the exhibit
displays both the loss exchange ratio (the number of Red weapons lost

divided by the number of Blue weapons lost) and the percent survivor

ratio (percent of Blue survivors divided by percent Red survivors) at the

end of each run (i.e., after 8.5 minutes of simulated combat). These

same two measures are plotted for several sets of the runs in exhibits

C-2 through C-li.

C.1.2 ANALYSIS WITH VECTOR-2

The analysis with the VECTOR-2 model employed a NATO scenario involv-

ing the US V Corps. A summary description of the scenario and data used

can be found in [VRI, 1981a].

C-2

"..:.
"."---'"J?.
-.- ''?'.''...................,....-...-.,........-........-..-..."-".- '-". ,-. , . -.- -
 EXHIBIT C-i: SUMMARY OF BLDM EXCURSIONS

RUN LOSS SURVIVOR

NUMBER DESCRIPTION EXCHANGE RATIO RATIO

I Baseline run 14.34 6.04

2 Double time to fire for Blue 4.56 0.93
3 Set priorities equal for Blue 8.90 3.41
4 Double time for stationary
Blue to acquire targets 14.34 6.04
5 Double time to fire first
round for Blue 5.52 1.21
6 Double time to fire subsequent
round for Blue 7.88 2.20
7 Halve the effective maximum
firing range for Blue
weapon types 1.23 0.27
8 Reduce maximum firing range
for Blue by 10% 10.44 3.85
9 Increase firing time for Blue
by 50% 6.53 1.59
10 Increase first round firinq
time for Blue by 50% 7.35 1.99
11 Reduce the number of Blue
weapons by 25% 6.86 1.05
12 Reverse order of Blue's
priorities 6.31 1.72
13 Increase firing time for Blue
by 25% 10.32 4.03
14 Increase firing time for Blue
by 75% 5.21 1.11
15 Reduce maximum firing ranqe
for Blue by 25% 4.54 0.92
16 Decrease number of Blue
weapons by 10% 10.15 3.08
17 Decrease number of Blue
weapons by 20% 7.88 1.57
18 Decrease number of Blue
weapons by 30% and 6.19 0.86
19 Decrease number of Blue
Red weapons by 10% 14.26 5.97
20 Decrease number of Blue and
Red weapons by 20% 14.09 5.84
21 Decrease number of Blue and
Red weapons by 50% 12.35. 4.85
22 Revised baseline run for use
with dispersion sensitivi-
ties 14.04 6.59

Continued ..-.

C-3

,# 4 # , *-- 'o.. - .. . . . -. - . . " . ' -. , - . , . ' ,. . ., "

'
... . . . .. ... - dW ,, "- ', ,..... b. . . L
 ", ° ~~~~~~~~~~~......j
.... .. . ..........---. ..... ... ,...... ''
-

EXHIBIT C-1: SUMMARY OF BLDM EXCURSIONS

(Continued)

RUN LOSS SURVIVOR

NUMBER DESCRIPTION EXCHANGE RATIO RATIO

* .23 Increase all Blue dispersions

by 25% using new base 10.48 4.12
. 24 Increase all Blue dispersions
by 50% using new base 9.30 2.75
25 Increase all Blue dispersions
by 75% using new base 7.95 1.96
26 Increase all Blue dispersions
by 100% using new base 6.84 1.51
27 Increase Blue first round
dispersions by 25% using
new base 11.90 5.06
28 Increase Blue first round
dispersions by 50% usinq
new base 10.96 4.43
29 Increase Blue first round
dispersions by 75% using
new base 10.07 3.46
30 Increase Blue first round
dispersions by 100% using
new base 9.49 2.96
31 Increase Blue subsequent
round dispersions by 25%
using new base 11.96 5.48
32 Increase Blue subsequent
round dispersions by 50%
using new base 10.13 3.64
33 Increase Blue subsequent
round dispersions by 75%
using new base 9.57 3.02
34 Increase Blue subsequent
round dispersions by 100%
using new base 8.48 2.25
35 Increase Blue round
dispersions following a hit
by 50% using new base 12.06 5.62
36 Increase Blue round
dispersions following a hit
by 75% using new base 11.41 5.18
37 Increase Blue round
dispersions following a hit
by 100% using new base 10.94 4.84
U,
-- Continued --

C-4

• ~~.......... ..-
' ... -. -. -' - '- '. " .• •. '. .
• , .,-.. ., .. / .". . . . . , .,,•••, ,.- - -, . .. ,,. a:..-.W ,. .m.
. .' , .. .w,-.. .,.,,.w .,a., .
.. ,a ,
 EXHIBIT C-i: SUMMARY OF BLOM EXCURSIONS
(Concluded)

RUN LOSS SURVIVOR

NUMBER DESCRIPTION EXCHANGE RATIO RATIO

38 Increase Blue round

dispersions following a
miss by 25% using new base 13.24 6.16
39 Increase Blue round
dispersions followinq a
miss by 50% using new base 11.72 4.91
40 Increase Blue round
dispersions following a
miss by 75% using new base 10.83 4.36
41 Increase Blue round
dispersions following a
miss by 100% using new base 10.04. 3.54
42 Increase Blue round
dispersions following a hit
by 25% using new base 13.37 6.27

"1

S.. ..,"

C-5

. . . . . .
 0
Lj

L6 0

ot

L.JL

4~~0 _ M

oN L
c-c

IC-
 0
wj 0

LU-

I- 4)

-C-
-W . 7. . T. 77

U-

LUJ

-0j

- U-c
000

U-C-8
 00

3]
Li 0

LL.

I-o

0-0

o1 00 c I

C-9)

s o -- )
p.'5,
0 j

-S-

Ro o
00

U.J-
0 EJ

L.J
=- z

00

I-O

U-1
 00
LLI

00

LAJ

10 9
- .,

00

00

LsL.

%- %
 0
wg

0.0

00

to.

00

U)-

0
LIJ 0

04 0 t N

C-1
 0

IdI

CDC
a- L.

C4

a0
C)

*C 1
 00

LLL

.9L
-. 4J

00

C-1
 0

ILLI

I-..

CD 0

Cl
00
LW

LoU

0C
..J 04 0

00

C--
 . .h.... .. . .. .. -..- - *

The following listing summarizes the runs that were made. Except
where noted otherwise, each run simulated a single day of combat in the

entire corps area. A quantitative comparison of run results follows the

listing.
(1) A baseline run was made with which to compare excursions. This
five-day run corresponds to the "NATO reinforced case"

described in [VRI, 1981a].

' '.
(2) Two runs investigated artillery delivery accuracy and/or target

location errors.
(3) Two runs modified artillery effects data. This could represent

changes in artillery ordnance selection or target posture.

Although not represented directly by these runs, the impact of

changes in artillery delivery patterns might be indirectly

inferred from these results.

(4) Two runs investigated target acquisition capabilities and/or

decisions to fire by air defense units. The second of these

was a two-day run.
(5) One run examined the suppressive effects of artillery in

counterbetery fire.
* (6) One run varied the effective fields of view of Blue

counterbattery radars, which could change as a result of

decisions concerning siting of sensors.

(7) Two runs examined variations in the time required to process an

acquired target by sensors used for fire support allocation.

(8) One run substantially increased communication delays (these

could also be viewed as decision delays).

'C-16
 (9) Three runs involved changes to the amount of artillery fire

allocated against a target.

(10) One run investigated the effect of round selection for

artillery (in particular, employing Copperhead rounds in

preference to conventional 155mm rounds against front-line

maneuver units).
(11) Two runs involved modifications to achievable firing rates for

artillery.

(12) One run modified the probability of finding a target for

Copperhead.

(13) Two runs investigated the impact of changes in the time to

restore to service a damaged command post.

Exhibit C-12 displays summary results of all the runs. The first

column of the exhibit indicates the group of runs involved, using the

numbers from the above listing. This is followed by a brief description -

of the modifications made for the run. Finally, the last four columns

indicate the overall combat vehicle loss exchange ratio (the ratio of Red

losses to Blue losses) and the average distance the Blue force withdrew,

and the percentage deviation from the baseline run for each of these two

measures.

C.2 SENSITIVITY ANALYSES FROM PREVIOUS STUDIES

Because of the large number of human factors areas of interest in -.

this study, it was not possible to perform model runs addressing model

sensitivity to all of them. Fortunately, model runs made for previous

studies provide some of the needed information; information concerning

these runs was assembled as part of task 3.

C-17

~j4~.........
 EXHIBIT C-12: SUMMARY OF VECTOR-2 EXCURSIONS

LOSS EXCHANGE AVERAGE BLUE

RUN DIFFERENCES FROM RATIO WITHDRAWAL
GROUP BASELINE RUN
ACTUAL % DEVIATION ACTUAL % DEVIATION

1. None
One day results- 4.7 0 10.0 0
Two day results- 4.0 0 17.5 0
2. A. Double CEPs for
all conventional
Red artillery 5.3 13 S.3 -17
B. Halve CEPs for
all conventional
Red artillery 4.2 -11 10.4 4
3. A. Double lethal
radii for all Red
conventional
artillery volleys 1.9 -60 17.4 74
B. Halve lethal radii
for all Red
conventional
artillery volleys 6.7 43 6.0 -40
4. A. Halve length of
time Red aircraft
are exposed to
air defense fire 4.7 0 10.0 0
B. Reduce length of
time Red aircraft
are exposed by 90,
(2 day run) 4.0 0 17.6 1
5. Require Blue artillery
engaged by oppos-
-. ing artillery
to be unavailable
for 30 minutes 4.7 0 9.7 -3
6. Halve effective range
of Blue counter-
battery radar 4.3 -9 10.0 0
7. A. Quadruple sensor
processing times
for Blue sensors 4.4 -6 9.8 -2
B. Set all Blue
sensor processing
times to 0. 4.7 0 10.0 0
8. Set all Blue communi-
cation delays to
-. 16 minutes 4.7 0 10.2 2
-- Continued--

C-"8

• ... ...-
.. . .-
-. '.-.. . . . . . . .. . ' .. .-

.,"
. """--
''-
..-.
'.,
'-"
-.
..-,
.,
;... . . . -. . .- - "-.-.,
, .. . . , ,,-.
, , . - -.-. . . .'--: . ::: _ i. . .. : -, . . -. ,.i:,. .. -
 EXHIBIT C-12: SUMMARY OF VECTOR-2 EXCURSIONS
(Concluded)

LOSS EXCHANGE AVERAGE BLUE

RUN DIFFERENCES FROM RATIO WITHDRAWAL
GROUP BASELINE RUN
ACTUAL % DEVIATION ACTUAL % DEVIATION

9. A. Double number of
Blue artillery
rounds allocated
per target 6.0 28 9.2 -8
B. Halve number of
Blue artillery
rounds allocated
per target 3.8 -19 10.0 0
C. Double number of
Blue artillery
rounds allocated
per target and
limit available
rounds to number
fired in baseline 5.5 17 7.9 -21
10. Modify priorities to
make Copperhead
higher priority
than conventional
155mm rounds
against Red front-
line forces 4.7 0 9.5 -5
11. A. Double Red
artillery firing
rates 4.0 -15 12.7 27 I
B. Halve Red
artillery firing
12. rates 5.7 21 6.5 -35
12. Multiply probability
of acquiring a
target for
Copperhead by 0.2 4.5 -4 10.0 0
13. A. Multiply rate at
which service is
restored to a
damaged Red com-
mand post by 0.1 4.6 -2 10.4 4
B. Multiply rate at
which service is
restored to a dam-
aged Red command
L post by 10 4.7 0 10.0 0

SC-19
 This section describes several studies which contribute evidence

concerning the sensitivity of models to human factors representations.

Each of the subsections below briefly describes one such study and

presents summary statistics concerning the human factors areas that were

investigated. These summaries are not intended as complete descriptions

of the analyses; they simply indicate the nature of any evidence concern-

ing human factors sensitivities. .

C.2.1 TERRAIN LINE-OF-SIGHT STUDY

This study [Farrell and Freedman, 1975] investigated the sensitivity

of predicted combat results to the choice of terrain lines of sight. A

Monte-Carlo combat simulation was used to represent an engagement involv-

ing 15 tanks attacking a position defended by three tanks and three

mounted anti-tank missile launchers. The study focused on the combined

effects of terrain area selection, defender location selection, and

attacker route selection on simulation results. Other model sensitivi-

' ties were also briefly addressed.

A general statement which summarizes some of the principal results

of the study is as follows: A standard 95 percent confidence interval

for the mean probability that the attacker wins the engagement, averaged

over scenarios (terrain and movement assumptions) which are believed a

priori to be equivalent, will cover a range of win probabilities at least

.50 wide (i.e., the probabilities of a win are estimable only within plus

or minus 25 percentage points). Other measures of effectiveness also

were highly sensitive to the assumed use of the terrain. Loss exchange

ratios, for example, were as much as 600 percent better for Blue on the

most favorable of "equivalent" terrains than on the least favorable.

C-20

H .1
' " ". . " ' ." "" " '. . '...- :::::
:.S.,'. .'.:. . ... ............
::'~~~~~~~~~~~~~. ".. ,,'."-q;,.. "...:."....."..
 Results of additional model sensitivity analyses are summarized in

exhibit C-13.

C.2.2 SENSITIVITY ANALYSIS OF BLDM ACQUISITION AND FIREPOWER ALLOCATION

A sensitivity analysis conducted at CACDA [Pickett and AuBuchon,

1976] found almost no difference in results of the BLDM model to changes

in maneuver unit weapon target acquisition capabilities in the presence

of high target densities. Capabilities to acquire both firing and non-

firing targets were varied. The report suggested that the insensitivi-

ties may be due to inadequacies in the detection and firepower allocation

logic of the model.

C.2.3 REACTION TIME ANALYSIS

In an analysis of scout helicopter operations [VRI, 1981b], the BLDM

model was used to investigate the effects of (among other things):

(1) scout helicopter hand-off time (the time between target detection

by a scout helicopter and the time at which an attack helicopter

learns of the target from the scout), and

(2) artillery reaction time (the time between a call for

support and the time at which that support -- a Copperhead

round -- arrives).

The results of these investigations are summarized in exhibit C-14.

C.2.4 VECTOR-O SENSITIVITY ANALYSES

A report describing detailed results of initial testing of the

VECTOR-O theater-level combat model [VRI, 1973] contains several sets of

results whose differences can be interpreted as caused by different

C-21
 EXHIBIT C-13: SENSITIVITY ANALYSES IN TERRAIN LINE-OF-SIGHT STUDY

ISSUE UNDER CONDITION PROBABILITY

INVESTIGATION VARIED DEFENDER WINS

Target assignment Select nearest target .28

logic
Select target at .26
4 random

Sensitivity to Instantaneous .26

acquisition acquisitions
Poisson-distributed .34
acquisition times

Sensitivity to .39, 2k= 51 .62

suppression parameters T = .39, z = 15 .70
T =.39, 9 = 30 .66
T= .63, z =5 .36
T = .63, Z = 15 .44
= .63, z =30 .44
t=1.0 .34
(no suppression)

it=risk threshold (accepted probability of a kill before a popdown),

2.down time per popdown (in seconds).

C-22

610 . . .
 EXHIBIT C-14: HAND-OFF TIME AND REACTION TIME VARIATIONS

VARIATIONS LOSS EXCHANGE RATIO

Scout hand-off time

(time to fire first round):

2 seconds 8.74

10 seconds 7.61

20 seconds 6.57

Artillery reaction time:

Message prep. FDC and qun
and trans. time lay time

60 sec. 180 sec. 4.26

10 sec. 180 sec. 4.56

10 sec. 30 sec. 5.90

. °.

C-2 3
- . . . - . . -.
 F71 . ..

assumptions concerning human factors. Sensitivities to the following

phenomena appear to be relevant to the present study:

(1) the ability to resupply or repair tanks;

(2) the ability to replace personnel; and

(3) attack helicopter operations in the target area (represented

via simultaneous changes in Blue helicopter effectiveness and

Red air defense effectiveness against the attacking heli-

copters).I
The results of these excursions, which are based on a NATO scenario, are

summarized in exhibit C-15.

C.2.5 DRAGON STUDY

In a study of alternative improvements to the Dragon missile system,

Bonder and Proegler, et al. [1978] used a company-level Monte Carlo simu-
lation to evaluate the effectiveness of an infantry platoon equipped with

Dragon systems in a defense against a Soviet tank company. Some of the

study results can be viewed as caused by variations in human factors

assumptions:
(1) The same forces were evaluated in each of three different
engagement types, corresponding to differing assumptions about

force deployment and maneuver.

. (2) A sensitivity analysis to assumed open-fire ranges was con-

ducted.

1This
could be interpreted as representing differing abilities of
helicopters to acquire the target, criteria for aborting a mission,
or decisions concerning standoff distance and maneuver pattern at the
target. It could also represent the effects of open-fire decisions by
the air defense weapons.

C-24

Y,:-__
" ; ,:,","-,'-"-,.''''...,-..............:..............•
.............. .. : ... •
- * *
 EXHIBIT C-15: VECTOR-O SENSITIVITY ANALYSES

RESULTS AFTER 100 DAYS OF COMBAT

CASE MEAN RED TANK LOSS

ADVANCE (KM) EXCHANGE RATIO .7%

Base 650 5.2

50% arriving individual replace- 170 6.5

ment tanks for both sides

25% arriving individual replace- 120 5.8

ment tanks for both sides

12 1/2% arriving individual 110 5.3

replacement tanks for
both sides

200% personhel arrivals for both 350 4.0

sides

Doubled Red effectiveness vs. 640 Not available

helicopters, doubled
Blue effectiveness of
helicopters

Halved Red effectiveness vs. 650 Not available

helicopters, halved Blue
effectiveness of
heli copters

C-25

. W ," W .4... % •. %%. .... _-,..... .......... ..

% % ", 4. , .* • ,4-*,'--,*.,, ., . . - • . ° . . . . . .
p. .. 7 7

(3) Two differing gunner launch decision rules were evaluated. In

the first, the probability of gunner launch was assumed to be a

function of the line-of-sight window length to the target. In

the second, the gunner was assumed to launch a round whenever a

target was detected.

While detailed results of the study are classified, the following general
statements can be made concerning sensitivities in these three areas:
(1) Results were highly sensitive to engagement type. For each

system examined, loss exchange ratios increased by more than

100 percent in changing from Blue's least advantageous to his

most advantageous engagement type. In some cases, the increase

was several hundred percent. Furthermore, the engagement type

which was most advantageous for Blue differed depending on the

system being evaluated, and the relative ranking of the systems

varied somewhat depending on the engagement type assumed.

(2) Results were moderately sensitive to open-fire ranges.

Comparison of best to worst loss exchange ratios always showed

differences considerably less than 100 percent.

(3) Changing gunner launch decision rules had no significant impact

on battle results.

C.2.6 NEUTRALIZATION RECOVERY RATE SENSITIVITY

In a study of nuclear combat using the VECTOR-i theater-level model

[VRI, 1977], it was found that results were extremely sensitive to the

recovery rate of neutralized Red units. That is, the size of the nuclear

strike required to defeat Red is sensitive to the distribution of time a

Red company which was not destroyed by the strike remains neutralized

C-26

.... ......................
because of the disruption caused by the attack. (Details of the st udy .. -

are classified.)

C.2.7 DIVOPS SENSITIVITIES

The initial study which applied the DIVOPS division-level model to a

European scenario [VRI, 1974] included an analysis of the sensitivity f

the model to assumptions used in the study. Some of these assumptions

involved human factors issues. The following excursions had a

significant impact on battle results (measured primarily by Blue's

ability to hold against Red's attack):

(1) The criteria used for initiation or termination of an attack

could affect which side was successful. For example, a

decrease in the maximum casualty level either side would acceot

before disengaging resulted in Blue being unable to hold his

position in a situation in which he was otherwise able to do

SO.

(2) A reduction in Red's division frontage resulted in Blue being

unable to stop Red's advance when he was otherwise able to do

so. (However, the reduction in Red frontages was significant,

and Blue was assumed unable to redeploy his force to counte,

Red's massing).

(3) The relative timing of reserve commitments was found to affect

results significantly, sometimes changing the course of the

battle. For example, delay of the commitment of a Red second

echelon unit until Blue was able to counterattack successfully

against unreinforced Red units meant that Blue gained the

ability subsequently to defend successfully aqainst an attac'

by the reinforced Red force.

C-27

- . -. ... . . . .. . . .° . . . .

• . .'";-.
. ',-'.
.'; .-. '.' -. ',.,,.-..
-'.-"
.. .',-'... .-.'. .-..- . .- . .. ... .- .- ..-.° .%... . . . . . . . .....-... ....-....-.-. .
 Additional excursions had little or no impact on battle results:

(1) Two ways of varying the amount of air support delivered by Blue

(dropping low priority missions versus flying all missions with

fewer sorties per mission) showed little difference in battle

outcomes for a given level of support delivered.

(2) Modifying artillery allocation priorities by removing

"stationary maneuver units with ten or more armored vehicles"

from the list of targets for Blue artillery had essentially no

effect.

(3) Reducing from ten to five the number of armored vehicles which

must be detected before requesting Blue air support had only

minor effects.

C.2.8 ENGINEER STUDY

In a study for the 1981 Army Engineering Systems Program Review, VRI

used the BLDM model to examine the benefits of improved capability to

prepare defensive positions. The purpose of the study was to assess the

additional effectiveness provided by the M9 ACE over that provided by the

07 dozer, but the results could also be interpreted in terms of different

assumptions concerning human factors. Results are summarized in exhibit

C-16 as a function of the percentage of naturally available hull defilade

positions.

C.2.9 ARAB-ISRAELI WPR STUDY

Thompson and White, et al. [1981] used the VECTOR-2 model in a veri-

fication study of the 1973 Arab-Israeli War. Various excursions were

made from the historical results to examine questions concerning the

C-28

- * *.-
',,, V 4 .44 , ,,,!,.- ,-
%-* .% . *.*-.-...,.. .,.. . .. .. .. ... , ..... . ,
-. . ., . , .- .. . .
 r -4
LAJ

0 ~44
*. ZA

tA 6A. C. CI.J C1 -

IAA

LL~J

IAI

L~AL
03
LU .~ ~ 1.6

L-1I

0 u0 -a

10

C-2
 capabilities and tactics of the forces involved; some of these excursions

involved changes in assumptions concerning human factors. Because the

study results are classified, they cannot be summarized here. However,

they have been taken into account in this assessment of model

sensit i vici es.

C.2.10 TARGET VALUE ANALYSIS

The VECTOR-2 model was used to generate combat resul ts for deter-

mining relative target values as part of the Army's Fire Support Mission

Area Analysis [Field Artillery School, 1980]. The scenario involved the

US V Corps in European combat in a 1986 time frame. Several of the

excursions from the base case can be interpreted as analyses of human

factors issues:

. (1) Four Red regiments were prevented from resuming their support-

ing attack for 11 hours on the second day of the war. (This

could be caused, for example, by a disruption in command and

control.) After five days of combat, this change had an

*" insignificant effect on loss exchange ratios, reduced Red's

penetration in the supporting attack area by 1/3, and slightly

reduced Red's penetration in the main attack area, when

compared with the base case.

(2) Modest delays (between 6 and 12 hours) were imposed on Red's

ability to commit second echelon divisions to the breakthrough

attempt in the main attack area. After five days of combat,

this change had only a minor effect on loss exchange ratios,

but reduced Red's penetration by at least 25 percent in the

main attack area and by 1/3 in the area of the supporting

attack.

C-30

. ~.%
 (3) A 12 hour delay was imposed on Red's ability to commit his

second echelon division to the supportinq attack. After five

days of combat, Red's penetration was sliqhtly less in the main

attack area and was reduced by 1/3 in the supoorting attack

are a.
4,.' "-

(4) Red's capability to detect Blue FARRPs (and thereby destroy

Blue helicopters on the ground) was denied him. In this run,

the maintenance units of two Red divisions were also orevented

from replacing or repairing -_portion of their losses. The

result was a reduction of Red's penetration by at least 25

percent in the main attack area and by 1/3 in the supporting

attack area. Examination of detailed outputs led to the

conclusion that all of this change was caused by the increased

availability of Blue helicopters; there was no apparent effect

associated with Red's decreased mainterance capability.

(5) The rate at which all Red following divisions could advance to

the front was decreased somewhat. (For example, availability

of Red second echelon divisions in the first echelon army was

delayed by one day.) This change had no major effect on loss

exchange ratios but had an extreme effect on Red's advance:

Blue was able to prevent Red's breakthrouah. Apparently the

delayed availability of second echelon forces allowed Blue to

defeat Red in detail.

C.3 SUMMARY OF RELATIVE SENSITIVITIES

This section contains estimates of the relative potential sensitivi-

" ties of models to human factors assumptions concerning each of a set of

C-31

.
. . ,,,*
. .",-*
. . .--. "'- -..- ,*" - "-'.-. , .'" . , - . . .
. '. . . . - .. . -
 .r-.-. 5V _1

model process phenomena. The estimates are based on the information

outlined in the previous two sections. The summary appears in exhibit

C-17. For each phenomenon, one of the following estimates of the

relative sensitivity is qiven:

(1) "H" if the available evidence seems to indicate that varyinq

human factors assumptions concerninq this phenomenon could

cause extreme variation in predicted combat outcomes;

(2) "M" if the evidence is mixed or if it appears that only a

moderate effect on model output could result from differinq

human factors assumptions related to the phenomenon;

(3) "L" if it appears that models are relatively insensitive to

changes in human factors assumptions concerning the phenomenon;

and
(4) "U" if no sensitivity information has been identified

concerning the phenomenon.

Most of the phenomena marked "U" concern model processes assumed to be of

little direct interest to Army behavioral scientists (e.g., air-to-air

combat). Following the estimates of sensitivity, the table contains an

indication of the previous sections of this appendix on which the

estimates were based.

C.'

.5--

C- 32 "

•. . .. ... • . . °.•

• o
S .° ",l . ." ." .. .° . . . . * . . . . . . . . . . . . . . .. . . . . . . . . . . . . ..-. . . . . . ° . .. " - "
 EXHIBIT C-17: RELATIVE SENSITIVITY ASSOCIATED WITH EACH MODEL PROCESS
PHENOMENON

PROCESS PHENOMENON SENSITIVITY RELEVANT

STUDIES1

Maneuver Unit Combat:

1. Force participation H 1.1, 2.6, 2.9 .-
2. Initial deployments of weapons H 1.1, 2.1, 2.5
3. Maneuver patterns of individual
weapons H 2.1, 2.5
4. Target acquisition capabilities L 1.1, 2.1, 2.2
5. Target identification capabilities M 1.1
6. Target selection (and target
switching) M 1.1, 2.1
7. Decision to fire M 1.1, 2.5
8. Ordnance selection U --
9. Time to fire or firing rate H 1.1, 2.3, 2.9
10. Accuracy of fire M 1.1, 2.9
11. Fire suppression H 2.1
12. Employment of obscuration U "-

Artillery Fire:
I. Ordnance selection M 1.2
2. Target element acquisition and
selection (if point fire) L 1.2
3. Decision to fire M 1.2
4. Delivery pattern M 1.2
5. Delivery accuracy M 1.2
6. Target posture M 1.2
7. Suppressive effects of fire M 1.2, 2.6
8. Effects of fire on communications M 2.10
and decision-making ability
9. Time to fire or firing rate M 1.2, 2.3
Air-To-Ground Attacks:
1. Ability to acquire (or reacquire).L 2 2.4
the target L2 2.4
2. Engagement decision (whether to
attack the target) L2 2.4
3. Ordnance selection U --
4. Target element acquisition and
selection (if point fire) U --
5. Aircraft standoff distance and/or
maneuver pattern over target L2 2.4
6. Weapon delivery accuracy U --
7. Target posture U --
8. Suppressive effects of the attack M 2.6
9. Effects of the attack on commnunica-
tions and decision-making ability M 2.10

'Numbers indicate sections within appendix C in which the studies are

descri bed.
2For
attack helicopters only. -- Continued --

C-33

" . .. • - +I + • " + • t " S+" " + ' " ] + " " * + "t • + ' + + ' ' " +° " "+ - ) + " " ' + , ,+ '. "
 EXHIBIT C-17: RELATIVE SENSITIVITY ASSOCIATED WITH EACH MODEL PROCESS
PHENOMENON
(Continued)

PROCESS PHENOMENON SENSITIVITY RELEVANT

STUDIES
Air-To-Ground Attacks (cont.):
10. Effects of the attack on air- U --
field sortie generation capability
(if target is an airfield)
11. Aircraft disengagement decision LI 2.4

Air-To-Air Combat:
1. Engagement decisions (whether to U --
engage and type of engagement)
2. Aircraft maneuver patterns U --
3. Target acquisition capabilities U --
4. Target selection U --
5. Ordnance selection U --
6. Time to fire U --
7. Accuracy of fire U -
8. Abort decisions U --

Air Defense:
1. Coordination of fire U --

2. Target acquisition capabilities L 1.2

3. Target selection U --
4. Decision to fire L 1.2, 2.4
5. Degraded capability of aircraft to U --
destroy targets on the ground
6. Aircraft mission abort decisions U --

Mobility, Countermobility, and Survivability:

1. Location, density, and extent of a
minefield or obstacle H 2.7, 2.9
2. Attrition suffered when encounter-
* ing a minefield U --
3. Delay in movement when encountering
a minefield or obstacle H 2.7, 2.9
4. Reduction in movement rates due to
minefields and obstacles H 2.7, 2.9
5. Increase in exposure due to
encountering a minefield or obstacle U --
6. Delays in traveling on damaged roads H 2.7, 2.9, 2.10
7. Delays in crossing rivers or gaps H 2.7, 2.9, 2.10
• 8. Degree of protection afforded by a
prepared defensive position M 2.8
, 9. Location and extent of a prepared
defensive position M 2.8
For attack helicopters only. -- Continued --

C-34

S aa
EXHIBIT C-17: RELATIVE SENSITIVITY ASSOCIATED WITH EACH MODEL PROCESS
PHENOMENON
(Continued)

PROCESS PHENOMENON SENSITIVITY RELEVANT

STUDIES

Intelligence and Electronic Warfare:

1. Collection management decisions -
what information to pursue M 2.10
2. Collection manaqement decisions -
how to collect desired information U --
3. Target acquisition capabilities M 2.10
4. Accuracy of target identification
(including false targets) U --
5. Target processing time M 1.2, 2.3
6. Target location error M 1.2
7. Timeliness of intelligence
processing U --
8. Accuracy of intelligence processing U --
9. Completeness of intelligence
processing U --
10. Decisions to jam M 2.10
11. Effectiveness of jamming M 2.10
12. Collection management decisions -
location of sensors L 1.2

Cormmuaications:
1. Time to transmit message L 1.2, 2.3
2. Message content U --
3. Transmission accuracy U --
4. Reception accuracy U .

Mission Assignment, Organization for Combat,

and Resource Allocation:
1. Initial deployments of maneuver
units H 2.1, 2.7
2. Changes in these deployments M 2.7, 2.9
3. Cross attachments among maneuver
units U
4. Allocation and/or positioning of
attack helicopters U -,

5. Allocation and/or positioning of

field artillery forces U -,
6. Allocation ?nd/or positioning of
CAS/BI aircraft -U
7. Allocation and/or positioning of
air defense forces U --

8. Allocation and/or positioning of

target acquisition resources U -.

9. Allocation and/or positioning of

supply points U

-- Continued --

C-35

.. .. ...... .......- . . . . • . -.
 EXHIBIT C-17: RELATIVE SENSITIVITY ASSOCIATED WITH EACH MODEL PROCESS
PHENOMENON
(Continued)

PROCESS PHENOMENON SENSITIVITY RELEVANT

STUDIES

Mission Assignment, Organization for Combat,

and Resource Allocation (cont.):
10. Allocation of aircraft to missions U --
11. Delays in making and implementing
decisions L 1.2

Maneuver Control:
1. Decisions to change the activity
and/or location of a unit M 2.7, 2.9
2. Decisions to request and commit
reserves H 2.7, 2.9
3. Decisions for unit retirement U --
4. Decisions to request fire support L 2.7
5. Decisions to request engineer
support M 2.8
6. Decisions for the tactical
relocation of supporting elements U --
7. Delays in making and implementing
decisions L 1.2
Fire Support Allocation:
1. Target and mission priorities L 2.7
2. Preferences for type and amount of
* support against a given target in
given situation M 1.2, 2.7
3. Preferences for types and amounts
of fire to be allocated to pre-
planned missions M 1.2 -

4. Criteria for ignoring targets or

dropping them from list of acquired
targets L 2.7
5. Delays in making and implementing
decisions M 2.3

Resupply and Replacement:

1. Locations of supply points U --

2. Stockpiling decisions at supoly

points U ".
3. Quantities of supplies and replace-
ments allocated to units on the
battlefield M 2.4
4. Delays in decisions to provide
%supplies or replacements U --

-- Continued --

C-36
1
 EXHIBIT C-17: RELATIVE SENSITIVITY ASSOCIATED WITH EACH MODEL PROCESS
PHENOMENON
(Concluded)

PROCESS PHENOMENON SENSITIVITY RELEVANT

STUDIES

Resupply and Replacement (cont.):

5. Time required to load and unload U
allocated supplies

Movement:
1. Aircraft speeds and flight paths U --
2. Movement rates of ground forces H 2.9, 2.10
3. Route selection of forces moving
on the ground U --

4. Degree of exposure of moving forces

(air and ground) U ."
Maintenance and Repair; Medical Support:
1. Rates at which weapons and
personnel are returned to duty
given availability of support M 2.4, 2.9, 2.10
2. Availability of repair capability
and medical support M 2.4, 2.9, 2.10
3. Rate at which damaged capability of
a facility is restored L 1.2

Construction, Maintenance, and Operation of

Airfields:
1. Sortie generating capability of an
airfield U -"

2. Rate of recovery of sortie

generating capability following an
attack on the airfield U.'-
3. Construction and repair rates for
runways, aircraft shelters, and
_ntire airfields U --
4. Delays in scrambling aircraft U --
5. Degree to which available shelters
are in use when an attack of an
airfield occurs U --
6. Aircraft basing rules U --

C-37

......................................
i...
 APPENDIX D

NOTES ON PRIORITIES FOR HUMAN FACTORS RESEARCH

As described in chapter 3.0, the information developed in tasks 1,

2, and 3 was combined to generate overall priorities for conducting human

factors research to improve combat models. Priorities were developed for

the model process phenomena (defined in appendix A), the 18 task clusters

(defined in appendix B), and the 37 Fleishman abilities (defined in

appendix B).

Exhibit D-1 contains the information used to rank the model process

phenomena. The column of the exhibit labeled "model sensitivity" con-

tains the same estimate of relative potential sensitivity to human fac-

tors assumptions as given in section C.3 of appendix C. The column

labeled "model fidelity" contains a subjective estimate of the fidelity

with which the human factors effects on each phenomenon is represented.

The estimates are based on the information presented in appendix A. A

"G" in this column means that model fidelity is typically reasonably

good; a "P" indicates generally poor fidelity. The column of the exhibit

labeled "abilities score" contains a number for each phenomenon intended

to represent the relative degree to which-human factors influence the

phenomenon. This number is the sum of the abilities ratings for all

task clusters associated with the phenomenon (see appendix B).

The final column of the exhibit ranks the phenomena in order of the

relative need of combat models for human factors research on each phenom-

enon. (A low number indicates a comparatively great need for such

research.) These ranks are based on the information in the previous

three columns of the exhibit combined as follows:

D-1

S........
--i ri_= rlit "".
iIm " . * m " . . .
* + . C .C ..+ u ..
" . . .. .. . . . * ..
+ - . .... . . .....
. . . . . ... . - • - .
 EXHIBIT D-1: DETERMINATION OF RELATIVE RANKING OF MODEL PROCESS PHENOMENA

MODEL MODEL ABILITIES OVERALL

PROCESS PHENOMENON SENSITIVITY FIDELITY SCORE RANK

Maneuver Unit Combat:

1. Force participation H G 281 5 (tie)
2. Initial deployments of weapons H G 281 5 (tie) .
3. Maneuver patterns of individual H P 199 1
weapons
4. Target acquisition capabilities L G 58 56
5. Target identification capabilities M P 58 17
6. Target selection (and target
switching) M G 69 27 (tie)
7. Decision to fire M G 69 27 (tie)
8. Ordnance selection U G 63 46
9. Time to fire or firinq rate H G 71 11
10. Accuracy of fire M G 74 24
11. Fire suppression H G 471 4
12. Employment of obscuration U P 82 37 (tie)

Artillery Fire:
1. Ordnance selection M P 63 16
2. Tarqet element acquisition and
selection (if poirt fire) L P 127 48
3. Decision to fire L P 69 51 (tie)
4. Delivery oattern M G 63 28
5. Delivery accuracy M G 74 or 71 25
6. Target posture M G 345 19
7. Suppressive effects of fire M G 349 18 (tie)
8. Effects of fire on communications
and decision-makinq ability M P 154 12 (tie)
9. Time to fire or firinq rate M G 71 26

Air-To-Ground Attacks:
1. Ability to acquire (or reacquire)
the target L P 58 52
2. Engagement decision (whether to
attack the target) L P 69 51 (tie)
3. Ordnance selection U P 63 42
4. Target element acquisition and
selection (if point fire) U P 127 34
5. Aircraft standoff distance and/or
maneuver pattern over tarqet L P 113 49 (tie)
6. Weapon delivery accuracy U P 74 or 71 39
7. Target posture U P 345 30
8. Suppressive effects of the attack M G 349 18 (tie)
9. Effects of the attack on communica
tions and decision-makinq ability M p 154 12 (tie)

-- Continued --

'°I

D-2
. . . . . . . . . .
. .°

S V %°
 EXHIBIT D-1: DETERMINATION OF RELATIVE RANKING OF MODEL PROCESS PHENOMENA
(Continued)

PROCESS PHENOMENON MODEL MODEL ABILITIES OVERALL

SENSITIVITY FIDELITY SCORE RANK

Air-To-Ground Attacks (cont.):

10. Effects of the attack on air-
field sortie generation capability
(if target is an airfield) -.
'.-
11. Aircraft disengaqement decision L P 69 51 (tie)

Air-To-Air Combat:
1. Engagement decisions (whether to
engage and type of engagement) - - -
2. Aircraft maneuver patterns - - -
3. Target acquisition capabilities - - -
4. Target selection - - -"
5. Ordnance selection - -°.
6. Time to fire - - "
7. Accuracy of fire
8. Abort decisions - - .

Air Defense:
1. Coordination of fire U P 151 33
2. Target acquisition capabilities L G 58 or 47 57
3. Target selection U P 69 41 (tie)
4. Decision to fire L P 69 51 (tie)
5. Deqraded capability of aircraft to 113 35 (tie)
destroy targets on the qround U P 69 41 (tie)
6. Aircraft mission abort decisions U P

Mobility, Countermobility, and Survivability:

1. Location, density, and extent of a
minefield or obstacle H G 167 7
2. Attrition suffered when encounter-
ing a minefield U G 177 43
3. Delay in movement when encountering
a minefield or obstacle H P 177 2
4. Reduction in movement rates due to
minefields and obstacles H G 177 6
5. Increase in exposure due to
encounterinq a minefield or obstacle U P 177 32
6. Delays in traveling on damaqed roads H P 162 3
7. Delays in crossinq rivers or qaps H G 162 8
8. Degree of protection afforded by a
prepared defensive position M G 284 20
9. Location and extent of a prepared
defensive position M G 167 22

-- Continued --

D-3

. iL ', .,.;,,. ,.- .. ,..' '. . '.. *.......'.-'.-... "... ... '.- . - " " . . ' - . .' . ' ,'
 EXHIBIT 0-1: DETERMINATION OF RELATIVE RANKING OF MODEL PROCESS PHENOMENA
(Continued)

PROCESS PHENOMENON MODEL MODEL ABILITIES OVERALL

SENSITIVITY FIDELITY SCORE RANK

Intelligence and Electronic Warfare:

1. Collection management decisions -
what information to pursue M P 82 14 (tie)
2. Collection management decisions -
how to collect desired information U P 82 37 (tie)
3. Target. acquisition capabilities M G 47 29
4. Accuracy of target identification
(including false targets) U P 87 36
5. Target processing time M P 87 13 (tie)
6. Target location error M P 87 13 (tie)
7. Timeliness of intelligence
processing 1G 87 44 (tie)
8. Accuracy of intelligence processin U G 87 44 (tie)
9. Completeness of intelliaence
processing U G 87 44 (tie)
10. Decisions to jam M P 82 14 (tie)
11. Effectiveness of jamming M P 72 15
12. Collection management decisions -
location of sensors L P 82 5G (tie)

Communications:
1. Time to transmit message L G 72 55
2. Message content U P 72 40 (tie)
3. Transmission accuracy U P 72 40 (tie)
4. Reception accuracy U P 72 40 (tie)

Mission Assignment, Organization for Combat,

and Resource Allocation:
I. Initial deployments of maneuver
units H G 82 9 (tie)
2. Changes in these deployments M G 82 23 (tie)
3. Cross attachments among maneuver
units U G 82 45 (tie)
4. Allocation and/or positioning of
attack helicopters U G 82 45 (tie)
5. Allocation and/or positioning of
field artillery forces U G 82 45 (tie)
6. Allocation and/or positioning of
CAS/BI aircraft U G 82 45 (tie)
7. Allocation and/or positioning of
air lefense forces U G 82 45 (tie)
8. Allocation and/or positioning of
target acquisition resources U P 82 37 (tie)
9. Allocation and/or positioning of
supply points U G 82 45 (tie)

-- Continued --

D-4

' .' , ' f - , > -.. - ' ......- .. -..---.... .'-. .- - .-.,] .- . ,-.
 EXHIBiT D-1: DETERMINATION OF RELATIVE RANKING OF MODEL PROCESS PHENOMENA
(Continued)

PROCESS PHENOMENON MODEL MODEL ABILITIE OVERALL

_ _ _ _ _ _ _ _ _ _ SENSITIVITY FIDELIT SCORE RANK

Mission Assignment, Organization for Combat,

and Resource Allocation (cont.):
10. Allocation of aircraft to missions U G 82 45 (tie)
"4 11. Delays in making and implementinq
decisions L G 272 53

Maneuver Control: ;j
1. Decisions to change the activity
and/or location of a unit M G 82 23 (tie)
2. Decisions to request and commit
reserves H G 82 9 (tie)
3. Decisions for unit retirement U G 82 45 (tie)
4. Decisions to request fire support L G 82 54 (tie)
5. Decisions to request engineer
support M G 82 23 (tie)
6. Decisions for the tactical
relocation of supporting elements U P 82 37 (tie)
7. Delays in making and implementing
decisions L P 82 50 (tie)

Fire Support Allocation:

1. Target and mission priorities L G 82 54 (tie)
2. Preferences for type and amount of
support against a given target in
given situation M G 82 23 (tie)
3. Preferences for types and amounts
of fire to be allocated to pre-
planned missions M G 82 23 (tie)
4. Criteria for ignoring targets or
dropping them from list of acquired
targets L P 82 50 (tie)
5. Delays in making and implementing
decisions M G 82 23 (tie)
Resupply and Replacement:
1. Locations of supply points U G 82 45 (tie)
2. Stockpiling decisions at supply
points U G 82 45 (tie)
3. Quantities of supplies and replace
ments allocated to units on the
battlefield M G 82 23 (tie)
4. Delays in decisions to provide
supplies or replacements U G 82 45 (tie)

-- Continued --

D-5

"'""'" "'""""""""*"
"":' " '".."'"".""""."'.""'
%L ' '" -------------------------------------------------
-"'"" """ ' " -"
 ~
. -*. - -- ~ 2 * ~ ~ .~..*~-. - - "- - - " • -o" . -.-. - - _"

EXHIBIT D-1: DETERMINATION OF RELATIVE RANKING OF MODEL PROCESS PHENOMENA

(Concluded)

PROCESS PHENOMENON MODEL MODEL ABILITIES OVERALL

SENSITIVITY FIDELITY SCORE RANK

Resupply and Replacement (cont.):

5. Time required to load and unload
allocated supplies U P 73 38

Movement:
I . Aircraft speeds and flight paths U P 113 35 (tie)
2. Movement rates of ground forces H G 77 10
3. Route selection of forces moving
on the ground U G 82 45 (tie)
4. Degree of exposure of moving forces
(air and ground) U P 190 31

Maintenance and Repair; Medical Support:

1. Rates at which weapons and
personnel are returned to duty
given availability of support M G 205 21
2. Availability of repair capability
and medical support M P 82 14 (tie)
3. Rate at which damaged capability of
a facility is restored L P 182 47

Construction, Maintenance, and Operation of

Airfields:
1. Sortie generating capability of an
airfield
2. Rate of recovery of sortie
generating capability following an
attack on the airfield --

3. Construction and repair rates for

runways, aircraft shelters, and
entire airfields
4. Delays in scrambling aircraft .-
5. Degree to which available shelters
are in use when an attack of an
airfield occurs - --

6. Aircraft basing rules ---

D-6
. . . ,

% %
. D -6 - .a- . - . . .

4%
 (1) Model sensitivity was taken to be the most significant cri-
terion, so the phenomena were sorted by the sensitivity esti-

mate in the following order -- H (most important), M, U, and L

(least important).

(2) Within a given model sensitivity class, those phenomena with -

poor (P) model fidelity were assumed to be of higher priority

than those with good (G) fideiity. The rationale for this

ranking was that a reasonable degree of fidelity probably means

that some knowledge already exists concerning the effect of

human factors on the phenomeno,.. Furthermore, a greater degree

of fidelity usually means that model data requirements are more

physically based, and thus the data are easier to develop.

(3) Finally, within a given model sensitivity and fidelity class,

the phenomena were ranked in order of decreasing abilities

scores. This ranking was based on the assumption that these

scores serve as a reasonable proxy for the degree to which

human factors can have an impact on each of the phenomena, and

hence the degree to which human factors research is likely to

have an impact on the representation of a phenomenon, in combat

-4 models.

Excluded from this priority scheme were those phenomena believed to be of

little direct interest to Army behavioral scientists; i.e., those involv-

ing air-to-air combat and airfield operations.

In addition to ranking the phenomena, the 18 tdsk clusters were

ranked as indicated in chapter 3.0 to provide an alternate basis for

prioritizing research on task performance. The priorities were deter-

*1 mined from a score computed for each task cluster; the score was the

:1D•
.p number of phenomena which involved the task cluster and for which model
sensitivity was not low (L).1 The score thus provides an indication of

the degree to which representation of performance of tasks in the task

cluster is likely to affect model output. The tasks were ranked in order p.

of decreasing scores. The scores and corresponding ranks can be found in

exhibit 0-2.

Finally, the 37 abilities were ranked as discussed in chapter 3.0 in

order to provide guidelines for prioritizing research on predicting abil-

ities from demographic data and estimating the effect of environmental

conditions, training, and experience on the abilities. To do this, a

score was computed for each of the abilities which was the sum of the

ratings for an ability for all task clusters associated with phenomena
for which the model sensitivity was not low (L).1 These scores are

thus intended to indicate the potential relative impact of each ability

on combat model output. The abilities were then ranked in order of

decreasing scores. The scores and corresponding ranks are displayed in

exhibit D-3.

44.

D-82

N
'phenomena involving air-to-air combat and airfield operations were not
included in computing these scores.
I
:

0)-8 "

~~A
,,w,," ...... ... . .. . .. . . . . . . . . . . . . -""
.,;'(:',
: :,.T ."."....,..".":".-. " ' ""''.""
..,'."-Tt '+. -""' . '".".' . '"."" ." . .+'- .". . +. ."-+.+ - "+-
,.:,
-,--.-.:,:-.'
, :-.-:i -:.-:;+ ."-
:..- " •? T.:-:--. . -. :. .. - - .- . +.:. + . . • -.: + - .. .
 EXHIBIT D-2: DETERMINATION OF RELATIVE RANKING OF TASK CLUSTERS

TASK CLUSTER SCORE RANK

1. Vehicle maneuver

2. Dismounted weapon maneuver

3. Visual acquisition
11

7
5
3 (tie)

5 (tie)
7
I
4. Aiming and tracking 6 6 (tie)

5. Weapon operation 7 5 (tie).

6. Individual decision making (threat) 11 3 (tie)

-. 7. Individual decision making (non-threat) 4 8

8. Command decision making 16 2

9. Command planning 23 1

10. Radar acquisition 1 11 (tie)

11. Target and intelligence development 6 6 (tie)

12. Communication 6 *6 (tie)

13. Driving 8 4
14. Aircraft operation 3 9
15. Construction 6 6 (tie)

16. Maintenance and repair 2 10

17. Medical care 1 11 (tie)

18. Resupply 1 11 (tie) 4

D-9

."•.r , ¢",,. -
x'
<,.-,-',
.'..'
.---.
',"
-..
....
.,,
..
...
..
."-.-....
...
•,
:..:;.
......
.
..
:-.-
.
...... ........... .. . ..... .
 EXHIBIT D-3: DETERMINATION OF RELATIVE RANKING OF ABILITIES

ABILITY SCORE RANK

1 355 12
2 231 18
3 341 14
4 434 7
5 388 9
6 517 2
7 356 11
8 307 15
9 357 10
10 345 13
11 285 16
12 223 19
13 248 17
14 417 8
15 459 4
16 448 6
17 582 1 (tie)
18 453 5
19 466 3
20 115 26
21 106 28
22 74 33
23 122 24
24 91 30
25 103 29
26 73 34
27 582 1 (tie)
28 169 21
29 135 23
30 108 27
31 59 36
32 121 25
33 65 35
34 88 31
35 81 32
36 152 22
37 174 20

D-1 0 082483

-*,-%*.-. , ',..r ", , ", ..- ,.. . ,-* • '. .. .- ... '. ,"...-.. . .',". .. ',"..".' ,'- - - .,. , '