Football Analysts Using Advanced Data and Models

As we're getting closer to the start of the new Premier League season, I wanted to see how much we could predict using nothing but open-source data. Rather than trying to produce one definitive league table, I built a model in R based on each team's attacking and defensive strengths when playing home and away. I then simulated the entire 2026/27 Premier League season 10,000 times. Instead of a single prediction, each team ends up with a distribution of possible finishes, which I think better captures the uncertainty and variability of football. One of the most interesting findings was just how close the title race appears to be. Man City, Arsenal and Liverpool overlap heavily across the 75–95 point range. Chelsea consistently emerge as favourites for fourth, while the middle of the table is exactly what you'd expect: incredibly difficult to separate. I was also curious to see how the model compared with the betting market. Comparing each team's average finishing position produced a Spearman rank correlation of 0.93, suggesting that a relatively simple model built using publicly available data captures a surprising amount of the same signal. It was a great reminder that probabilities tell a much richer story than a single predicted league table ever can. I've written up the full methodology, visualisations and results on my Substack if you'd like to see how the model works and what 10,000 simulated Premier League seasons think will happen this year. https://lnkd.in/ePpwM5_v #DataScience #RStats #SportsAnalytics #PremierLeague #DataVisualization

Football clubs have never had more data, however access and speed have been a consistent issue in this industry. Every ad-hoc question was reliant on an analyst, a dashboard, or a new report. Our new integration changes that. By combining our Semantic Layer with Claude, coaches, analysts, scouts and sporting directors can now interrogate Kognia data and ask tactical questions in plain language, from "tell me which young left-backs attack space best" to "show me one player's ball receptions, event by event, from his last match", and get accurate, contextual answers. Gone are the days of waiting for every follow-up question or request. A full breakdown, with real examples and worked queries, can be read here: https://lnkd.in/eQ_KBrHx

Referenced in my post yesterday, Kognia Sports Intelligence 's Semantic Layer is now available for use. People that know me know I love to work with data and I work in a data informed way. But I have no business conducting passing cluster analysis or visualising the average position change of defenders when defending crosses relative to cross location. But the integration makes it easier than ever. Plug your AI agent of choice straight in to Kognia's data and produce detailed data insights to affect how you see and analyse the game. Let me know if you want to see more.

Day 7 of My Sports Data Analytics Journey Today, I started learning the football metrics that analysts use to explain matches beyond the final score. I explored concepts such as: Expected Goals (xG) Possession Pass Accuracy Shots on Target Big Chances Ball Recoveries One lesson stood out: The score tells you what happened. The data helps explain why it happened. This shift in thinking is changing how I watch football. Instead of focusing only on the scoreline, I'm learning to understand the patterns and decisions behind each match. Through D_BRAIN, my goal is to make football analytics simple, insightful, and accessible for everyone. Question: Which football statistic do you think is the most underrated? #SportsAnalytics #FootballAnalytics #ExpectedGoals #DataAnalytics #SportsData #LearningInPublic #D_BRAIN

Most football platforms show you the output. We wanted to show what is happening behind it. Today, we launched Live Intelligence on TopStreaks — a real-time view into the football monitoring network that powers our products. It shows: 🌍 Football coverage across 28 countries and 56 competitions 🔥 More than 42,000 active streaks monitored network-wide ⚽ Streak activity connected to today’s fixtures 📊 Categories including BTTS, goals, unbeaten runs, player cards and player goals 🤖 The infrastructure powering AI Tipsters, Live Signals, Opportunity Finder, Value Bets and Match Intelligence The rotating globe highlights where football activity is being monitored, while selecting a country reveals its fixtures, competitions and streak activity. This isn’t another tips page. It’s a window into the data and monitoring layer working behind TopStreaks. Live Intelligence is now available at: https://lnkd.in/etQZMqb5 I’d love to hear what you think. #FootballTech #SportsTech #ArtificialIntelligence #DataAnalytics #BuildInPublic

Football data isn’t the problem. Understanding what it means is. A club’s transfer history can tell you where they’ve recruited, what they’ve spent and which players have moved through the squad. But the real value starts when you can connect those signals. Where are the gaps? What recruitment patterns are emerging? Why might they matter? And what could the club do next? That’s the direction we’re taking with X-1. From data → context → intelligence → action. The goal isn’t to give football professionals more information. It’s to help them make better decisions with the information they already have. Still building. Still testing. Still learning.🧠⚽ #FootballRecruitment #FootballAgent

What happens when a football club stops asking, "What can we do with data?" and starts asking, "What should we build next?" That's the conversation happening at Los Angeles Football Club (LAFC). In this third episode of Behind the Black and Gold, get a behind-the-scenes look at how LAFC and SAS are exploring everything from email optimization to digital twins and why they're already thinking years ahead about what's possible for fan engagement.- https://lnkd.in/exMrK4KY

For a hundred years, football analysis has been a one-way street. The analyst watches, takes notes, and hands the coach a report. If the coach has a question the report didn't answer, they wait; or they never get the answer at all. But that's not how understanding actually works. Understanding is a conversation. You see something, you ask why, you follow the thread, you ask again. The most valuable tools in football won't be the ones that hand you more data. They'll be the ones you can talk to, ask a question in plain language, get an answer, ask the next one. The report told you what it decided to tell you. The future of analysis answers what you actually want to know. That's the shift we're building toward. #TactiSport #FootballIntelligence #FootballAnalytics #FutureOfFootball #SaudiFootball

What happens when a football club stops asking, "What can we do with data?" and starts asking, "What should we build next?" That's the conversation happening at Los Angeles Football Club (LAFC). In this third episode of Behind the Black and Gold, you'll get a behind-the-scenes look at how LAFC and SAS are exploring everything from email optimization to digital twins and why they're already thinking years ahead about what's possible for fan engagement.- https://lnkd.in/e9v-ZWzD

What happens when a football club stops asking, "What can we do with data?" and starts asking, "What should we build next?" That's the conversation happening at Los Angeles Football Club (LAFC). In this third episode of Behind the Black and Gold, you'll get a behind-the-scenes look at how LAFC and SAS are exploring everything from email optimization to digital twins and why they're already thinking years ahead about what's possible for fan engagement.- https://lnkd.in/e4cqc28V