One Saturday night, my husband and I were sitting on the couch waiting for the Manchester United match to start when something seemed very wrong. The Coach had just announced the team lineup, and instead of playing their new striker, he had decided to use a midfielder in the striker position. My husband, who has always loved football, was immediately shaking his head, unhappy about the lineup. "The analyst got this all wrong," "The analyst got this all wrong," He continued, going into detail about why the formation didn't make any tactical sense. "They just bought a new striker, and he's clearly better than the midfielder they are using up front, but the coach is benching him because he thinks he's 'not ready.'" "They just bought a new striker, and he's clearly better than the midfielder they are using up front, but the coach is benching him because he thinks he's 'not ready.'" I kept thinking in my data analyst mind,"The numbers don't lie," as I listened to his analysis. But what numbers were the Coach really using in this case? And most importantly, were they the appropriate numbers? Was there bias? "The numbers don't lie," That question took me into a rabbit hole that changed how I see football forever. That question took me into a rabbit hole that changed how I see football forever. That question took me into a rabbit hole that changed how I see football forever. As a data analyst, I know how to be interested in how decisions are made. When someone makes a choice that doesn't make sense to me, my first instinct is to investigate the data and methods behind it. So, I looked at the most important statistics that modern football experts use to rate players. What I found out was very interesting. Modern football creates a huge quantity of data, such as pass completion rates, tackles and interceptions, Expected Goals (xG), sprint speed and distance travelled, and heat maps that show where players are and how they move. Every touch, run, and defensive move is tracked and counted.But this is where things get interesting and annoying. Even though they have all this data, coaches often make choices that don't seem to match what the statistics say. Look at Manchester United as an example. The team obviously needed a Stricker from my research. They had just spent millions on one. But the Coach decided to put a midfielder in that spot instead. From a purely data point of view, this didn't make any sense. The striker's stats were better than the midfielders in that capacity. For example, the striker scored more goals, was better at holding up play, and was better at positioning himself in the box. So why make the choice? My husband said it all comes down to one word: experience. So why make the choice? My husband said it all comes down to one word: experience. experience When Gut Feeling Beats Analytics When Gut Feeling Beats Analytics The more I watched football as an analyst, the more this trend showed itself. I remember another game where Casemiro played instead of Mainoo. It wasn't because Mainoo's stats were worse; it was because Casemiro had "more experience." From a data point of view, this choice didn't make sense. When you look at the main stats, Mainoo's numbers show that he has been completing more passes in tight places, making more progressive passes, and making more defensive plays each game in recent games. His heat maps showed better ways to move and position himself; in other words, Mainoo is faster and can cover more ground than Casemiro. The information about Casemiro told a different story. In the past, he had great performance data, but recently, they indicated slower sprint speeds, fewer successful tackles, and fewer times he kept the ball. But the coach picked experience over proof. This made me very angry as a data analyst. In my world, choices are founded on facts that can be measured. If Player A routinely beats Player B on all important criteria, Player A wins the field and not the bench. It's that easy. But the coach picked experience over proof. I was discovering, though, that football has its own set of rules. I was discovering, though, that football has its own set of rules. The more I learnt about football statistics, the more I saw that there were important factors that traditional stats don't do a good job of capturing. This example helped me understand how complicated this is. Ugarte is a good defensive midfielder; he is not good with ball possession. His numbers for tackles and interceptions are great, which is exactly what you want from a defensive midfielder. But when he's under pressure, his pass completion rate decreases a lot, and his heat maps reveal that he doesn't like to take risks with the ball. Using traditional analytics, you might think he's the best defensive player. The data don't show how his unwillingness to take risks affects the whole team's attacking rhythm, though. His teammates start to make safer passes since they know he won't try to make the hard forward pass. Even while his individual defensive stats look great, the team's overall inventiveness suffers. This is when the Coach's experience and gut feeling come in handy. An experienced coach can observe these ripple effects that aren't shown by regular measurements.This is what actually made me think that football analytics is harder than I first thought: The numbers from after the game frequently tell a very different story than the numbers from before the game. Do you remember the Manchester United game when the midfielder played as a striker? There was something noteworthy in the analysis after the game. Even though the midfielder didn't score (which was predicted), his mobility opened space for other players, which contributed to two goals. His heat map showed that he dipped deeper than a normal striker, which drew defenders with him and opened space for midfield runners. The Coach's choice seemed odd from a purely positional point of view, yet it worked out tactically. The numbers after the game confirmed a decision that looked wrong at the time. This taught me something crucial: in football and in complicated data analysis projects, the big picture is more important than the little details. This taught me something crucial: in football and in complicated data analysis projects, the big picture is more important than the little details. the big picture is more important than the little details. The Bias Problem: When Old Ways Are Better Than New Ones But I still have some doubts as an analyst. There are clearly times when football decision-making is affected by the same flaws that affect business analytics, such as confirmation bias, relying too much on prior success, and not accepting fresh facts. The bias against people based on their age and experience is really worrying. I have seen Coaches repeatedly choose older, "experienced" players over younger ones with better current performance stats, just because "they've been there before." This is like a corporation giving someone a promotion based on how long they have been with the company instead of how well they do their job, which is something most modern businesses have stopped doing. Even when new data shows that specific formations or player combinations aren't working anymore, coaches often fall in love with them again. They'll keep a player who had a tremendous season two years ago, even when their performance metrics have been going down for months. The Future: Where Data and Gut Feelings Must Come Together What I have learnt from my time with football analytics is that the sport needs to do a better job of combining data science and human intelligence. Neither method is adequate independently. Pure analytics misses important things like team rapport/spirit, psychological pressure, and tactical details that aren't shown in typical measures. But relying only on gut feelings ignores useful information that could help you make better choices. The best modern coaches are those who use data to help them make decisions instead of telling them what to do. They look at the passing stats and heat maps, but they also trust their own eyes and experience to figure out what those figures mean in the big picture. As someone who works with data every day, I like that football is educating me about the limits of using only numbers to figure things out. Numbers are honest, but they don't always reveal the whole story. The Numbers Game My football lessons on Saturday nights continue, with my husband still shaking his head at some of the tactical choices I point out. What I’ve come to realise, though, is that managing a football team just like doing good data analysis requires both quantitative rigour and qualitative judgment. The best decisions come from combining solid facts with a clear understanding of context. Numbers provide the foundation, but it’s people who give them meaning. So, the next time you see a coach make a move that doesn’t seem to match the stats, remember football, like life, is more complex than any single number can reveal. Often, the most important factors are the ones hardest to measure. That doesn’t mean we should ignore the data. Instead, we need to learn how to interpret it more wisely, seeing both what the numbers reveal and what they leave out. For me, the beautiful game has taught a simple truth: the real story often lies in the space between the statistics and the narrative they create. Whether it’s a football formation or a company’s performance, that’s a lesson worth carrying forward.

The is an opinion piece based on the author’s POV and does not necessarily reflect the views of HackerNoon.

Every

Analytics vs. Instinct in Football

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

ChatGPT Is Not Just a Tool; It's a Mirror for How You Think

10 Ways to Optimize Your Database

10 Best React Native Chart Libraries

10 Best Datasets for Time Series Analysis

12 Mistakes that Data Scientists Make and How to Avoid Them

13 Best Datasets for Power BI Practice

ChatGPT Is Not Just a Tool; It's a Mirror for How You Think

10 Ways to Optimize Your Database

10 Best React Native Chart Libraries

10 Best Datasets for Time Series Analysis

12 Mistakes that Data Scientists Make and How to Avoid Them

13 Best Datasets for Power BI Practice

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps