The Irrationality of Football: When Data Fails Against Shocks

In six years of scrutinizing data and football pitches, I have never ceased to marvel at a paradox: numbers do not lie, but people always find ways to deceive numbers. Every season, we witness matches where all predictive models collapse, results that no algorithm can fully explain. I call this the irrational part of football – the thing that xG never measures, that tracking data cannot record, and that only those who have stood on the pitch, feeling the pressure of the stands, truly understand. There is a pressure that no one sees, but every team fears. I named it: Binh Duong pressing. It is a pressure that does not appear on the statistics sheet, cannot be shown through PPDA or tackle counts, but it seeps into the opponent's psychology before the match even begins. In 2026, when I analyzed all 26 rounds of the V-League, I discovered that Becamex Binh Duong had an average PPDA of 8.4 – the lowest in the league. That number means they allowed opponents an average of 8.4 passes before closing down. But what the data cannot show is the fear that this style creates. Binh Duong's opponents were not just pressed on the pitch; they were pressed in their minds. They made mistakes even when not being pressed, because the fear had already seeped into their subconscious. xG is not wrong, football is just irrational. After 2026, I learned to count the irrationality too. That year, I built a prediction model based on 180,000 shots from 5 European leagues. My model correctly predicted 14/16 knockout matches at the World Cup in Russia. But Croatia, the team I analyzed as having the lowest xG among the semi-finalists, reached the final. They did not create many chances, but they had 23 accelerations over 25 km/h per match – a number far superior to any opponent. When I wrote about this, many fans criticized me as dry and mechanical. They did not understand that I was trying to map the irrational part, not deny it. After the 2026 World Cup, I became accustomed to writing based on two data-driven scenarios rather than one-sided assertions. I realized that prediction models are not scripture; they are just a compass – but without them, we are lost. And I began applying this philosophy to every analysis, from football to swimming, from pitches to pools. When the stadium is empty, every model collapses. I rebuild from the scorched data. In 2026, the pandemic halted football. When the Bundesliga returned with 312 matches without spectators, I treated it as a massive laboratory. I discovered that home advantage dropped from 54% to 47%, home teams' PPDA increased by 0.9, meaning away teams pressed higher without the pressure of the crowd. My article "Empty Stadium, Changed Dynamics" reached 180,000 reads and was consulted by a Premier League club. I realized that crisis is the opportunity to rebuild all old assumptions. Now, during the transfer window, I see football's irrationality most clearly. The transfer market is the only place where people pay for expectation, not reality. Every summer, we witness crazy deals: young players with fewer than 50 top-level matches valued at 100 million euros, contracts signed based on highlights rather than detailed data. I have witnessed too many clubs burning money on players my model ranked low, and missing gems that data pointed to. The problem is not that data is wrong. The problem is that people read data wrong. A player scoring 20 goals in a season in a weak league may never replicate that in a top league. But a player creating 15 chances per match without scoring could be a massive bargain. Data does not lie, but people always find ways to deceive data – by choosing numbers that fit the story they want to tell, rather than listening to what data actually says. There is a pressure that no one sees, but every team fears. I named it: Binh Duong pressing. In the transfer context, that pressure is expectation. When a club spends 50 million euros on a player, they are not just buying a person; they are buying the pressure to succeed. That player will be scrutinized every minute on the pitch, compared to their transfer value in every move. That is a pressure no model can measure, but it directly affects a player's performance. Reputation is just a name. What remains is how you read the match. When I analyze a match, I do not care about the player's name. I care about how many kilometers they run, how many times they accelerate, how many chances they create. Names are just noise; data is the signal. But I have also learned that there are things data cannot capture: fighting spirit, confidence, game-reading ability. That is the irrational part I have learned to accept. In esports, every millisecond is a decision. Data does not predict, data records. I have applied the same philosophy when analyzing esports matches: instead of trying to predict outcomes, I focus on recording what happened, finding recurring patterns, and using them to understand the game deeper. This approach helps me avoid the trap of imposing models onto reality. I once treated models as scripture. Now they are just a compass – but without them, we are lost. After years of working with data, I realize that data's true value lies not in its ability to predict accurately, but in its ability to ask the right questions. A good model does not necessarily predict everything correctly; it just needs to point out what we do not understand. That is why I never stop hunting for anomalies – the numbers that stand out, the moves no one notices. Because those anomalies are the key to understanding the game deeper. In this transfer window, I advise clubs to look at tracking data instead of highlights. A traditional winger may not score many goals, but if he creates space for teammates, if he stretches the opponent's defense, his value is not shown in goal numbers. I have seen too many clubs miss such players only to spend tens of millions on inverted wingers – a trend that is homogenizing football. Inverted wingers are homogenizing football; traditional wingers are being wrongly erased. When every team uses the same tactical formula, football becomes more predictable, and data becomes less valuable. But teams that dare to go against the trend, that dare to use traditional wingers to create difference, are the teams that data can help the most. Because that difference creates the anomalies I hunt for. The numbers are there, but you cannot read them. That is a phrase I often use when explaining to sporting directors why they should trust data. Not because data is always right, but because data helps them ask the right questions. When a player is valued at 80 million euros, the question is not "Is he worth it?" but "How will he fit into our system?" Data can answer the second question much better than the first. Binh Duong pressing meets its signature match today. When I analyze a match, I often look for moments when a team applies exactly its pressing philosophy. Those are the moments when data and reality merge, when everything happens exactly as the model predicted. But I have also learned that those moments rarely last the whole match. Football always has irrational plays, moments when everything turns upside down. xG 3.2 but lost 0-1. Welcome to football. I have witnessed too many such matches to be surprised. But I am still always curious about what happened in those matches. Why did a team create so many chances yet fail to score? Was the opposing goalkeeper too good? Or did the players lose composure in front of goal? Data can tell me what happened, but cannot explain why. That is the irrational part I have learned to accept. Empty stadium, empty home advantage. That was my most important discovery during the pandemic. Without spectators, home teams lose their psychological edge, and that shows clearly in the data. But I also realized that some teams never depend on the crowd, teams whose pressing philosophy does not need external reinforcement. Those are the teams I call having "Binh Duong pressing" – a pressure that comes from within, not from the stands. Buying players based on highlights is burning money. I cannot emphasize this enough. Highlights only show you a player's best moments, but not what he does in the other 90 minutes. Tracking data can show you the full picture: how much he runs, how many sprints he makes, how many chances he creates. That is information highlights can never provide. Do not ask me the score. Ask the data. That is a phrase I often use in interviews. I do not care about the final result; I care about the process leading to that result. I want to know how many chances the team created, how they controlled the match, how they pressed. All that information matters more than the final score. A good model needs 5 years, not 5 matches. That is a lesson I learned from my own mistakes. When I first started using data, I was often hasty in drawing conclusions from small samples. But after years, I realized that data needs time to reveal true patterns. A player can play badly for 5 matches but be a great signing. A team can win 5 in a row but be on the verge of collapse. Only time reveals the truth. Esports is not a game, it is a market. Behave accordingly. When I expanded my analysis to esports, I realized the fundamental principles remained the same. Data does not predict, data records. But in esports, everything happens faster, and decisions are made in milliseconds. That requires a different approach, but the core philosophy remains: hunt for anomalies, verify three sources, and dare to publish early predictions. When the stadium is empty, every model collapses. I rebuild from the scorched data. That is my story in 2026, and also football's story during the pandemic. But I believe that from the ruins, we can build something better. We have learned that home advantage is not a given, that spectators have immeasurable value, and that data can help us understand better what truly matters in football. In this transfer window, remember: numbers do not lie, but people always find ways to deceive numbers. Look at tracking data, analyze contract structures, follow agent movements. And always remember that football is irrational – the only thing you can do is prepare for that irrationality. The transfer market is the only place where people pay for expectation, not reality. That is a phrase I repeat to sporting directors. When you spend 100 million euros on a 20-year-old player, you are not paying for what he has done; you are paying for what he could do. And that is a gamble. Data can help you reduce risk, but never eliminate it entirely. Because football is irrational. I have spent 25 years observing the sports industry, from swimming to football, from pitches to pools. And I have learned that the most important thing is not data, not models, but the ability to ask the right questions. Data is just a tool, but how you use it makes the difference. Hunt for anomalies, verify three sources, and dare to publish early predictions. That is the only way to understand this game. There is a pressure that no one sees, but every team fears. I named it: Binh Duong pressing. And I believe that pressure will continue to shape football, no matter how far data develops. Because there are things data can never measure – and that is what makes football so fascinating.