When Data Lied: AC Milan, San Siro and the 0.2-Second Sensor

I remember one evening in late April 2026 in Milan. Head coach Vincenzo Montella put in front of me a table of expected goals from AC Milan's first 20 Serie A matches. At home, at San Siro: 1.85. Away from home: 1.02. Almost double. Yet the team's actual goals scored in those two settings were level. If you only read the numbers, you would say the problem was finishing, luck, or the strikers' form. I did not think so. In a closed meeting room, I asked an uncomfortable question: what if the problem was not the players, but the measuring equipment? At that time, AC Milan had no margin for error. They were fighting for a Europa League spot against direct rivals such as Inter, Atalanta and Lazio. One wrong tactical step could cost them a European place. Our analysis department was asked to audit movement data from 20 matches. The data came from cameras and sensors placed around the pitch. Every build-up, every movement, every press was turned into coordinates. But when I compared the data with the broadcast images, I noticed something strange: whenever the ball was played out from the goalkeeper through the south-west corner of San Siro, the positioning data of the receiving player was shifted by one beat compared to the image. Not by much — just 0.2 seconds — but enough to distort the whole story. 0.2 seconds sounds harmless. It is not enough to make a pass look as slow as a defensive error, but it is enough for a sensor to record a pass later than it actually happened. When you rebuild pressing, space and attacking direction from that data, every phase that started from the south-west corner became deformed. Why did this matter? Because most of Milan's attacking sequences started from the goalkeeper. They built their game from the back, and the mispositioned sensor made the system believe that the team could not circulate the ball quickly through that zone. If we had trusted the old numbers, we might have concluded that Milan needed to avoid building down the left side, or we would have blamed the deep-lying midfielder. We would have changed a working tactical design because of a wrong number. My internal report was 14 pages long, and the most important conclusion was not about tactics: calibrate the sensors before making decisions. The coaching staff did not rush to change the starting eleven. Montella chose to increase ball circulation down the right flank only after the new data was clean. The result: AC Milan won five of their last eight matches and secured a Europa League place. Not because of a new formation, not because of a star, but because they were willing to check the origin of the number before using it to steer the team. This may sound obvious, but in modern football, few clubs are willing to do it. Most coaching staffs ask 'how should we play?' before asking 'is the data we are looking at correct?'. They love a beautiful xG number because it confirms their philosophy. They call pressing, tracking and expected goals something that cannot be questioned. But data only tells part of the story; the rest depends on knowing how to listen. I sat in meetings where engineers talked about camera latency, but nobody wanted to hear it because they wanted to talk about tactics. That is why a 0.2-second error could survive for half a season. It does not appear on screen in red letters; it hides inside a metric that looks perfect. Every collapse has a precursor; only a few people are willing to see it in advance. When Montella left AC Milan, people usually told a story about unstable tactics and dressing-room pressure. But if you look more carefully, what began to crack earlier may have been a blind faith in a data system. We managed to fix it in 2026-17, but not every club is so lucky. Now, whenever I look at a statistics table on a screen, I remember the question: which sensor produced this number, and was the measurement condition clean? Every tracking number needs to be put on the operating table, not on an altar. If AC Milan had not stopped to check the equipment back then, they might have adjusted their attack in the wrong direction and lost Europa League qualification. But they stopped, just for a beat, and what they found was not a new star. It was an old truth: before believing in what you measure, make sure the ruler is not bent.