Vietnamese Basketball and the Blank Cells in the Box Score

One morning in March, I opened the scouting file of a domestic basketball team and found almost every stat column returning blank. Not a margin of error, not a unit mismatch. Blank. Effective field goal percentage, or eFG%, was empty. Assists per 100 possessions were empty. Even the defensive rating after a screen, the number I needed most to judge a defense, was empty. I stared at the screen longer than usual, because a wrong metric can still be fixed, while an empty cell leaves nothing to fix.

My job is to read data and say what will happen next. Data only answers when it exists.

Across roughly thirteen years observing professional sport, I learned something counter to most people's intuition: the greatest fear of a data analyst is not a wrong metric but a missing one. When a stat sheet returns empty values, everything behind it loses its footing. You cannot compute a standard deviation on a blank column.

In Vietnamese basketball, most teams still track only the basics: points, rebounds, assists, and occasionally shooting percentage. Those numbers are correct, but they describe only the surface of a game. They do not show how many points a guard creates per possession, whether a center defends better or worse when pulled beyond the three-point line, or how many seconds a pressing team needs to shift from attack to defense.

The problem is not laziness. The problem is that the data infrastructure has not been built. Motion-tracking cameras, systems that record player positions every fraction of a second, synchronization software: all cost money and people, and only pay off after several seasons. When budgets are tight, that is the first line item cut. And when it is cut, nobody sees the loss, because what disappears never existed in the first place.

I once built a nine-dimension analysis framework for a team: tactics, player data, team operations, league landscape, rules and governance, coaching staff and locker room, risk, media, and industry ripple effects. It sounded impressive. But when real data went in, eight of the nine dimensions returned empty values.

The tactics dimension: no OffRtg, no DefRtg, no Pace. That does not mean the team has no tactics. It means their tactics were never recorded as metrics, so they cannot be compared with anyone. A disciplined system and a chaotic one look identical when viewed through an empty stat sheet.

The player dimension: no advanced metrics, no USG%, no age curve. A rising player and a declining one can look the same across three straight games if you only read scoring average. The difference lies in efficiency per shot attempt, which nobody measures.

The operations dimension: no salary structure, no reserve fund, no tradeable assets. That means every buy-or-sell decision rests on the gut feeling of whoever sits in the meeting room.

This is the point I want to make clear. Blank space is not neutral. It does not say 'unknown.' It says 'fill me with something.' And people always fill it with whatever is closest at hand: the memory of a great game, a beautiful play, a compliment on social media. Blank data is always filled with narrative, and narrative has no error bar. When there are no metrics, people do not become more uncertain; they become more confident, because nothing contradicts them.

In 2026, I pointed out that striker Gastón Merlo of SHB Đà Nẵng averaged 0.8 xG per match but scored only 0.4 goals. I published the raw data of the next 12 matches, with shot counts and shot locations. The team took 9 of 36 points, exactly as predicted. In 2026, I analyzed the German national team and predicted they would exit in the group stage. People argued against me not with data but with past reputation. In 2026, the whole world mourned Germany. I quietly re-read the model's log file. The data already knew, but the data was not read.

Most people believe the danger of sports analytics is bad data. I believe the real danger is missing data, and it is more dangerous precisely because it is invisible.

A wrong metric produces a wrong prediction, and a wrong prediction gets corrected by the match. That is a healthy feedback loop. A missing metric produces no feedback at all. It sits in silence, and people fill it with intuition, with experience, with a feel for the game. Every coach talks about feel. I have no feel; I have standard deviation. But standard deviation needs data, while feel needs nothing.

The paradox is this: the less data a team has, the more it trusts feel, and the more it trusts feel, the less it invests in data. That is a self-locking loop. Insiders do not escape it by working harder, but by accepting that what they lack is not effort but infrastructure.

I do not want to be understood as dismissing people. A coach who reads a game well by eye can be right more often than a bad model. But once a model has enough data, the gap closes very fast. The problem is not that the human eye is worse than a machine. The problem is that the human eye cannot store a million possessions.

Vietnamese basketball is at a stage where the right question is not which software to buy, but what to record first. I believe the signal of the coming season will lie with teams that start recording shot locations instead of only points. Data is a monastery: the less noise, the more clearly you hear something trying to speak. Our current problem is that the monastery is still too silent, and we are mistaking the sound of ourselves for the sound of the game.