A 3-Cushion Qualifier With No Stat Sheet: Four Days of Hand-Notation and One Conclusion I Had to Withdraw

Inning number 34. The player in the blue shirt placed the cue ball near the middle of the table, needing a three-cushion shot to close the inning. He missed by half a ball. On my phone screen, the qualifier's online statistics page was still empty: no average column, no inning count column, no high-run column. The organisers did nothing wrong. At a regional qualifying round, scores are kept on a paper board with a marker, and nobody is obliged to feed data to a man sitting in the third row.

I opened my notebook and drew four columns. Four days, seven matches, forty-seven innings written in blue ink. By the end of the fourth day I had a very tidy conclusion, and then I had to withdraw it after re-reading the very lines I had just written.

Context: a sport rich in data at the top, poor in data at the bottom

Three-cushion is a discipline where the average is treated as the standard measure. At UMB World Cup events, every single inning is recorded; the results page carries inning counts, averages, high runs, even per-round averages for each player. In the PBA system for pool, the data goes further still: pot success rate, break success rate, the number of times a player leaves the table.

Drop down to a regional qualifier and that entire apparatus disappears. There is no electronic board logging each inning, nobody pressing an update button, and nobody accountable for whether those figures exist at all. The Vietnamese generation of Trần Quyết Chiến, Nguyễn Quốc Nguyện, Dương Quốc Hoàng, Bao Phương Vinh and Nguyễn Trần Thanh Tự has lifted Vietnamese three-cushion into Asia's leading group, yet the data layer behind them at domestic level remains as thin as a paper score sheet.

Based on my experience following matches at domestic qualifiers and World Cup events, I always begin with a checklist of conditions that need verifying: who measured the data, how they measured it, on which table, and what it is hiding. At a qualifier with no data at all, that checklist shrinks to a single line: the measurer is me, and I carry the error margin myself.

Playing conditions are not a minor detail. Worn cloth, high humidity, a room temperature that shifts between morning and evening, all of it makes the same shot produce two different outcomes. A player used to an air-conditioned practice table will lose points in a hall with open doors. No statistics column records that, yet it sits inside every inning. When the match table stops being a fortress, I learn to listen to the sound of balls colliding on the table next to mine.

Core: the four columns I drew myself, and what they actually said

I recorded four columns. Column one was the score of the opening inning, meaning break quality. Column two was the number of innings needed to reach ten points. Column three was the average the opponent scored immediately after a defensive inning. Column four was the highest run of the match.

Forty-seven innings produced a sample average of 0.91. That figure is meaningless standing alone, and I knew it before I wrote it down. What matters is the spread. Two players finished the tournament with an average around 1.02, but one had three scoreless innings and two innings worth seven points, while the other scored almost exactly one point every inning. On the official score sheet they look identical. On the rhythm of the match they are two entirely different players.

The break column is where I was surprised. The player who won the semi-final averaged 2.3 points on his opening inning; the loser averaged 0.8. A gap of 1.5 points on the first inning, multiplied across more than ten innings per match, is enough to decide a race to forty points. One bad break is a mistake. Three bad breaks in the same session is a signal. When I checked column four, the high run, the winner was ahead there too, but not by as much as I had assumed. The high run is only the visible part; the break column is the submerged part.

Column three, the opponent's average after a defensive inning, was the hardest to record and the one I trust most. The winner of the final held his opponent to 0.6 points per defensive exchange, while the loser conceded 1.4. In a sport where innings are a scarce resource, forcing an opponent back to the table with one and a half fewer points per inning is worth more than any beautiful shot.

For a reference point, I went back through my notes from recent World Cups. Semi-finalists at those events typically finish with averages around 1.4 to 1.6, roughly a third higher than the 0.91 recorded at this qualifier. That gap says nothing about innate talent. It says that at qualifying level the break and the ability to finish an inning are still loose, and every wasted inning is paid for in points.

I did not run a statistical test on this dataset. Forty-seven innings, seven matches, one qualifier: the sample does not permit me to state anything with certainty, and I learned that lesson too expensively to forget it. A few years ago I built a model on the expected-value index of a single match and lost everything. Since then, before any conclusion, I ask myself: if the table changed, if the session changed, would the result hold?

Contrarian: the empty sheet is not the biggest problem

The tidy conclusion I had to withdraw was this. After day three I wrote in my notebook that the youngest player in the draw lacked nerve in decisive innings. The evidence was three missed shots on the final inning of matches. It sounded reasonable, and it matched the story the crowd keeps telling: young Vietnamese players lack experience in decisive moments.

On day four I recorded the cue ball position before each of those misses. All three began from a passive position left by a poor break, not from a key shot in a good position he had created himself. What failed was the opening inning, not the nervous system. Had I looked only at the average column and the three late misses, I would have written a conclusion about psychology while the data was talking about technique.

That is why I do not trust judgements built on a single column of figures. An empty statistics sheet is honest in its own way. A statistics sheet with exactly one indicator is more dangerous, because it creates the impression of being fully equipped to pass judgement.

Data never lies, but I have misheard it before.

One more thing needs saying. Data analysts are moving ever closer to the dressing rooms of many sports, and billiards is not outside that trend. But a model cannot know what a player feels stepping into his thirty-fifth inning with the score level, on a table he has never played on. What I wrote in my notebook is only the measurable part. The crowd pressure here is the sound of balls clacking on the neighbouring table, and I have to state clearly that I have no numbers for that part.

Takeaway

At the end of the qualifier I kept only two lines. First, at qualifying level break quality matters more than the high run, and that is what I will record in the first column of every match from now on. Second, when there is no data, the most honest thing an observer can do is measure it himself, publish the sample size, and mark the gaps.

I do not write to convince anyone. I write so that the data has a witness.

Next round I will add one more column: the cue ball position at the start of the inning. If the next forty-seven innings repeat this same story, I will have grounds to speak a little louder. If not, I will rewrite the piece and make fun of myself for mishearing it again.