When the Data Board Goes Blank at a Vietnamese Billiards Event: A Lesson on the Integrity of Numbers

On the last Saturday of October, at a billiards club in Hai Phong, the qualifying round of an open 3-cushion tournament reached its final session of the third match day. Table four. Tran Quyet Chien led Nguyen Quoc Nguyen 28-19 after the seventh inning. The electronic screen hanging beside the table — the one that should have displayed successful cushion-entry rate, positional play counts, and average rest time between turns at the cue — showed nothing but a blank space. Not a single number. Not a single average line. The organizers announced that the data recording system had failed starting from the previous round and would be fixed at the earliest opportunity. I sat in the third row, my notebook already open, my pen resting at an angle on the squared page. After four years in the billiards betting analysis trade, it was the first time I felt I was practicing my profession without raw material. The people around me still followed the match, still applauded every beautiful three-cushion shot, still gasped at every cunning positional play. Only I noticed something troubling: when data disappears from the screen, people keep telling stories — through feeling, through memory, through what they call intuition. And that story will appear in the papers tomorrow as if it were a verified truth. That night I decided to sit down and write this piece. Not about a shot, but about the foundation beneath every shot. Billiards is a sport of numbers, though very few people admit it. In snooker, a player's break strength is measured by points per visit, defensive ability by safety success rate, and consistency by the number of 50-plus and 100-plus breaks. In 3-cushion — the discipline where Vietnam genuinely holds a place on the world map — players are measured by average, by best run, by the number of innings needed to finish a frame. Every shot leaves a numeric trace. The problem is this: who records those numeric traces, how they are recorded, and under what conditions. Worldwide, billiards data systems have long been standardized. For snooker, the World Professional Billiards and Snooker Association and World Snooker Tour publish official statistics for each event; aggregator platforms preserve head-to-head records, century counts, and 147 maximums. For carom, the Union Mondiale de Billard and specialized platforms provide frame-by-frame and even shot-by-shot scoring, sometimes including shot speed. In Vietnam, national events and several open tournaments have begun using electronic recording systems, but completeness still varies between events, between tables, and between playing sessions. That disparity creates what I call an input gap. A data table can look complete while actually missing an entire column. A numeric column can look precise while actually being hand-entered. A result can look reasonable while actually being inferred from a single match. And the most dangerous thing in the analysis trade is failing to realize you are standing on such a gap. People usually think bad data means numbers higher or lower than reality. In my experience, the most dangerous bad data is data that looks perfectly fine. Years ago, when I was new to the trade, I built a match assessment on a statistics table I believed was complete. That table had everything — pot rate, safety count, number of 50-plus visits. But when I checked it against match video, I discovered the safety count column had been tallied using a definition different from the one I was using. What the data-entry operator called safe included shots where the opponent still had a finishing opportunity. The table was not wrong in its data entry. It was wrong in its definition. The first lesson I drew from it: before using a number, ask who measured it, how they measured it, under what conditions, and what it is hiding. Data never lies, but I have misheard it before. I wrote that line in my notebook that year and have never crossed it out. Back to the Hai Phong night. When the screen went blank, I started recording by hand. Each visit by Tran Quyet Chien, I marked a stroke. Each successful positional play, I added a cross. Each time he forced his opponent into a safety option, I circled it. After forty visits, my page looked like a battle map. And something interesting emerged: what I recorded by hand did not entirely match what I perceived with my eyes. My feeling said Tran Quyet Chien was controlling the match through attack. But the strokes on paper showed that the number of visits he ended with a safety shot exceeded the number he ended with points. In other words, he was winning by keeping his opponent silent, not by scoring in bulk. Looking only at the 28-19 score, one would assume this was an attacking display. But the handwritten data told another story: this was a defensive match disguised by the score. This is precisely what I mean by the input gap. When the official system stops working, each person has two choices. The first is to ignore it and keep telling the story through feeling. The second is to create a temporary system of one's own, and disclose openly that it is temporary. I chose the second, and wrote clearly at the top of the page: handwritten data, unverified, small sample size. Disclosing limitations matters this much because in sports analysis, overconfidence is a greater enemy than ignorance. Someone who knows they lack data will be cautious. Someone who believes they have complete data will reach a wrong conclusion and then defend it to the end. I have seen an entire analysis group argue for hundreds of comments over a single column of numbers that was misunderstood in definition. In 2026, when I was only seventeen, I first applied an expected-goals model to Vietnamese football and was taught a memorable lesson by one match. I predicted a strong team would win comfortably based on the chances it created. The result was the exact opposite, because my model did not factor in the form of the opposing goalkeeper — something no single metric captures. I realized that the most beautiful metrics often miss the exact moment that decides a match. From then on, I began recording twenty consecutive matches by hand to compare the numbers with what my eyes saw. In 2026, I wrote an analysis of one national team's high pressing and was mocked for a conclusion that ran against the crowd. Two weeks later, the results on the pitch confirmed my analysis, and my inbox received twelve acknowledgments. The crowd laughed. The numbers did not. But I also learned that being right once does not mean being right forever. A year later, I had to sit down, reopen that old piece, and correct myself wherever I had been overly optimistic. In 2026, mid-pandemic, when tournaments returned without spectators, I collected data from entire final rounds and found home advantage had declined markedly. I proposed an adjustment coefficient, disclosed the sample size and statistical significance, and noted my limitations clearly. When home is no longer a fortress, I learned to listen to empty stands. Some objected that the sample was too small to conclude. They were partly right. But staying silent and waiting is also a way of concluding wrongly. Those three stories, in the end, are variations of the same problem: when data is incomplete, what should we do? Back to billiards, where I truly live with the trade. The question is not only how to get enough data, but how to know whether the data you have belongs to whom and to which phase. An average of 1.5 in qualifying says nothing about a player's ability in the semifinal, where pressure and tempo are entirely different. An impressive pot rate in one match may simply reflect an opponent leaving easy positions. One missed cushion shot is an error. Three missed cushion shots at the same tournament is a signal. But to read that signal, you need to know whether those three misses happened on the same table type, in the same climate conditions, and in the same mental state. That is why I never use a single metric to conclude. Every analysis I write begins with a checklist of conditions to verify. Who recorded this data? Where? What is the sample size? Is there a second source to cross-check? If the answer to any item is unclear, I lower the confidence of the entire conclusion. With Vietnamese billiards, where even a national event lacks a uniform data system across tables, that checklist is often longer than the conclusion itself. I have also learned to distinguish between two kinds of gaps. The first is a technical gap — a system fails, a screen goes blank, a recorder is absent. The second is a conceptual gap — a system runs fine but measures the wrong thing. The second is far more dangerous, because it produces no blank screen to warn you. It simply, quietly, produces numbers that look entirely reasonable. There is a common belief in the analysis world that when data is missing, the solution is to collect more. I do not oppose that. But I believe that before collecting more, you must clarify what exactly you are missing. If you are missing definitions, collecting another thousand data points only produces a thousand wrong points. This holds for billiards no less than for football or basketball. That Hai Phong night, I left the club with two things in hand: a final score still unfinished on the electronic board, and a page of unverified handwritten data. The next morning, a few news sites ran the result with commentary like a devastating attacking display. Reading those lines, I understood that most spectators would never know the screen had been blank throughout the match. They would believe a story written through feeling, and that story would become collective memory. There is an objection I always raise with myself: if data is missing, why not wait until there is enough and then write? My answer after many years is: waiting is also a decision, and a bad one if it leads readers to believe the gap does not exist. But here I want to push the argument one step further. The emptiness of data is not only an obstacle — sometimes it is itself data. A recording system collapsing at a tournament says something about that tournament: about its level of investment, its operational capacity, and how the organizers treat the value of information. When a column of numbers disappears, the right question is not what we are missing, but why we allowed it to disappear. This is where I frequently disagree with my colleagues in the data analysis field. Many of them enter the locker room of sport with absolute faith that everything can be measured, and everything measured can be optimized. But billiards, like football or basketball, is not merely an optimization problem. It is a sequence of moments decided by a trembling hand, by a short breath, by noise from the row behind you. No model measures the heart rate of a player standing before the decisive shot. I do not write to persuade anyone. I write so that data has a witness. What I fear is not missing data. What I fear is a world where data becomes so complete that people stop asking how it was created. A perfect scoreboard can make us forget that behind every number is a person who sat typing at two in the morning, tired, and may have typed a single digit wrong. In the billiards industry chain, this deserves even more attention. Upstream, practice halls and clubs are where young players are produced, but also where data is lost most — a practice frame no one records, a grassroots tournament no one aggregates. Midstream, professional players and events are where data is generated but often only serves live media, not archived for later analysis. Downstream, sponsors, equipment, and derivative markets are the places that need data most but reach it least. If I had to pick one thing to improve the situation, I would not buy more screens. I would choose to unify definitions. A tournament does not need ten metrics. It needs three metrics defined clearly, recorded consistently from qualifying to the final, and disclosed to anyone who wants to verify. That is the foundation on which everything else can be built. The Hai Phong night ended with a score few will remember tomorrow, and a lesson I will carry through my career. The next round of the season begins in three weeks. In those three weeks, the organizers have a chance to fix the recording system, while I have the task of reopening my notebook and pushing every handwritten number into a small box clearly marked small sample, unverified. When the data board goes blank, the only thing left is the name of this trade: honesty about what we know and what we do not. I do not write to persuade anyone. I write so that data has a witness. And sometimes, the witness must be the pen itself.