Three Billiards Disciplines, Three Data Languages — and the Analyst's Trap

On 6 June 2026, the World Professional Billiards and Snooker Association (WPBSA) announced sanctions against ten Chinese cueists after a months-long match-fixing investigation. Liang Wenbo and Li Hang received lifetime bans; the other eight were suspended for between a few years and nearly a decade. For someone like me, who makes a living reading billiards data in London, it was not merely an ethical scandal. It was a methodological shock.

It took me a few weeks to see why. The problem was not any number in the case file. It was the question I had never forced myself to answer clearly: when I say "billiards", which discipline am I talking about?

Billiards is not one sport. It is an umbrella containing at least four rule systems, and each system produces a kind of data that cannot be shared.

Snooker has a two-year rolling ranking system, the concept of the century break, the 147 maximum, and the Triple Crown of the World Championship at the Crucible Theatre in Sheffield, the UK Championship and the Masters. Nine-ball is measured by break-and-run, by race-to formats and by the Mosconi Cup — the annual USA-versus-Europe team event. Chinese eight-ball is a hybrid, played on a pocketed table with rules close to eight-ball, but its prize structure has swollen with domestic capital, through events such as the Joy Cup. European carom is an entirely different story.

These reference systems cannot be converted into one another. If you apply snooker's yardstick to nine-ball, you are not analysing badly — you are analysing a different sport. That is why the first step of any billiards analysis must be identifying the discipline, before the data table is even opened.

Back to the June 2026 case. The lesson for a data worker is not the identity of the punished cueists, but how the investigation was built. The WPBSA did not conclude from a single match. It cross-referenced samples across many seasons, compared anomalous betting patterns with actual patterns of play at the table, and only then reached a conclusion. That process has four steps: form a hypothesis, extract data, cross-check, then judge. Skipping any step is manufacturing your own error.

I once ran a small exercise to test my own professional reflexes. I took an event I was not sure belonged to which discipline, with only a name and a few lines of description, and tried to analyse it. The result: I could not identify the discipline at all. No distinctive event name, no rule terminology, no description of table or balls, no player names. When the discipline is unidentified, every metric downstream is meaningless. You cannot discuss the "quality of the break" in a snooker event, nor speak of a "147" in a nine-ball event. Each discipline has its own measuring toolkit, and using the wrong toolkit does not give you an approximate answer — it gives a completely wrong one.

The trophy is not on the scoreboard, it is in the table of chance quality. And that table is designed separately for each discipline, in its own system of units.

The same applies to rankings. Snooker uses two-year accumulated points, so a cueist can slide down the ranking despite deep runs in several events. Matchroom's nine-ball revolves around invitational events and per-event ranking points. Chinese eight-ball has its own system, tied to domestic events. The three systems cannot be converted. A cueist who wins a Chinese eight-ball event is not thereby ranked alongside a Crucible champion — even though both may be lumped together as "billiards world champion" in the press.

And this is where the data becomes dangerous. Journalism tends to collapse everything into one word: "billiards". A headline like "cueist X is world champion" can be true for one discipline and false for another. The reader has no way to check if the article does not state the reference system.

Based on my experience watching matches, I learned this most clearly in 2026, when arenas were empty of spectators because of the pandemic. When the crowd noise disappeared, variables normally drowned out suddenly surfaced. In billiards, the empty arena revealed the rhythm of reading the table, the time taken before a shot, the hesitation on deciding frames. "The empty arena, the coach's voice clearer than ever, and the data too." Only, an empty arena is a clean laboratory only if you know which discipline you are measuring in it.

Prize structure also says a great deal. Major snooker events concentrate prize money heavily on the winner, making the ranking race brutal at the top. Many Chinese eight-ball events have larger and more widely distributed prize funds, enough to draw cueists from other disciplines across. The flow of cueists between disciplines is a market signal, but it can only be read if we separate each system. Merged together, we see only a jumble of names and prizes that cannot be compared.

There is a temptation greater than misidentifying the discipline: concluding from too small a sample.

In football, after the 2026 World Cup, I wrote that Morocco's four knockout matches had an average xGA of 0.6 — the lowest of the tournament — but that a sample of four matches was far too small to call it a sustainable tactic. The same test applies to billiards: a cueist winning three events in a row proves nothing about the long term if we do not know whether those three events share a discipline, a ranking system, or a field of opponents. A winning streak can be a signal. It can also be noise arranged attractively.

A Morocco-style miracle in billiards — an unfancied cueist reaching the deep end of an event — is the same: it is measurable, but only within that sample. The miracle is not in the magic, it is in the square metres defended with intent — and in billiards, in the safety shots placed exactly right.

Worse, reputation is not data either. The 2026 case showed that some famous cueists were at the centre of a fixing network. A big name is not a reliable variable in the equation. Correlation is only correlation; to reach causation you must isolate variables, and billiards gives us very few chances to do that cleanly.

The next cycle will not reward the fastest reader, but the one who classifies correctly first. The question I am setting for myself next season is not "which cueist is playing best", but "which discipline am I talking about, under which reference system, and what am I willing to assert before the sample is large enough". The three billiards disciplines will keep sharing one name in the press. The data writer's job is not to let them share one number.

Data limits: This article rests on public facts about tournament structure and the 2026 case. Claims about cueist flows between disciplines are inferences from public data, not causal conclusions. Sample sizes for most examples are small; readers should treat this as an analytical framework, not a forecast.