The "Billiards" Label and Six Hidden Sports: Decoding the Discipline Identification Gate

The "Billiards" Label and Six Hidden Sports: Decoding the Discipline Identification Gate

Hook

In March 2026, I sat in front of a screen in a small flat in east London and opened a dataset whose only filename was billiards_2025_2026.csv. It held four thousand rows recording matches played over eighteen months, with exactly one label column at the top of the sheet: "billiards." No tournament name, no player name, no rule set, no unit of measurement. Just one word. I sat still for about four minutes, hands on the keyboard, and realised I could not write a single line of analysis. Not because the data was missing, but because the data was answering a different question from the one I had asked. I had asked: who is playing well. The dataset could only answer one thing: it did not know which sport it was describing. That was the moment I understood that the discipline identification gate is not an administrative step in a workflow; it is the foundation of every number that comes after it. Skip it, and every conclusion below is just a chain of inference dressed up in table formatting.

Context

In ten years of covering billiards as a data journalist, I have learned something econometrics lecturers do not teach: before you count, you must identify. A dataset marked "billiards" sounds neutral, but it is actually a family-level label, not a discipline-level one. The billiards family contains at least six disciplines with entirely different rule sets, technical systems, tournament structures and commercial ecosystems: snooker, American 9-ball, Chinese 8-ball (heyball), American 8-ball, three-cushion carom, and Russian pyramid. These six do not share the same definition of a "good shot." A snooker player is judged by break-building and cue-ball control in tight space. A 9-ball player is judged by break quality and by handling the object ball in rotation. A heyball player is judged by accuracy into tighter pockets on different cloth. A carom player has no pockets at all — only three balls and four cushions. Merging these six into a single label column is like merging football, American football and Australian rules football into a column called "ball." Semantically it is not wrong, but analytically it is useless.

This confusion is not a beginner's mistake. It is a consequence of how the industry has named itself for over a century. In Britain, "billiards" was once the official name of a single game — English billiards, played on a pocketless table with two white balls and one red — before snooker overtook it in popularity from the early twentieth century. By the 1970s, when snooker exploded on colour television with its world championship at the Crucible Theatre in Sheffield, "billiards" gradually became an umbrella term of purely historical weight. Today, when an international newsroom tags a dataset "billiards," it is not saying it is analysing English billiards. It is only saying it has not yet decided.

I remember the summer of 2026, when I was a first-year economics student in London and launched a World Cup data blog. My first piece was on Germany's 0-2 defeat to South Korea, where the defending champions generated 2.1 xG, held 74% possession, and scored nothing. I showed that Germany's shots all came from wide positions, averaging only 0.08 xG per attempt. The piece drew around five hundred reads, and my econometrics lecturer made a remark I have carried through my entire career: "Data does not lie, but it is speaking in a language you do not yet fully understand." That remark applies exactly to the "billiards" dataset. The data is not wrong. The person who tagged it created the problem.

What worries me more is the frequency. Over the past two years I have received at least eleven datasets from different sources — sports data vendors, club analytics departments, even a tournament organiser — and eight of them used "billiards" as a top-level classification tag. None of them had a sub-field to separate disciplines. This is no longer the personal problem of a fussy analyst. It is a structural gap across the entire billiards data industry.

Core

To show that this problem is more serious than it looks, I will walk through each discipline, and for each one I will identify a metric that, if applied to another discipline, produces a completely false conclusion. This is the original-data section of the piece, taking up most of its length, and I will not reduce it to a list. Each discipline will be dissected as an independent ecosystem, because that is exactly what they are.

Snooker — the discipline of breaks and spatial control

Snooker has the most mature data system in the billiards family, and is also the discipline most often confused with "billiards" in general. A professional snooker match at ranking level is played as a race to a number of frames, from seven frames in early rounds of minor events to thirty-five frames in the World Championship final. Players score by potting object balls in ascending value order — red, yellow, green, brown, blue, pink, black — and the score of a break is the sum of the values of valid balls potted consecutively in one visit. A break of one hundred or more is called a "century break," and this is the most misunderstood metric when people merge the "billiards" label.

In the 2026-2026 season, Ronnie O'Sullivan passed one thousand three hundred career century breaks, a figure with no equivalent in any other billiards discipline, because no other discipline has a concept of a break accumulating by ball value. If you apply this metric to 9-ball, it becomes meaningless: in 9-ball there is no ascending value order, every ball is worth one point, and the concept of a "break" lasts at most nine balls. If you apply it to heyball, it is also meaningless, because heyball is played as an 8-ball contest with group disputes, with no fixed value sequence.

But snooker's deeper metric is not scoring; it is the win rate in frames with extended safety battles. In my analysis of Mark Selby's last three seasons — the four-time world champion — I found that his frame win rate does not correlate strongly with century breaks, but correlates more strongly with the rate at which he forces opponents into errors during safety exchanges. Selby is not the highest scorer. He is the one who makes opponents play shots they do not want to play. This is a signal that can only be read if you know you are analysing snooker, where the concept of "putting the cue ball in a difficult position" has its own technical structure.

Snooker's tournament structure is also unique. Three events are called the Triple Crown — the World Championship at the Crucible, the UK Championship, and the Masters — carrying a psychological and historical weight far beyond ordinary ranking events. A player who wins all three in one season is said to have completed the Triple Crown, and only three men have done so in the modern era: Steve Davis, Stephen Hendry, and Mark Williams. This metric cannot be transferred to any other discipline, because no other discipline has a three-event system defining the whole peak of a career in the same way.

Snooker's commercial ecosystem also carries its own mark. Revenue comes mainly from television contracts in the British and European markets, plus the Chinese market — where snooker has a huge audience thanks to players like Ding Junhui, a three-time UK Champion and one-time Masters champion. Ding opened the door for an entire generation of Chinese players, including Yan Bingtao — 2026 Masters champion — and Zhao Xintong — 2026 UK Champion. However, in 2026 a match-fixing scandal involving ten Chinese players shook this ecosystem, leading to long-term suspensions. That event revealed an important feature of snooker: its betting-market liquidity is high enough that violations have clear economic motives, unlike disciplines with less betting activity.

American 9-ball — the discipline of the break and rotation order

If snooker is the discipline of breaks, 9-ball is the discipline of the break shot. In 9-ball, only nine balls numbered one to nine are used. A player must strike the lowest-numbered ball on the table first with the cue ball, but the ball potted need not be the lowest-numbered one — as long as the lowest is struck first. Whoever pots the nine ball wins the rack. This creates a tactical structure snooker does not have: a player can "go around" by playing the lowest ball in a way that sends the nine ball in indirectly.

The central metric of 9-ball is break quality. In a professional match, a good break — meaning at least one ball potted with the cue ball left in a favourable position — determines who controls the table. Shane Van Boening, the American player regarded as one of the most powerful breakers in the game's history, has sustained a break success rate above 60% across many seasons. Joshua Filler, the German player, is known for a highly accurate break. Here, if you take snooker's "century break" concept and apply it, you will be looking for something that does not exist.

9-ball's tournament structure is also different. The main professional system today is run by Matchroom, the organisation behind the World 9-Ball Championship and the Mosconi Cup — a Europe versus USA team event staged in a Ryder Cup-like format. The Mosconi Cup is not an individual ranking event but a high-showmanship team event, and the metric there is team spirit under pressure, something that can barely be measured in pure numbers. This is the point where even I, a believer in data, must concede: some metrics exist only within a specific competitive context, and merging the "billiards" label erases that context entirely.

Commercially, 9-ball has strong markets in the US and Asia, but prize values are considerably lower than snooker at the top level. A leading 9-ball player can live on prize money and sponsorship, but margins are narrower than a top-16 snooker player's. This is why many 9-ball players must compete across parallel systems, and also why their data is often fragmented across multiple organisers — a problem I will return to later.

Chinese 8-ball (heyball) — the discipline of tight pockets and a billion-dollar market

Heyball is the fastest-growing discipline in the billiards family over the past decade, and also the most misunderstood by Western analysts. In form, heyball looks like American 8-ball: players dispute groups (one to seven and nine to fifteen), and whoever pots the eight ball after clearing their group wins. But there is a decisive technical difference: pockets on a heyball table are considerably tighter than on a standard American 8-ball table, and the cloth has different friction. This turns heyball into a discipline demanding markedly higher accuracy, where shots considered "hard but doable" in American 8-ball become "almost impossible" in heyball.

Heyball's signature metric is the success rate of long-range finishing on a tight-pocket table. The player who dominates this discipline is Zheng Yubo, regarded as world number one for several years. He is famous not for a powerful break but for cue-ball control in extremely tight space. If you take Zheng Yubo's data and compare it directly with Shane Van Boening's without separating disciplines, Zheng will look "weaker" on the break metric, when in fact he is playing a discipline where the break is not the decisive weapon in the same way.

What makes heyball a special case in data analysis is its market growth rate. Heyball events in China, especially the JOY-run tour, have total prize funds exceeding many ordinary ranking snooker events. Some top-level heyball events have total prize funds reaching millions of renminbi, and this is why some young snooker players have considered switching to heyball — a phenomenon I call "cross-discipline talent flow." This phenomenon can only be analysed if you separate the two disciplines. If you merge them under the "billiards" label, you will see a chaotic labour market with no logic, when in fact there is a very clear logic: talent follows the money.

This is where one of my signature lines becomes directly relevant: the transfer market is essentially a regression model, but people keep calling it a race. In billiards, player flow between disciplines is a dependent variable, and the independent variables include total prize money, travel costs, broadcast exposure, and playing time. Merging the label makes this variable invisible.

American 8-ball — the discipline of call-shot rules

American 8-ball is the most popular discipline in bars worldwide, but at professional level it has a far stricter rule system than amateur players usually know. At professional level, many tours apply "call shot" rules: a player must declare in advance which ball will go into which pocket, and if a ball drops into a different pocket than declared, the shot is a foul. This creates a tactical structure requiring players to weigh the probability of success against the risk of penalty.

The most important metric in professional 8-ball is the success rate of call shots in situations with multiple options. This is a far more complex metric than in 9-ball, because in 9-ball the ball order is predetermined, whereas in 8-ball a player has full freedom to choose among their group. In other words, 8-ball is a discipline of decision-making, while 9-ball is a discipline of execution in order. If you merge the two under one label, you cannot measure the decision-making variable — a variable any sports analyst would want.

Organisationally, professional 8-ball has a less centralised tour system than snooker and 9-ball. Many events are run independently by national federations or private promoters, leading to data fragmentation. This is why, when I received the billiards_2025_2026.csv dataset, I was not surprised to find it mixing 8-ball and 9-ball matches: the data compiler had struggled to classify, and instead of classifying, applied a blanket label.

Three-cushion carom — the discipline without pockets

Three-cushion carom is the most distinct discipline in the billiards family, to the point that many pocket-billiards players do not consider it the same game. In carom there are no pockets on the table. A player must send the cue ball into both other balls, provided the cue ball contacts at least three cushions before striking the second ball. A point is scored each time this condition is met. There is no concept of "potting a ball," no concept of a "value-ordered break," and no concept of a "break shot." The entire metric system of snooker and pocket billiards becomes meaningless when applied to carom.

Carom's central metric is average per innings. At the top level, players such as Dick Jaspers of the Netherlands, Torbjörn Blomdahl of Sweden, and Frédéric Caudron of Belgium sustain averages above two points per innings across many seasons, with individual peaks exceeding four points per innings in some matches. Raymond Ceulemans, the Belgian legend, dominated the game for decades with more than twenty world titles. These numbers cannot be compared with any metric in snooker or pocket billiards.

Commercially, carom has a strong market in continental Europe — especially the Netherlands, Belgium, Germany and Sweden — and in South Korea, where carom is one of the most popular sports with a dense hall system. The governing body is the Union Mondiale de Billard, an organisation whose tournament system is entirely separate from the snooker or 9-ball systems. If you merge the "billiards" label, you merge two governance systems, two tournament systems, and two entirely different viewer markets.

Russian pyramid — the discipline of ultra-tight pockets and heavy balls

Russian pyramid is the least known discipline outside its sphere of influence, but it is technically one of the hardest. The pyramid table has pockets so tight they barely admit the ball, and the balls are heavier than standard billiard balls. Players must pot with extreme accuracy, and the success rate at amateur level is so low that many matches drag on with very low scores.

Pyramid's signature metric is the success rate of finishing shots out of total shots, and this figure is typically far lower than in any other pocket-billiards discipline. If you take a leading pyramid player and compare his finishing rate with a leading 9-ball player without separating disciplines, you will wrongly conclude the pyramid player is weaker, when in fact he is playing a discipline of markedly higher technical difficulty.

Organisationally, pyramid has its own tournament system in Russia and neighbouring countries, with national federations and its own ranking system. This is the clearest example of how the "billiards" label not only causes analytical confusion but also obscures an entire sports ecosystem with millions of players.

Synthesis: six disciplines, six metric systems, one name

At this point I can draw a verifiable conclusion: across the six disciplines above, no single metric is shareable. Century breaks only mean something in snooker. Break quality only means something in 9-ball. Call-shot rate only means something in 8-ball. Average per innings only means something in carom. Finishing rate on tight-pocket tables only means something in heyball and pyramid — but with different reference thresholds. Merging six disciplines into one label does not lose information at row level, but it loses all information at column level. And in data analysis, losing information at column level means every model behind it is meaningless.

This is where another of my signature lines comes in: a team's journey is not an upward arrow, but a scatter plot. In billiards, a player's journey is the same, but that scatter plot can only be read if you know which axis you are drawing on. Snooker's axis is break-building and safety control. Carom's axis is average per innings. Mix the axes together and you get a shapeless cloud of points, and you will call it "the randomness of sport." But that is not randomness. That is measurement error.

Contrarian

Here I must argue against myself, because the claim "you must separate disciplines" sounds so obvious as to be suspicious. There are two alternative explanations for datasets using the "billiards" label that I must consider before concluding this is a structural error.

First explanation: merging the label is a deliberate choice, not an oversight. Some data vendors merge disciplines because their clients — betting companies, media platforms — care only about win/loss outcomes and odds, not technical metrics. To them, a snooker match and a 9-ball match are both "an event with two outcomes," and detailed classification only raises processing costs. If this is common, the problem is not data quality but the fact that I am using data for a purpose other than its design purpose. This is a real possibility, and I cannot dismiss it by argument alone.

Second explanation: separating disciplines may create an illusion of precision. If I separate snooker from 9-ball, I may feel more confident in my conclusions, but that confidence does not necessarily correspond to actual accuracy. Sample sizes for each discipline individually are smaller than the merged sample, and a smaller sample usually leads to wider confidence intervals. In other words, by separating labels, I may have traded one type of error for another. This I must concede: separating disciplines is not a free solution.

However, weighing these two explanations against the available evidence, I still lean towards the conclusion that label merging is a more serious problem than a reasonable design choice. The reason is that among the eight merged-label datasets I examined, none contained documentation stating that merging was intentional. No metadata field said "this label covers multiple disciplines." If it were a design choice, it would have to be recorded. The silence of the metadata is a signal, and I read it as a sign of carelessness rather than intent.

On sample size, I agree that separating labels reduces it, but I regard this as a price worth paying, not a reason not to separate. A small but discipline-correct sample remains more useful than a large sample mixing different phenomena. This is the principle I applied from 2026, when analysing Morocco's four knockout matches at the World Cup: a sample of four matches is too small to claim a sustainable tactic, but it is enough to describe a specific phenomenon, provided I state the limits clearly. Morocco's miracle was not magic; it was deliberately defended square metres — and to see that, I had to know I was measuring square metres, not something else.

There is one more counter-intuitive point I want to raise: sometimes label merging itself creates a valuable new metric. If I merge all billiards disciplines and measure total televised playing hours per year, I have an indicator of the industry's overall health — one that belongs to no single discipline. The problem only arises when I try to infer from that aggregate metric conclusions at individual or technical level. In other words, the "billiards" label is not absolutely wrong; it is only wrong when used at the wrong level of analysis. This is a subtle distinction that I believe many analysts overlook, and it explains why this problem persists so stubbornly.

Takeaway

What I take from this analysis is not a call to overhaul the entire billiards data system — that is beyond the reach of a journalist. What I take is a question I will ask in every future project: what am I measuring, and at what level does that measurement mean anything. In the current transfer window, as players weigh moves between disciplines for prize money and sponsorship, this question becomes more urgent than ever. A dataset that does not know which sport it is describing cannot tell you where any player is heading. And if you do not know where players are heading, you do not have a market. You only have a list of names.

Data limitations

The analysis above rests on observation of eight datasets labelled "billiards" that I have had access to over the past two years. A sample of eight is small, and insufficient to claim that the entire billiards data industry commits this error at a similar rate. Figures on Ronnie O'Sullivan's century breaks, on the Triple Crown records of three players, and on carom players' averages are drawn from the public records of the respective governing bodies; readers should verify the original sources before citing them. Some assessments of "break quality" and "cue-ball control" are qualitative and cannot be converted into absolute numbers, and I have flagged this at each point. Finally, I do not yet have data over a long enough span to judge whether cross-discipline talent flow is a sustainable trend or a short-term phenomenon. That is the signal I will keep tracking in the rounds ahead.