How the BTR works, and how far you should trust it
The BilliardToday Rating is our own rating for three-cushion players. It is not an official federation ranking. This page explains exactly what goes into it, what it measures well, and where the data is still thin — including the parts we would rather not have to admit.
The short version
- It rates three-cushion only. A player's results in libre, balk-line, one-cushion or five-pins are ignored completely. A player who mostly plays other disciplines will have a rating built on very few matches.
- It uses who you played and how you played. Beating a strong opponent moves your rating more than beating a weak one, and winning with a high average is worth more than grinding out a win.
- It fades when you stop playing. A rating fades toward the baseline while a player is inactive, so the list shows who is competing now rather than who was strong years ago.
- The data is not evenly spread around the world. We hold deep domestic results for Greece and mainly international results for everyone else. That asymmetry is the single biggest limitation of this rating.
- Uncertainty is published, not hidden. Every rating carries a confidence level. A number marked with an asterisk is genuinely provisional and should not be read as a ranking claim.
How a rating moves
Every match updates both players. The size of the change depends on three things: the result, the strength of the opponent, and how well you played.
1. The result
Win, draw or loss. A three-cushion match can be drawn when the innings limit is reached with both players on the same points.
2. The opponent
Beating a stronger player earns more than beating a weaker one, and losing to a stronger player costs less.
3. Your performance
Your average per inning is compared with what is normal for a player at your level, measured over your last ten matches so a single good day changes nothing.
A worked example
A 1800-rated player beats an 1800-rated opponent. That is exactly what the rating expected, so the gain is small. The same player beats a 1400-rated opponent while scoring well below their usual average: the win was expected and the performance was poor, so the rating can actually fall. Winning is not automatically rewarded — winning as expected, in the way you were expected to, is.
Two things that are deliberately not part of it
Past titles do not count toward the rating, and neither does reputation. A former champion who has stopped competing loses ground to an active player. Where a match was played is weighted, though: a World Cup match counts more than a club match, because the field is stronger.
Event weighting
| Event level | Examples | Weight |
|---|---|---|
| World | World Cup, World Championship | 1.5× |
| Continental / national | European Championship, national championship, Euro Grand Prix | 1.0× |
| Club / other | Club events, open and invitational tournaments | 0.5× |
Age-restricted events are then scaled down on top of this, because their entry rules exclude the strongest players: junior and under-19 events count at half the value above, under-21 events at three quarters. Women's events are not scaled — see the reasoning below.
Where the results come from
The rating is only as good as the results behind it, and that is where our coverage is uneven. Being upfront about this matters more than the formula. The split below is a snapshot and shifts as events are added.
| Event level | Events | Matches | Where |
|---|---|---|---|
| World Cup | 78 | 17,366 | Worldwide |
| National championship | 86 | 10,521 | Greece only |
| European Championship | 34 | 1,617 | Worldwide |
| World Championship | 26 | 1,413 | Worldwide |
| Other rated events | 22 | 1,257 | Mostly Greece and Europe |
The asymmetry, stated plainly
We have nationwide domestic results for Greece and no other country. Every player outside Greece appears in the rating only through international events. A Greek club player can therefore build a rating from a hundred local matches, while a Korean or Turkish player of similar ability may have only a handful of World Cup appearances to be judged on. Their confidence levels are not comparable, and their ratings should not be treated as directly comparable either.
Why Greek players are not simply ranked higher
National matches carry the lower 1.0× weight, and they are played against a shallower field than a World Cup. A high domestic average still has to be earned against international opponents to move a player far up the table — which is why the top of the list is still dominated by players who compete worldwide.
How much to trust a number
Every rating carries an uncertainty figure, which grows when a player has few matches or has not played for a long time. We publish it rather than hide it, and the leaderboard marks the weakest cases with an asterisk.
Enough recent matches behind them for the number to mean what it says.
Marked with an asterisk on the leaderboard. Shown for completeness, not as a ranking claim.
Roughly 89 out of every 100 numbers on the leaderboard are still provisional. That is the honest state of the data today.
Why a rating can look wrong
Too few matches
A player with a handful of recorded matches can sit far from their true level in either direction. This is common outside Greece, where a player may only appear in one or two events.
Long gaps
Ratings fade while a player is inactive. Someone who was world class five years ago and has not competed since will have drifted toward the middle of the table, which is deliberate: we no longer know their current level.
Playing mostly other disciplines
Strong carom players who compete mainly in libre or balk-line will show a low three-cushion rating built on very few matches. That is not a judgement on their ability, only a statement that we have little three-cushion evidence.
Age-restricted events
A tournament that only admits players below a given age cannot have the same field as an open tournament of the same level, because the strongest players in the world are excluded by its entry rules. Those events therefore carry a lower weight: under-19 and junior events count at half weight, under-21 events at three-quarter weight. Under-25 events are unchanged, since those players are adults.
This is a statement about the event's entry rules, never about the player. A 17-year-old who reaches the main draw of an open World Cup is rated exactly like an adult — same opponents, same field, same weight. Only results from age-restricted brackets are scaled down, and only for the players whose entire record sits inside them; most juniors in this database also compete in open events and are unaffected.
Women's events are not down-weighted
Women's competitions are open-age elite events: the players are adults competing at the top of their discipline, and several also compete in open fields. The pool is thinner, but that is a different thing from a weaker field per player, and the rating already gives little credit for beating low-rated opponents. Reducing the weight would amount to asserting that women's competition is worth less, which the data does not support — so we disclose the thinner pool instead of adjusting it away.
Countries we simply do not cover
Federations that do not publish results in a form we can read are invisible here, however strong their players are. Absence from the leaderboard usually means missing data, not a weak player.
How it gets better
Ratings mostly get more accurate on their own as players accumulate matches and as the gaps between their appearances shrink. Newly added events also raise the confidence of everyone who took part.
The three things that would move the needle most are: more clubs recording their results in a format we can import, national federations sharing their domestic championships with us, and more frequent World Cup data. Two of those are outside our control, and we would rather say so than pretend the coverage is even.
Coverage already exists in machine-readable form for some federations, so the constraint is integration work rather than availability. We add sources as time allows and recalculate the whole table afterwards, which is why ratings sometimes shift for everyone at once.
What the BTR is not
- Not an official ranking. It is not sanctioned by the UMB, the CEB or any national federation, and it carries no seeding or qualification rights.
- Not a measure of a career. It describes recent competitive form in one discipline. It says nothing about a player's record, titles or contribution to the sport.
- Not the same as the official UMB ranking. The UMB rank is built on a season of World Cup results and decides entries. The BTR is built on everything we hold, over the whole history, and is intended to be descriptive.
- Not comparable across disciplines. A strong BTR is a three-cushion statement and nothing more.