ELO ratings for pool leagues
Most bar pool has no rating at all. You know who is good because you have watched them. That works in one venue and falls apart the moment anyone travels.
What ELO actually is
ELO is a rating system invented for chess and now used for everything from football to video games. Each player has a number. When two players meet, the system compares their numbers and predicts a result. After the match, both numbers move, and how far they move depends on how surprising the result was.
The intuition is simple. Beating someone far stronger gains you a lot of points. Beating someone far weaker gains you almost nothing. Losing to someone far weaker costs you a lot. That is what makes the number meaningful: it is not a score, it is an estimate of strength that is constantly corrected by results.
Why the maths matters less than you think
You will find long arguments online about whether to use standard ELO, Glicko, TrueSkill, or something else. For a bar league, the difference between them is small compared to two much bigger questions.
- What counts as a match? A single game and a best-of-five should not be worth the same. Decide this before you start.
- How fast should ratings move? A new player should reach their true level in a handful of matches. An established player should not swing wildly after one bad night. Every system has a dial for this, and it is the setting that actually affects your league.
The case for per-venue ratings
This is where most pool rating attempts go wrong. They build one global ladder, and within a month it is a mess, because a strong player at one venue may never meet a strong player at another. The numbers drift apart, and a 1500 at one bar is not a 1500 at another.
Pool Queue keeps the rating per venue. That has three practical effects.
- The numbers stay comparable inside a room. Everyone in a given venue is connected through shared matches, so the ratings mean something relative to each other.
- Regulars get a real ladder. A venue's own history, its own rivalries, its own standings. That is the thing people actually care about.
- New venues start clean. A fresh bar does not inherit a decade of drift from somewhere else.
The trade-off is real: a player with ratings at three venues has three numbers, and they are not directly comparable. In practice that is less confusing than a single global number that is quietly wrong.
Handicapping follows from the rating
Once you trust the number, you can use it to even up games. The difference between two ratings predicts the chance of each player winning. A few hundred points is a meaningful edge; a hundred is not much.
For bar pool this turns into simple formats. Give the weaker player a game start, or alternate breaks, or spot a ball. The exact adjustment matters less than applying it consistently, which is only possible if the underlying rating is not being hand-waved each week.
What breaks a rating system
Three things, in order of how often they occur.
- Players who never rejoin. Someone wins five, leaves, and parks a rating that follows them forever. There is no clean fix. Soft-delete and a minimum match count before a rating counts publicly is a reasonable mitigation.
- Smurfing and favours. Two players agreeing to let one win to move a number. Any rating system is vulnerable, and the answer is social as much as technical.
- Duplicate profiles. The same regular appears twice with different names and half the history in each. This is why player merge exists in the venue app, with rating recalculation from match history rather than a hand-added number.
Starting with no history
Everyone starts at the same number, or you seed a few known regulars and let the rest settle. Either works. What matters is that the ratings are allowed to move quickly at the start and slowly later, so new players converge on something true within a few weeks rather than a year.
If the first month of numbers looks wrong, that is normal. Rating systems are not accurate on day one; they become accurate by being corrected.
Last updated 16 September 2026