How to read win rate, pick rate and ban rate

Tier lists are built on three numbers. Read well, they tell you a lot; read badly, they lead to wrong conclusions. What you should know about each one.

Win rate: the share of games won

Win rate is the percentage of games a champion wins. Since exactly one team wins every game, the average across all champions is 50%. A champion at 51% wins a bit more than normal; one at 48%, a bit less.

The differences that matter are small. Between a 49% champion and a 52% one there is a big gap in tier-list terms, even if it looks like just three points. That is why tier lists rank champions around 52–53% as S or A, not 60%.

How big a sample you need

A win rate calculated from few games means almost nothing. A useful rule: the approximate margin of error of a win rate near 50% is about ±3 points with 1,000 games, ±1 point with 10,000 and ±0.3 with 100,000. If a champion has 54% over 800 games, it may be genuinely strong… or a lucky 50%.

This explains why Grandmaster and Challenger data swings so much: there are few players and each champion gets few games. Pick My Champ marks those ranks with low reliability for that reason.

Pick rate: how popular it is

Pick rate is the percentage of games in which someone picks the champion. It is useful for two things:

Knowing how much to trust the win rate. A high pick rate means many games and a stable win rate.
Spotting bias. Champions with a very low pick rate are mostly played by specialists who have played them for years, so their win rate can be inflated: do not expect the same the first time you try them. Conversely, a newly released champion, or one that became popular after appearing in pro play, usually has a low win rate because many people are still learning it.

Ban rate: what annoys people

Ban rate is the percentage of games in which the champion is banned. It mostly shows how frustrating or strong a champion is perceived to be: some heavily banned champions have a normal win rate because they are very annoying for players who do not know how to play against them. A high ban rate also reduces the number of games for the champion and therefore the precision of its win rate.

Why they change by rank

The same champion can have 53% in Iron and 48% in Diamond, or the other way round. The most common reasons:

Mechanical difficulty. Hard champions (demanding combos, little margin for error) perform worse in low ranks and better in high ranks, where they are played more precisely.
Coordination. Champions that rely on the team following up perform better the more organised the game is.
Punishing mistakes. Mistakes are punished less in low ranks, and easy, durable champions take advantage of that.

That is why Pick My Champ lets you filter recommendations and tier lists by rank: the ideal champion in Silver is not always the ideal one in Diamond.

Matchup win rate

The win rate of a specific matchup (for example, one champion against another in the same lane) follows the same rules, with one caveat: there are far fewer games of each pairing than of the champion overall, so the margin of error is larger. That is why we only flag a matchup as favourable or unfavourable when there is enough sample and the difference is clear. You can look them up in Counters.

In short

50% is the average; the relevant differences are a few points.
Always check how many games are behind a percentage before trusting it.
Very low pick rate: careful, it may be inflated by specialists.
Compare data for your rank, not the overall numbers.