How to read win rate, pick rate, and sample size

A 53% win rate does not mean a champion or build is universally strong. It reflects who selected it, the opponents they faced, and the number of games observed. A difficult champion may be strong but have a lower overall win rate, while a niche counter build can post a high result in favorable situations.

Win rate is most useful when comparing choices with similar conditions and sample sizes. A build at 51% over 500 games is generally more stable evidence than one at 62.5% over eight games.

In a ten-game sample, one result moves win rate by ten percentage points. At 100 games it moves one point, and at 1,000 games only 0.1 points. The same displayed win rate can therefore carry very different levels of confidence.

infolol excludes champion-role samples below 100 games from recommendations and public detail pages. This is our editorial quality gate, not a threshold set by Riot Games.

A high pick rate means the choice appeared across more players and situations. A low pick rate may be concentrated among specialists, counters, or experiments. Popularity does not prove that a build is best, but it helps describe how general the evidence is.

• When win rates are close, start by evaluating the larger and more common sample.
• Treat low-pick alternatives as situational until you understand why they were selected.
• Expect every sample to be volatile immediately after a patch.

Confirm the patch and role, review the role-level sample, then compare the leading builds. Finally, adjust for both team compositions and your recall gold. Statistics should narrow the decision, not replace in-game judgment.