Skip to content
How You Rank
All ranking methods
Scales globally

Multiplayer Elo

Elo generalised: a six-player free-for-all becomes every pair of duels inside it, with the K-factor scaled down so one match does not count six times over.
FamilyElo-like

What does a rating gap mean?

Interactive example · no real player data

Expected score50%
−8000+800100%

Equal ratings split the expectation 50/50. Expected score includes draws; it is not a guaranteed win rate. In multiplayer Elo, each pair is compared and updates are scaled by the field size.

How it works About 1 min read

Multiplayer Elo extends the classic Elo rating system to work with free-for-all games where 3 or more players compete. It decomposes each match into all pairwise comparisons and updates ratings based on placement order.

How it works: A 4-player game produces 6 pairwise comparisons (every possible pair). Each pair gets an Elo update based on who placed higher. The K-factor is scaled down by (n-1) so a 4-player game shifts ratings by roughly the same total as a 1v1 game -- no inflation.

When to use it: Board game groups, party games, or any context where 3+ players compete and you want a single skill rating. Works alongside standard Elo -- use Elo for your 1v1 games and Multiplayer Elo for group games.

Watch out for: With more players per match, there's more noise per game. Ratings take a few more matches to stabilize compared to standard 1v1 Elo. The system works best when there's a clear placement order (scores or explicit rankings).

Will it fit your results?
Works well with

Games whose results this method reads cleanly

Also in Elo-like

Adaptive skill ratings that update after every match.

Multiplayer Elo — How You Rank