Post
A mathematical system that assigns skill ratings and predicts match outcomes based on win/loss history.
Originally created by physicist Arpad Elo for chess, this system assigns every player a numerical rating. When you beat someone rated higher than you, you gain more points than if you beat someone lower. The beauty is in its self-correcting nature: over enough games, your rating converges on your true skill level. Most modern games use modified versions, like Microsoft's TrueSkill or Riot's MMR system, which factor in additional variables like performance metrics and team composition.
Example
Chess.com uses a direct ELO implementation where grandmasters sit around 2700+. Overwatch 2's competitive mode uses a modified system under the hood, even though it displays skill tiers like Gold and Diamond on the surface. Valorant's Ranked Rating similarly wraps an MMR-based system in visible rank labels.
Why it matters
Without ELO-style matchmaking, competitive games would be chaos. New players would constantly face veterans, and nobody would have fun. It is the invisible engine that makes ranked play feel fair and progression feel meaningful.
Related concepts