How Chess Elo Ratings Rise and Fall
The Elo rating system predicts a probability of winning based purely on the rating gap between two players, then adjusts the rating according to how far the actual result deviated from that prediction. A player rated 200 points higher is expected to win about 76% of the time, and the gap between that expectation and the real outcome determines how much the rating moves.
The same result can swing your rating very differently depending on the opponent. Beating a much stronger player yields a big rating gain because the expected win probability was low, while losing to a much weaker player causes a big rating drop for the same reason. Expected outcomes, like a strong player beating a weak one, barely move the needle.
The K-factor controls how much weight a single game carries. New players, whose ratings haven't stabilized yet, use a higher K so their rating converges quickly toward their true strength. Established masters use a lower K so one or two results don't swing a well-calibrated rating too much.
Frequently Asked Questions
Elo rating derives an expected win probability from the rating gap between two players, then multiplies the difference between the actual result and that expectation by a K-factor to get the rating change. Beating expectations raises your rating; falling short lowers it.
The K-factor controls how much a single game affects your rating. New players get a higher K (often 40) since their rating is still unstable, while established masters get a lower K (10-20) for smoother, more gradual changes.
Yes, but only slightly. Beating a lower-rated opponent is already the expected outcome, so the gap between expected and actual result is small, which means a small rating gain.