Chess Ratings and ELO Explained
A chess rating is a single number that estimates a player's strength, based on the results of their games. Here's how the system actually works, and what counts as "good" at every level.
What is the ELO system?
Chess ratings are calculated using the ELO system, named after Arpad Elo, the physics professor who designed it in the 1960s for the US Chess Federation — it was later adopted by FIDE, the World Chess Federation, and eventually by every major online platform.
The core idea is simple: every player has a number representing their current strength. After a game, both players' ratings are adjusted based on the result and on what the result implied — beating a much higher-rated player gains you a lot of points; beating a much lower-rated player gains you very few, because you were expected to win anyway.
How rating changes are calculated
Before a game, the rating difference between the two players predicts an "expected score" — the probability each player is expected to win. After the game, each player's rating moves toward or away from that expectation:
- Win against a favourite (higher-rated opponent): you gain a large number of points.
- Win against an underdog (lower-rated opponent): you gain only a few points.
- Draw against a much stronger opponent: counts almost like a win for rating purposes — you still gain points.
- Loss against a much weaker opponent: you lose a large number of points.
The size of each rating change is also controlled by a factor called K — newer or lower-rated players typically have a higher K-factor, so their rating moves faster and settles on an accurate number sooner; established, high-rated players have a lower K-factor, so a single result barely moves their rating.
What counts as a "good" chess rating?
This depends heavily on which rating pool you're comparing against — online ratings (chess.com, Lichess) and FIDE (official, over-the-board) ratings are not directly comparable, and are often several hundred points apart for the same player. Roughly, for FIDE-style over-the-board ratings:
| Rating range | Typical level |
|---|---|
| Below 1000 | Beginner — still learning basic tactics and rules of thumb. |
| 1000–1400 | Casual club player — knows the rules well, makes regular tactical mistakes. |
| 1400–1800 | Solid club player — understands opening principles and common tactics, still drops material sometimes. |
| 1800–2000 | Strong club player — often holds a title like Candidate Master in some federations. |
| 2000–2200 | Expert / Candidate Master — a serious, dedicated player. |
| 2200–2300 | FIDE Master (FM) territory. |
| 2300–2400 | International Master (IM) territory. |
| 2500+ | Grandmaster (GM) territory — the highest standard title in chess. |
| 2800+ | World top-10 level. The all-time peak rating, held by Magnus Carlsen, is around 2882. |
On online platforms, the numbers look different because the player pool skews younger, more casual, and includes many beginners and bots — a 1500 on Chess.com's Rapid rating is a notably different strength than a 1500 FIDE rating.
Chess titles and how they relate to rating
FIDE titles are earned, not just reached by rating alone — most require both a minimum rating and specific tournament performances called "norms," and once earned a title is held for life even if rating later drops. In ascending order:
- CM — Candidate Master
- FM — FIDE Master
- IM — International Master
- GM — Grandmaster (the highest title FIDE awards)
There are equivalent titles for women (WCM, WFM, WIM, WGM), and separate titles for correspondence and online chess.
Why do online and FIDE ratings differ so much?
A few reasons: online games are usually played at faster time controls (blitz/rapid), which tend to produce more mistakes and wider rating swings; the online player pool includes far more beginners and inconsistent players, which shifts the whole scale; and online platforms each run their own separate rating pool, so a rating on one site isn't directly transferable to another. Treat any single rating as an estimate within its own pool, not an absolute measure of chess skill.
Provisional vs established ratings
A new player's first several games produce a provisional rating — it moves quickly and can swing hundreds of points as the system figures out roughly where you belong. After enough games, the rating becomes established and changes much more slowly, a few points at a time.
Working on your own rating? The fastest way to climb from beginner level is fixing tactical blunders — see Chess Tactics for the patterns worth learning first.
Frequently asked questions
What is a "good" chess rating for a beginner?
Anything from 400–1000 is normal while you're still learning the rules and basic tactics — there's no rating to be embarrassed about while starting out. Most players who stick with the game comfortably reach 1000–1200 within their first year or two of regular play.
Why did my rating drop after a draw?
If you drew against a significantly lower-rated opponent, the expected result was a win, so a draw still counts as an underperformance relative to that expectation, and costs you a small number of points.
Is a 2000 rating good?
Yes — a FIDE rating of 2000 puts you in the top few percent of tournament players worldwide, and is often enough to hold a national or regional expert-level title.
Do online ratings and FIDE ratings mean the same thing?
No. They're separate rating pools calculated independently — the same player will typically have noticeably different numbers on Chess.com, Lichess, and FIDE.