US Chess (USCF) rating estimator
This estimator shows how your US Chess rating changes after one event. It uses the standard formula from the published US Chess rating system: the score your rating predicts, a K factor set by how many games your rating rests on, and a bonus for an unusually strong result. Enter your rating, your games and the results.
Your rating before the event, 100 to 3000.
A whole number above 8. It sets how far one event can move you.
Use 1 if you met everyone once. It decides whether the bonus can apply.
Only the Regular rating gets this K. Leave it unticked for a Quick or Blitz rating. It only changes K for players rated above 2200.
Example: a 1700 player with 30 games on record in a four-round event. Type your own rating and games to see your result.
- Rating change
- +23
- New rating 1,723
- Exact new rating before rounding
- 1,722.65
- Score and expected score
- 2.5 of 4, expected 1.86
- Expected score is the points the rating gaps say you should score.
- K factor
- K = 33.32
- K sets how far one event can move your rating. 30 previous games count as 20.01 effective games at a rating of 1700, and you played 4 games in the event: K = 800 / (20.01 + 4).
- Bonus points
- +1.33
- K x (score - expected) is 21.33, which is 1.33 above the bonus threshold of 20.00 (10 x the square root of 4), so that excess is added a second time.
Formula: The US Chess Rating system, Mark E. Glickman and Thomas Doan, April 6, 2026, read 2026-11-05. An estimate, not the rating US Chess publishes.
| Game | Opponent | Result | Expected |
|---|---|---|---|
| 1 | 1,650 | Win | 0.57 |
| 2 | 1,700 | Draw | 0.50 |
| 3 | 1,750 | Loss | 0.43 |
| 4 | 1,800 | Win | 0.36 |
How the new rating is computed
US Chess compares what you scored with what your rating said you should score. The expected score for each game comes from the gap between the two ratings, and the sum over the event is your expected points. K sets how far the gap between your points and your expected points moves you.
New rating = rating + K x (points scored - expected points) + bonus
Expected points for one game = 1 / (1 + 10 ^ (-(your rating - opponent rating) / 400))
K and effective games
K is not a fixed number. It shrinks as your rating rests on more games, but only up to a cap that depends on your rating, called the effective number of games. A rating that counts for few games moves quickly; one that counts for many moves slowly.
K = 800 / (effective games + games in the event)
Effective games = the smaller of your games before the event and N*, where N* = 50 / square root of (0.662 + 0.00000739 x (2569 - rating)^2) up to a rating of 2355, and 50 above it
| Rating | Most games that count |
|---|---|
| 1,000 | 11.51 |
| 1,400 | 15.24 |
| 1,700 | 20.01 |
| 2,000 | 28.61 |
| 2,300 | 45.71 |
| 2,355 | 49.99 |
| 2,400 | 50.00 |
Dual-rated events
When an over-the-board regular event has a time control from 30 to 65 minutes, it is rated in two systems at once. For a player rated above 2200 the K for the Regular rating is then 800 x (6.5 - 0.0025 x rating) / (effective games + games in the event) below 2500, and 200 / (effective games + games in the event) from 2500. The lower K applies to the Regular rating only, so tick the box only when the rating you typed is your Regular rating; a Quick rating keeps the standard K.
The bonus
A result far above what your rating predicts earns a second payment. It needs at least 3 games and no opponent met more than twice (once in a 3-game event). A 3-game event counts as 4 rounds when the threshold is set.
Bonus = the larger of 0 and K x (points scored - expected points) - 10 x square root of the larger of 4 and the games in the event
Worked examples
These rows were produced by the same code as the estimator above. Set the estimator to the same inputs and you will see the same numbers.
| Event | K | Score | Expected | Bonus | Change | New rating |
|---|---|---|---|---|---|---|
| Four games, a win, a draw, a loss and a win (the calculator opens with it) | 33.32 | 2.5 | 1.86 | +1.33 | +23 | 1,723 |
| Four wins against equal opponents, 30 games before the event | 33.32 | 4 | 2.00 | +46.63 | +113 | 1,813 |
| The same four wins, 100 games before the event | 33.32 | 4 | 2.00 | +46.63 | +113 | 1,813 |
| The same four wins, one opponent met three times | 33.32 | 4 | 2.00 | 0.00 | +67 | 1,767 |
What the estimate leaves out
US Chess rates an event in two passes and uses each opponent's first-pass rating in the second; the estimator applies the formula once with the opponent ratings you type, so typing your opponents' ratings after the event brings it closer. It also leaves out rating floors above 100, opponents with no rating, and players whose rating rests on 8 games or fewer or who have only won or only lost, which US Chess rates with a different algorithm.
For the FIDE version of this calculation, use the chess rating calculator. For how the systems differ, read chess ratings explained.
Sources
The US Chess Rating system, Mark E. Glickman and Thomas Doan, April 6, 2026, read 2026-11-05. The US Chess Rating system: effective games in section 3, the standard formula, K and the bonus in section 4.2, the floor of 100 in section 5. The bonus multiplier of 10 has applied since January 1, 2025.
Only the rating US Chess publishes after it rates the event is official.
Questions people ask
- Is this my official US Chess rating?
- No. It estimates the result of one event from the numbers you enter. US Chess rates an event in two passes and uses each opponent’s first-pass rating in the second, while this tool uses the ratings you type for both. The official figure can also differ when a rating floor applies or an opponent has no rating yet.
- Why does the tool refuse a rating built on 8 games or fewer?
- US Chess uses the standard formula once a rating rests on more than 8 games. Below that it runs a different, iterative algorithm over the whole event, and so it does for a player who has only ever won or only ever lost. The tool stops and says why when you enter 8 games or fewer. It has no box for a history of only wins or only losses, so it cannot see that case: if that is you, the standard formula it shows is not the one US Chess uses.
- What are effective games?
- A rating built on many games is less reliable at a low rating than at a high one, so US Chess caps the games that count. A 1700 rating counts for at most 20.01 games however many you have played, which keeps K near 33 for a four-round event. Above 2355 the cap is 50 games.
- What is the bonus?
- It is a second payment for a result far above what your rating predicts. When K times the gap between your points and your expected points is larger than 10 times the square root of the rounds (counting a 3-round event as 4), the excess is added again. It needs at least 3 games and no opponent met more than twice, or once in a 3-game event.
- Does it work for blitz, quick and online ratings?
- The formulas are the same for all six US Chess rating systems, and the rated systems are computed separately, so enter the rating and game count from the system the event belongs to. The dual-rated tick applies only to a Regular rating, at over-the-board events with a time control from 30 to 65 minutes; leave it unticked for a Quick or Blitz rating.
- What do I enter for an unrated opponent?
- The tool needs a number for every opponent. For an opponent with no rating in that system, US Chess builds a starting rating from the opponent’s other ratings (the other US Chess systems, FIDE and the Canadian federation) or, with none, from age. This tool does not reproduce that step, so enter your best figure and expect the result to differ more from the official one.
- Is this the FIDE formula?
- No. FIDE reads expected scores from its own table and uses its own K factors. The chess rating calculator on this site follows the FIDE Handbook; this page follows the US Chess system.
Want a plan for the games behind the numbers?
The first lesson reviews your games, reads your level and ends with a written study plan.