US Chess Correspondence Chess ratings are calculated as each game is reported the Correspondence Chess Director. Below is an explanation of how rating calculations are computed for Established and Provisionally rated players. If you have any questions relating to the calculation of CC ratings or other US Chess CC matters, please contact the US Chess Correspondence Chess Director Alex Dunne, PO Box 367, Sayre, PA 18840 or email Alex at [email protected].
ESTABLISHED RATED PLAYERS
CC ratings are easy to compute. Here's how it works.
Rn = Ro + .04(ED) +/- 16
This means that a new rating (Rn) is determined by taking the old rating (Ro), adding or subtracting 4 percent of the difference in ratings between opponents (.04(ED)), and adding or subtracting 16 points.
A 1600 rated player beats a 1400 rated player.
Ro + .04(ED) +/- 16 Rn
A) Winner 1600 -8 +16 = 1608
B) Loser 1400 +8 -16 = 1392
Had these players drawn, the +/- 16 would be disregarded and only the .04(ED) would be used to figure their new ratings.
Ro + .04(ED) +/- 16 Rn
A) 1600 -8 = 1592
B) 1400 +8 = 1408
Rating differences that exceed 350 points are figured as 350 points.
For players rated 2100-2399, the formula Rn = Ro + .03(ED) +/- 12 is used.
For players rated 2400 and above, the formula Rn = Ro + .02(ED) +/- 8 is used.
Crossing boundries occurs when players cross 2100 or 2400. They will have their ratings calculated according to the K factor appropriate to their current rating. However, the points under or over these boundries will be proportionally adjusted to reflect the new K. (For example, a player goes from 2090 to 2112 with a K of 16. The player has 12 points in the range where K = 12, or 3/4 of 16. The proportional adjustment is 3/4 x 12, or 9. The new rating is 2109.)
Provisional ratings are figured differently. During the first 25 games played by a correspondence chess player, we are trying to determine his or her relative playing strength. There are two things worth noting: l. provisional ratings usually fluctuate wildly, and 2. a win does not guarantee a rating will go up, while a loss does not guarantee a rating will go down. If the provisional player has beaten someone (established or provisional) 400 points or more below him or lost to a player (established or provisional) 400 or more above him, each player will be given credit for one more game played, but neither player's rating will change at all.
John Doe, a previously unrated player, plays five games.
He beats a 1200 rated player.
He beats a 1500 rated player.
He loses to a 2200 rated player.
He beats a 1000 rated player.
He draws with a 1000 rated player.
Here's what happens to his rating:
This game gives him a first rating of 1200 + 400, or 1600.
This game gives him 1500 + 400, or 1900. The previous 1600 plus this divided by two games gives him a rating of 1750.
This loss is so predictable that he suffers no penalty. He gets a game credit for the experience but no rating change.
This victory is so predictable that he gets no rating credit, only another game credit.
This draw represents a rateable result. He gets only the opponent's rating for a draw (without the 400 points added or subtracted for wins and losses). All his game results are summed (1600 + 1900 + 1750 + 1750 + 1000 = 8000) and divided by 5, the number of games. Therefore, after five games, John Doe has a provisional rating of 1600.
If both players are unrated and the result was a win/loss, the winner receives a rating of 1700 and one game credit. The loser receives a rating of 1300 and one game credit. If the game was drawn, each player receives a rating of 1500 and one game credit.
When a win is achieved over a quitter and full rating credit does not apply (example:opponent deceased)the winner gets 20 percent of the points awarded as if the win were over an equally rated opponent but no more than l0 points. No game credit is given in this situation.
Codes: 1.0=win; 0.1=loss; 0.5=draw; Q=forfeit loss; Ox=special forfeit loss.