NOTE: This is a very old SSDF Chess Ratings notes - they are not current.The rating numbers that you see on the Your Move Chess & Games website are based upon the United States Chess Federation rating system. This system allows players to judge their chess strength relative to other rated chess players.
Chess computers and chess playing programs also have ratings, and by using the rating numbers supplied for each chess computer/software, you can determine just how strong the computer will play under tournament time conditions of 40 moves in 2 hours (3 minutes per move).
You may find that the ratings indicated for the computers are different than ratings that may be claimed from other sources. The reason why ratings do not always agree are several fold:
Computers always play better, relative to humans, at fast time controls. So companies and organizations that make claims based upon Action Chess (each side gets 1/2 hour for entire game) or Blitz Chess (each side gets 5 minutes for entire game), are stating numbers that may be easily misconstrued by the consumer. For instance, a computer that plays 2200 at tournament time controls (40 moves in 2 hours or 3 minutes a move) is likely to play 2260 to 2350+ at Action Chess and 2400+ at Blitz Chess. Your Move Chess & Games has always chosen to quote tournament time ratings.
The ratings system is not a perfect science. It is a system based upon estimates. The more games played, the smaller the standard deviation of error. Those who estimate ratings sometimes do so based upon small samplings of games, based upon problem solving speeds, based upon games against human players, or based upon games against other chess computers or software. If sheer numbers of games played is the most accurate measure of a computer's rating, then the Swedish Rating List would be the gospel of computer ratings because as of the last list (they come out every month or two), over 54,000 games have been played using 163 different computers/software. Since the Swedish List is done using the Swedish rating's system, those of us in the U.S. must add 180 points to their numbers.
Historically, many manufacturers and retailers of chess computers and chess playing software have found it necessary to exaggerate the ratings of their own products in order to distance themselves from the pack. Your Move Chess & Games has never found that necessary because we have always attempted to sell every chess playing product on the market. The ratings quoted by Your Move are reliable and accurate or at least as reliable and accurate as estimates can be. The rating system is based upon numerical values, sort of like bowling averages. If you bowl against a opponent with a better average, you are more likely to lose than to win. And the bigger the differential, the worse your thrashing is likely to be. The same applies to rated chess players. A higher rated player is likely to beat a lower rated player, and the larger the difference in rating the more likely that the beating will be severe.
When you play against a chess computer or chess playing software, the same dynamics are in place. If you play 600 rated chess, and you set some of the finest chess playing software to play at 3 minutes per move on your 200mhz Pentium, you are likely not to know what hit you because it is playing at a 2700 level. This is no different from inviting the town Grandmaster for a game of chess over a cup of coffee. You will be dead after the third move.
So how do you make a determination as to how strong your chess computer opponent should be? The rule of thumb (invented by Your Move Chess & Games) is that unless you are the beginner of beginners or a child just learning, you should attempt to buy the strongest (highest rated) chess opponent that you wish to afford. No, this is not intended only for masochists because the really strong programs/computers can be made fairly stupid so that if you wish to win every once in a while, it will be possible. But one learns in chess by being beaten, and the electronic opponents are nice enough to tell you why you are losing. You will become a better player by losing against a grandmaster than by beating a patzer.
The really nice thing about buying strong is that it will never waste your time making you wait for each move because the stronger a chess computer is, the faster it will make good moves. If you buy something that is not very strong, you will be forced to wait long periods of time to get decent games, and the time will eventually come (if you are capable of improving) when you will wait long periods of time but win every game every time anyway. Not a good scenario. Remember, we are not advising spending more money than you want to, just suggesting you buy the strongest program/computer for the money you wish to spend.
The rating distribution chart below is what tournament chess players use to compare their rating with the rating of their opponents and to compare their rating with the ratings of all tournament players in the United States.
Please keep in mind that ratings are estimates and not exact measurements of chess playing ability. Generally if you are rated 200 points higher than your opponent, you would expect to win 75% of your games. If you are rated 800 points higher than your opponent, you can expect to win 99+% of your games.
As an amateur, if you beat an opponent with the same rating as you, you will add 16 points to your own rating. The most you can gain from winning any single game is 32 points (if your opponent is 400 or more points higher than you). The least you can gain is 1 point (if your opponent is 400 or more points lower than you). Points are lost in the same way. As your rating gets above 2100, the points gained or lost begin to decrease.
The chart contains the ratings of all United States Chess Federation members who played in over-the-board games and is accurate as of January, 1997. The classes are designated on the left. The rating range is in column two. The number of players in a given class are in column three, and the percentile is listed in the last column on the right. There were 59,517 rated players in the United States in January and their ratings are distributed in the following fashion:
SETS FROM AMERICA'S LARGEST CHESS STORE