 |
|
|
| Recently I received some very interesting comments from John P. McCumiskey (a fellow APCT member) on the topic of chess ratings. My response, naturally, was to invite him to write an article on the subject detailing his views. Thanks to John for sharing his thoughtful viewpoints on this ever-interesting subject. He invites your comments on this subject, as do I. John can be reached at jmc-lmc@pacbell.net. Thanks for the very interesting article, John! -- J. Franklin Campbell |
| By John P. McCumiskey
Such is the comment frequently heard from the lips of players ever concerned about their chess ratings. For many, this is a number that never rises fast enough and always seems to drop far too quickly. It would seem, for these players, their rating is the end-all, be-all of chess. Achieving a high rating is their "reward" or is "deserved" for playing "good" chess. Ratings, however high or low, are not a reward. I believe one's rating is a number that shows the overall performance of one player versus others over time and is calculated using a complicated statistical formula. I recently followed, with great amusement, a rec.games.chess.misc usenet thread dealing with a high rating. Someone started it by complaining of an on-line player with an extraordinarily high rating (well in excess of 2700). It seems this high rated player was able to achieve this level of performance by following a simple formula: Only play those rated at least 200 points higher than his rating. As I interpreted the original post, this was an unfair way to obtain such a high rating. In the world of on-line play, it is possible to manipulate your rating upward. On the opposite end of the spectrum are the "sand baggers," those who intentionally lose games to keep their rating from getting too high. Organizations such as the US Chess Federation (USCF) and Continental Chess try to combat such practices with rating floors and minimum assigned ratings based on previous performances. Sand bagging is an appalling practice, but difficult to prove and/or control.
Fortunately, the vast majority of chess players do not fall into either of these categories. They play chess for fun and try to improve their rating. Within a chess organization, ratings are a necessary "evil", but they are more difficult to manipulate. An attentive tournament director or tournament administrator will detect sand bagging. Participation in correspondence/e-mail and over-the-board tournaments do not allow a person to pick and choose opponents based on personal criteria. In correspondence/e-mail play, the tournament administrator sends you a section assignment for your event. In over-the-board play, the tournament director provides round by round pairings. This means you will have games with players both higher and lower rated. Over time, this gives you a fairly accurate assessment of your level of play. An example of an organization that makes good use of its rating system is American Postal Chess Tournaments (APCT). APCT does run some events without rating restriction, but also has several events restricted to players based on their rating. These events include a class championship, a "Semi-Class" event where players with similar ratings are paired together, as well as other events that are restricted to players above or below a specific rating. Participating in these events has worked out very nicely during my APCT membership. My rating has the primary function of telling me what my overall performance is, based on all the games I have played. If my rating goes up, it can be assumed that my chess play has been successful; if my rating goes down, it can be assumed I've been doing something wrong. In either case, I have the incentive to work to improve my play! Secondary to giving me feedback on my performance, I use my rating to set goals for where I want my level of chess play to be. My highest USCF rating was 2097 and APCT rating 2004. Today, those ratings are 1976 and 1870, respectively. In each of these organizations, I have set a goal: In USCF, the goal is to get my rating over 2200; in APCT, my goal is to qualify for the US Correspondence Chess Championship preliminaries, which will require a rating 2000 or above. In both cases, hard work will be required to achieve the desired ratings of my overall chess-playing performance. One example of how I have used ratings to set goals occurred while I was playing chess in Alaska. For many years, the Alaska State Chess Championship was a 10 player round-robin, with invitations being sent to the defending champion and the nine other highest rated players. In January 1980, I decided I really wanted to earn an invitation, but my rating was only 1700 (a rating of 1850 took the last invitation in 1979). I had to pick the rating level that I believed would guarantee an invitation into the tournament for me. The goal was 1875. By the time the event arrived in November, I had achieved a rating of 1832, short of my goal, but it was good enough in 1980 to receive my first invitation to Alaska State Chess Championship. A chess rating will have different meanings for each individual: something earned or deserved, a reward, a gift, etc. Always remember, a true rating, based on your results over time against many players, is a fair assessment of one's overall performance. Hmmm ... I see the August 1998 USCF rating supplement. What, my rating's gone down!! ;-) If you have any comments about this article, I can be reached via e-mail at jmc-lmc@pacbell.net. All comments are appreciated. |