10 Feb '06 14:35>1 edit
Originally posted by masscatWhen these stats were true (if they were), USCF ratings were terribly inflated. Arpad Elo's theory upon which the rating system is built, would have 1500 at the 50th percentile. According to the current distribution chart, the median is closer to 1000, and just under 1400 if you exclude scholastic members. You have it somewhere close to 1700 it seems.
Rough USCF rating breakdown by percentile:
2400+ 0.4%
Masters 2.6%
Experts 10%
Class A 27%
Class B 55%
Class C 81%
Class D 95%
Class E 98%
e.g. an 1800 rating puts him in the top 27% of the rated players in the country. These stats are at least 20 years old, but I doubt they've changed much.
The influx of scholastic members has driven the ratings down in recent years, as well as making adult events more dangerous. Many youth are underrated, and a 1200 kid in an adult tournament is often, in truth, a 1500 player in actual strength. Consequently, the kids wins most of the money for biggest upset.
BTW, the figures you reported would be "percentile" if you subtracted each from 100 and typed the answer. 0.4% represents the proportion of players that have reached that rating or higher, As a percentile, this figure would be expressed as 99.6%, which is fairly close to the 99.58% presented in the USCF 2002 Ratings Distribution Chart for non-scholastic members. http://www.uschess.org/ratings/ratedist.html