Jeff Sonas is a statistical chess analyst who invented the Chessmetrics system for rating chess players, which is intended as an improvement on the Elo rating system. He is the founder and proprietor of the Chessmetrics.com website, which gives Sonas' calculations of the ratings of current players and historical ratings going back as far as January 1843.[1] Sonas has written dozens of articles since 1999 for ChessBase.com and other chess websites.[2] He was a participant in the FIDE ratings committee meeting in Athens, Greece in June 2010.
In 2002 he wrote for chessbase.com that a 24 k-factor would be ideal for predicting game results, instead of the standard 10 for high level players.[3] In 2009 John Nunn wrote he needed more proof for the change.[4] Sonas replied saying Nunn didn't thoroughly read his first article, but also says there is room for improvement, as he didn't analyze what it should be for low level players.[5]
In 2011 after analyzing 1.5 million FIDE rated games, he demonstrated according to the Elo formula, two players having a rating difference of X, actually have a true difference more like X(5/6). This means players are consistently losing points for winning, and winning points for losing, since the Elo formula is incorrectly predicting results. He also says the 400 point rule should be raised to 700-900, or be gotten rid of entirely. The rule says that players with an Elo score more than 400 points apart will be treated as though they are only 400 points apart.[6]
Sonas graduated with honors with a B.S. in Mathematical and Computational Sciences from Stanford University in 1991.
References
edit- ^ January 1843 rating list. Retrieved on 2008-12-19.
- ^ Man vs machine - who is winning? (October 8, 2003). Retrieved on 2008-12-19.
- ^ "The Sonas Rating Formula – Better than Elo?". 22 October 2002.
- ^ "Nunn on the K-factor: Show me the proof!". 30 April 2009.
- ^ "Rating debate: Is 24 the ideal K-factor?". 3 May 2009.
- ^ "The Elo rating system – correcting the expectancy tables". 30 March 2011.