Author: Walter Koroljow
Date: 10:44:28 07/29/99
Go up one level in this thread
On July 29, 1999 at 10:48:52, KarinsDad wrote:
>On July 28, 1999 at 14:52:58, Dann Corbit wrote:
>
>>On July 28, 1999 at 10:09:08, Paulo Soares wrote:
>>[snip]
>>>Laurence, a 1700 player can win a 2200 player, this is a statisc
>>>question. The problem is: what the probalility of a 1700 player
>>>win a 2200 player?
>>win expectancy = 0.0532402152020224 {which includes wins+draws}
>>On average, in a 100 game match, the 1700 player would score 5 points.
>>
>>Don't forget that Kasparov, Fischer, etc. were all counted as 1700 players at
>>some point (admittedly, not for long). These things factor into the statistics
>>also.
>
>Dann,
>
>Have you done a win expectancy for a 2500 player to a 2800 player?
>
>The reason I ask the question is that vitor pointed out elsewhere in this thread
>that once you get to a certain rating, you only win or lose 1/2 the rating
>points per win/loss. To me, this is an aberration in the formula which would
>result in a different win expectancy between a 2500 player and a 2800 player
>than between a 1500 player and a 1800 player (i.e. the equation should be
>different).
>
>Is this the case? And if so, what would be the win expectancy? Would the win
>expectancy between a 2500 player and a 2800 player be equivalent to a 600 spread
>in the lower ratings?
>
>KarinsDad :)
KarinsDad,
The win expectancy is identical for a 300 point difference no matter where it
occurs in in the rating scale. The different "k-factors" control the rate at
which the ratings follows results but they do not control the ultimate values
the ratings attain.
More specifically, the ratings for people below 2100 USCF are approximately the
average of their performance rating for their last 25 games. For people in the
2100-2300 bracket, they are the average of the last 36 games, and for people
above 2300, they are the average of the last 50 games.
Cheers,
Walter
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.