Author: Chris Carson
Date: 12:48:24 05/13/01
Go up one level in this thread
Joshua,
Just FYI, I am not with the SSDF.
There are more than enough games for statistical analysis (183 is more
than enough). The broad range of results is part of the statistical
analysis. You will find this also in human ratings.
Example:
TWIC 339: http://www.chesscenter.com/twic/twic339.html
3rd Julian Borowski A Essen GER (GER), 2-11 v 2001 cat. XV (2616)
---------------------------------------------------------------------
1 2 3 4 5 6 7 8 9 0
---------------------------------------------------------------------
1. Kasimdzhanov, Rustam g UZB 2693 * . = . . = 1 1 1 . 4.0 2832 +139
2. Jussupow, Artur g GER 2645 . * . = = 1 . = . = 3.0 2671
3. Lutz, Christopher g GER 2614 = . * 0 = 1 . . . 1 3.0 2703
4. Romanishin, Oleg M g UKR 2559 . = 1 * 0 . . = . 1 3.0 2690
5. Gurevich, Mikhail g BEL 2688 . = = 1 * . . = 0 . 2.5 2586
6. Sutovsky, Emil g ISR 2604 = 0 0 . . * 1 . 1 . 2.5 2632
7. Dautov, Rustem g GER 2631 0 . . . . 0 * = 1 1 2.5 2604
8. Bischoff, Klaus g GER 2533 0 = . = = . = * . . 2.0 2571
9. Nijboer, Friso g NED 2580 0 . . . 1 0 0 . * = 1.5 2496
10. Smagin, Sergey g RUS 2613 . = 0 0 . . 0 . = * 1.0 2365 -248
---------------------------------------------------------------------
Also, I do not claim a perfect system. Stastistics is
probability, not exact, there is a range of expected results.
Finally, the SSDF shows 337 human vs computer calibration games at
40/2 (see tony's page) and I have collected another 183+ games that is
well over the arbitrary number you picked of 500 (337+183=520).
Tony' Page (SSDF calibration and my list with 183+ games):
http://home.interact.se/~w100107/welcome.htm
However you are free to believe what you wish. :)
Best Regards,
Chris Carson
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.