Author: Harald Faber
Date: 13:38:57 02/03/05
Go up one level in this thread
On February 03, 2005 at 09:45:26, Joshua Haglund wrote: >On February 03, 2005 at 02:50:57, Harald Faber wrote: > >>On February 02, 2005 at 12:37:37, Joshua Haglund wrote: >> >>>On February 02, 2005 at 10:43:37, gerold daniels wrote: >>> >>>>when tony tests it than all will know how good the new shedder 9 is. >>>> >>>>gerold. >>> >>>When tony tests it, it'll be done on obsolete hardware, at one time control, >>>with probably less than 1k games. Leave it to the online chess servers to test >>>it :) >> >> >>Totally nonsense. On the servers people play with different machines and their >>own so-called "tuned" books, sometimes even with own settings. It often occurs >>that their "tuning" has an opposite effect and hurts the engine's play and thus >>the results. So how can you say that such results have more weight than playing >>on equal hardware and with the delivered opening book? >> > >That's why I give chess engines an ELO RANGE. It covers the wide variety of time >controls, hardware, books and settings, Not just one setting, long time controls >on slow hardware. So you measure everything else but not what the out-of-the-box product is able to play. >>>When it has 5000 games it'll be ELO List Update, which will be a day or two >>>after it comes out. >> >> >>Wasted time. >> > >The more games played the more accurate the results. Waiting 1 or 2 days isn't a >waist of time to see 5000 games played at Playchess.com versus the mere handful >played. You did not understand my point. Forget it. >Joshua Haglund >toneewa@yahoo.com
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.