Author: Uri Blass
Date: 15:09:27 02/28/04
Go up one level in this thread
On February 28, 2004 at 16:41:25, Luis Smith wrote: >On February 28, 2004 at 16:14:48, Uri Blass wrote: > >>The last results of movei00_8_174 against Junior5 were positive and movei seem >>to be able to beat Junior5 with time handicap of 3:1 in the nunn2 match at 90 >>minutes against 30 minutes(ponder off) and that seems to be a good reason for a >>new release. >> >>00_8_174 has result of 15-9 in the last 24 nunn match games against Junior5(it >>is 26-24 for movei if I combine this result with result of 008_168a and 008_170 >>that have a bug of using swap_list[-1]) >> >>Some known bugs or disadvantages of 00_8_174 >> >>1)Movei00_8_174 can fail to deliver the fastest mate because as soon as it see a >>mate it may prune other mates but I do not know of cases of missing a win >>because of that problem. >>I also do not know of very long games like smarthink's games because of that >>problem. >> >>2)Movei00_8_174 does not use tablebases except internal kpk tablebases of Dieter >>that were posted here with permission of Dieter. >> >>3)Movei00_8_174 has a very small book and the book still does not know about >>tranposition. >> >>4)Movei is using logfile with the name moveiloginfo143.log(the name is based on >>movei00_8_143 and I was too lazy to change it) >> >> >>I did some small changes in movei from the last version that are probably not >>changes in the engine unless I have some bugs but I assume that every change may >>do the program weaker because of some bug so I plan to release 00_8_174 and not >>a version with small changes like changing the name of the logfile and having >>special functions to write to logfile. >> >>I plan to write some information in hebrew and english about movei and send it >>as a zipped file to Leo who is going to release it. >> >>I am not sure if I am going to finish writing information about it today and >>even if I finish it then I am still not sure if Leo is going to release it in >>his page in the next few hours but you can expect it soon. >> >>Uri > >Hi Uri, > > This is good news. Preliminary test results at g/15 AMD XP 2200+ 64 MB hash >for all engines. > >Movei AMD XP 2200+ 384 MB RAM 2004 > >Movei 00_8_174 - Amateur 2.70 3.0 - 5.0 +2/-4/=2 37.50% >Movei 00_8_174 - Frenzee 146 6.0 - 2.0 +5/-1/=2 75.00% >Movei 00_8_174 - SlowChess 2.9 4.5 - 3.5 +4/-3/=1 56.25% >Movei 00_8_174 - Tao 5.4 2.0 - 6.0 +1/-5/=2 25.00% >Movei 00_8_174 - Quark 2.05b 5.0 - 3.0 +4/-2/=2 62.50% >Movei 00_8_174 - Hagrid 0.7.56 8.0 - 0.0 +8/-0/=0 100.00% >Movei 00_8_174 - Movei 00_8_173 3.0 - 5.0 +0/-2/=6 37.50% > >Hagrid is probably not a good opponent to test against Movei as it is a bit weak >but I wanted to include a wide range of engines, and also Quark wasn't using its >own book. > >Movei00_8_174 vs. Junior 5 at g/5 with no handicap resulted in 1 - 8 in favor of >Junior and Movei only managed 2 draws so I stopped the tournament. They were >both using Perfect5.ctg from Sedat's website. Thanks for your results. Note that the only difference between 173 and 174 is removing a bug in swap_list. I guess that 173 was lucky to get 5-3 against 174 and I am not considering going back to 173. 8-0 against hagrid is a very good result Hagrid is weak but not so weak to expect this result. On the other hand losing 5-3 against amateur is a small disappointment. Uri
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.