Author: Vincent Diepeveen
Date: 11:03:55 08/30/99
Go up one level in this thread
On August 29, 1999 at 01:12:53, Ed Schröder wrote: >On August 28, 1999 at 16:13:27, Robert Hyatt wrote: > >>On August 28, 1999 at 11:21:22, Frank Quisinsky wrote: >> >>>Hello, >>> >>>the result from the WT-5 tournament from Kai Skibbe, Christian Koch and me ! >>> >>>40 moves in 40 minutes, >>>AMD K6-2 333 MHz, Pentium III 504 MHz and AMD K6-3 450 MHz, >>>~ 32 MB for Hash-Tables, >>>Ponder = Off >>> >>>Crafty, Voyager, Comet, Patzer, Bringer using 4-pieces Nalimov Table-Base >>>Gromit and AnMon using 4-pieces Edwards Table-Bases ! >>> >>>IMPORTANT NOTE >>>Zarkov, Nimzo, Voyager, LGoliath, Patzer, Francesca und LambChop >>>are versions of programs in development and have been offered by >>>authors to Kai, Chritian and Frank only for the explicit purpose >>>of organizing this tournament. THEY ARE NOT PUBLICLY AVAILABLE ! >>> >>>01. Nimzo 2000a 102,0/144 70,83% 2572 ELO >>>02. Zarkov 4.5h 95,5/144 66,32% 2538 ELO >>>03. Crafty 16.11-16.15 92,0/144 63,89% 2521 ELO >>>04. Voyager 5.08-6.01 89,5/144 62,15% 2510 ELO >>>05. LGoliath 2000 75,5/144 52,43% 2445 ELO >>>06. Comet B.04-B.05 75,0/144 52,08% 2444 ELO >>>07. AnMon 4.26-4.27 70,0/144 48,61% 2421 ELO >>>08. Patzer 3.0 64,0/144 44,44% 2394 ELO >>>09. Phalanx 21 60,0/144 41,67% 2375 ELO >>>10. Francesca 0.68d-0.70 58,5/144 40,63% 2369 ELO >>>11. Gromit 2.20 58,0/144 40,28% 2367 ELO >>>12. LambChop 8.2 51,0/144 35,42% 2333 ELO >>>13. Bringer 1.4-1.5 45,0/144 31,25% 2303 ELO >>> >>>You can download the 936 games under ... >>>http://members.xoom.com/VolkerPi/ >>> >>>Nimzo 2000a play very strong and this version is more than 100 ELO better than >>>Nimzo 2000 ! Voyager win the matches against Nimzo 2000a, Zarkov 4.5h and Crafty >>>16.15 with 6,5 : 5,5 lose against Gromit with 5,0 : 7,0 ! >>> >>>LambChop 8.2 play more than 100 ELO better than LambChop 7.1, this is very >>>great, LambChop play an interesting chess, I like this program (and AnMon :-) >>>and the others) >>> >>>For the new chess program Bringer ist this result sensational ! LGoliath 2000 >>>play ca. 75-100 ELO better than LGoliath Gold 2.05 ! >>> >>>From Francesca we become in the last round an update. This new Francesca version >>>play stronger. OK, when we play the next tournament, but not in the next 3-4 >>>months I think ! >>> >>>The Crafty versions play an fantastic chess and I can not understand that Dr. >>>Robert Hyatt say that the time managment from Crafty is bad when Crafty play on >>>one machine. I can not see that. Crafty play under Winboard (without ponder) >>>matches with a good time control and blitz not the last moves before the time >>>control ending !!! >> >> >>that is because you don't know the program. I disable a good bit of the >>time management code when pondering is disabled. But it _definitely_ hurts >>it as it now doesn't assume it will save any time, which is not the way to >>play chess... IE as a human I use more time when the game is complex, near >>the beginning after I am out of my 'book'. Ponder=off disables this, and I >>consider this bad, whether it plays well for you or not. >> >>It is simply not the way to play matches... Because that is not the way the >>program is developed, nor is it the way it is tested. And no it doesn't have to >>blitz the last few moves, because I modified this several versions ago. But >>just because it doesn't get low on time, doesn't imply that it is using its >>time _correctly_. It definitely isn't... > >We (the programmers) can argue what we want Bob but this is a lost case >on before hand because the formula of playing 2 programs on one machine >is too good to be true. People are not going to give this up. > >Same story as with book-learning, it hides the real strength of a chess >engine. Still people take the numbers for real. Another lost case :-) > >Ed Correct Ed, A very fine feature of winboard version of DIEP is that it eats up always 100% of the systemtime, so when starting both engines and both in book: diep eating up 98%, other program fighting to slowly get 50% instead of 100%, as when diep is first out of book it usual gets 98-99% of the system time as the other prog which still is in book is idling.... ...fair way of matching other compu's... Note that Zarkov here comes out as best prog in all winboard matches i did (last months i didn't do anymore). Weird that nimzo is above it, probably the fast level that's getting played (30 seconds a move an engine). I played at a PII450 dual a level of 40 in 2. Would like to see how nimzo is doing at a somewhat slower level... Greetings, Vincent
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.