Author: Albert Silver
Date: 13:27:38 12/09/04
Go up one level in this thread
On December 09, 2004 at 14:07:04, Murat wrote:
>
>
>White: Fritz 6 on a fixed depth of 6 plys and selective option in the engine
>parameters changed to 0 from the defoult 2.
I really don't see what this proves. Suppose you take the Mephisto Genius and
run it at 6 plies, do you think they will be examining the same depth? Not in a
million years as Lang's programs were highly selective. The fact is that if you
were able to speed up the Mach III to the speed of a P90 and play it on equal
hardware so to speak, it would most likely get *slaughtered* by Fritz 6. One
need only look at the SSDF list and see the huge disparity of ratings on
identical hardware of older and newer programs.
Let's take Shredder 7 for example. On a K6/450 it is rated 2699. This rating
represents its ability to beat other programs, which is exactly the subject of
your post. Mind you, it is probably rated a good 30 points less than the 7.04.
Well, I can see the culmination of Lang's programs on this list, the Genius 6.5,
also tested on the K6/450. It is rated 2478, or well over 200 points less. The
Mach III was far less developed, so it is safe to assume that had it somehow
been ported to the PC as was, it would do a lot worse.
Albert
>I did take Fritz out of book after
>move 17 since the Mach III run out of its book at that point.
>
>Black: Chess Challenger Excel68000 Mach III at level G6 which is fixed at 6ply.
>
>[Event "Fixed level at 6 ply"]
>[Date "2004.12.09"]
>[White "Fritz 6"]
>[Black "ChessChallengerMachIII"]
>[Result "0-1"]
>[ECO "B90"]
>
>1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Be3 e5 7. Nb3 Be6 8. f3
>d5 9. exd5 Bxd5 10. Nxd5 Nxd5 11. Qd3 Nc6 12. Rd1 Nxe3 13. Qxe3 Bd6 14. Bd3
>Qh4+ 15. g3 Qe7 16. O-O Nb4 17. Bf5 O-O 18. c3 g6 19. f4 gxf5 20. cxb4 e4 21.
>Nd4 Bxb4 22. Nxf5 Qc5 23. Nd4 Qe7 24. a3 Bc5 25. b4 Bb6 26. Rf2 Rac8 27. f5 Qe5
>28. f6 Qh5 29. g4 Qg6 30. g5 Rfd8 31. Rfd2 Qh5 32. Re1 Qg4+ 33. Kf1 Rxd4 34.
>Rxd4 Rc2 35. Rd8+ Bxd8 36. Qg3 Qf5+ 37. Kg1 Bb6+ 38. Kh1 Bc7 39. Rd1 Rxh2+ 40.
>Kg1 Rh3 41. Qxh3 Qxh3 0-1
>
>I think the high ratings today's programs enjoy are 80% due to the advances of
>hardware and 20% in actual programming improvement.
>
>I must say bravo Dan&Kate Spracklen, hell of a job.
>
>All the best
>
>Murat...
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.