Author: Sune Fischer
Date: 11:30:21 07/19/02
Go up one level in this thread
On July 19, 2002 at 14:16:34, Matthew Hull wrote:
>On July 19, 2002 at 14:03:24, Sune Fischer wrote:
>
>>On July 19, 2002 at 13:39:32, Matthew Hull wrote:
>>
>>>On July 19, 2002 at 09:54:08, Uri Blass wrote:
>>>
>>>>world championship.
>>>>
>>>>I suggest that the rules can say that the games are
>>>>played under winboard(pondering off,animation off).
>>>>
>>>>Every 2 programs can play 100 or even more games between
>>>>them so the total number of games of every program can be
>>>>at least 5000.
>>>>
>>>>I believe that we may get significant results
>>>>by that idea.
>>>>
>>>>Uri
>>>
>>>I don't suppose we would see many "evergreen" games from such a contest. The
>>>quality of the chess would be quite low. For that kind of entertainment, an
>>>elemetary school chess tournament might suffice. Some might say it would be
>>>more interesting to watch people race their electric belt sanders.
>>>
>>>To me, the compelling attraction of computerchess is its potential to some day
>>>definitively define what Lasker called "the reason of chess", or at least make
>>>some asymtotic approach to that answer.
>>>
>>>Bullit chess is an idea seeminly calculated to suck the very life out of a game
>>>which is loved for its beauty and complexity.
>>>
>>>As much fun as blitz can be personally, it's not real chess in the classic
>>>intellectual sense. I think computerchess can offer more than what an extreme
>>>bullit event proposed here can offer.
>>>
>>>Just my opinion. ;-)
>>>
>>>Regards,
>>
>>No, you are right it isn't chess in the classic sense, 1 sec/G they move so fast
>>you have no idea who is being checkmated or what is going on :)
>>
>>But, you will be surprised that something interesting can actually develop in
>>such games. E.g. things that are found by static eval need no search at all.
>
>I suppose this could be a high motive for the contest. But I still think they
>just want to "race their electric belt sanders". :-)
>
>>
>>Here is a "nice" 1 sec game selfplayed by my engine, I like the opening for
>>white, it sort of reveals the opening code I have (no book was used) :)
>>
>>1. d4 d6 2. e4 Nd7 3. Nf3 Ngf6 4. Nc3 e6 5. Bf4 Qe7 6. Bc4 c6 7. O-O a6 8.
>>Re1 a5 9. e5 dxe5 10. dxe5 Ng4 11. Qd6 Qd8 12. Qd4 Bc5 13. Qd2 b6 14. Bg5
>>Qc7 15. Nd4 a4 16. a3 b5 17. b3 bxc4 18. bxa4 h6 19. Bf4 h5 20. h3 f6 21.
>>exf6 Nde5 22. fxg7 Qxg7 23. hxg4 Nxg4 24. Nxe6 Bxf2+ 25. Qxf2 Qxc3 26. Qd4
>>Rxa4 27. Be5 Qxc2 28. Bxh8 c3 29. Nc5+ Qe4 30. Qxc3 Nf2 31. Rab1 Bh3 32. g3
>>Bd7 33. g4 Bxg4 34. Ra1 Nh3+ 35. Kh2 Rc4 36. Qxc4 Qxe1 37. Rxe1+ Kf8 38.
>>Bc3 Ng5 39. Qf4+ Nf7 40. Qb8+ Bc8 41. Qxc8+ Nd8 42. Qxd8+ Kf7 43. Qe8#
>>{White mates} 1-0
>>
>>-S.
>
>Nice. Now at the end of the contest, you can wade through the 5000 odd results
>to find the gems, eh?
No, I only played three :)
The first was lost on time after 83 moves, the second did mate but was "not a
good example" :)
Anyway, I think this clears up a few things about the static eval.
It shows you what things are found by search and what by evaluation.
For instance you capture a knight on c6 and how do you recapture, do you use the
bishop on d7 or do you create a nasty double pawn by recapturing with the
b-pawn?
If you let it search, it will probably find the right move, but it should get
the rudimentary ideas correct at the first ply!
I have also begun printing out the static eval after setting up a position, I
found a horrible bug in my evaluation (causing up to 1.5 pawn miseval!).
It is just another powertool in the toolbox :)
-S.
>Regards,
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.