Computer Chess Club Archives


Search

Terms

Messages

Subject: My CSTal plays different! Why ?? What is wrong with the autoplayer ?

Author: Thorsten Czub

Date: 13:13:46 02/09/00

Go up one level in this thread


Hi Harald,
i have replayed your autoplayer-games with my machine. you know that we
benchmarked and
your machine
is a little slower.
I have some strange moves, i cannot reproduce.
please : i don't say YOU have made anything wrong. i am just trying to find out
whats going on.
thanks.

[Event "Test"]
[Site "Haltern"]
[Date "2000.02.07"]
[Round "11"]
[White "Shredder 4.00"]
[Black "CSTal 2.03"]
[Result "1-0"]
[ECO "E05"]
[PlyCount "95"]
[EventDate "2000.02.??"]

1. d4 {Buch 0s} 1... Nf6 {6s} 2. c4 {Buch 0s} 2... e6 {1s} 3. Nf3 {Buch 0s} 3...
d5 {2s} 4. g3 {Buch
0s} 4... Be7 {1s} 5. Bg2 {Buch 0s} 5... O-O {1s} 6. O-O {Buch 0s} 6... dxc4 {1s}
7. Qc2 {Buch 0s}
7... a6 {1s} 8. a4 {Buch 0s} 8... Bd7 {1s} 9. Qxc4 {Buch 0s} 9... Bc6 {1s} 10.
Bg5 {Buch 0s} 10...
a5 {2s} 11. Nc3 {Buch 0s} 11... Na6 {1s} 12. Bxf6 {Buch 0s} 12... Bxf6 {1s} 13.
e4 {Buch 0s }
13... Nb4 {1s} 14. Rfd1 {Buch 0s} 14... b6 {1s} 15. Rac1 {Buch 0s} 15... Bb7
{1s} 16. Qe2 {Buch
0s} 16... Rc8 {6:20m} 17. Qe3 {-0.14/11 3:57m} 17... Kh8 { 8:51m (c6)} 18. Bf1
{+0.01/11 3:40m}
18... Be7 {13:36m (c6)} 19. Bc4 { -0.02/11 3:50m}
19... Ra8
{I have problems to reproduce this move. my cstal had Bc4 in permanent brain. i
waited 3'50"
but the move did not appear on my machine, my cstal wanted to play
d9, 71s, -0.02 , (f1c4) d8e8 c3b5 c7c6 b5c3 c8d8 e3f4 b7a6 c4a6 b4a6
d10, 346s, -0.22 (f1c4) d8e8 c3b5 e8d8 f3e5 c7c6 b5c3 e7g5 f2f4 g5f6 e3f2 c8c7
h2h3 f6e5
please - i don't say anything is wrong with your data, i do only say: i have
problems to get this
move on my machine.
strange... let's see if i find other  moves. and when i find more of these moves
i think we have
the strangeness.
and when my version plays different without autoplayer, i could play from THIS
position but now
moving Qe8 instead of  Ra8.
1s (c6)} 20. Na2 {+0.10/10 3:00m} 20... f5 { 10:13m (Dd6)} 21. exf5 {+0.07/11
4:38m} 21... Rxf5
{11:19m} 22. Ne1 { +0.08/11 0s} 22... Rf6 {7:34m (Ld5)} 23. Nxb4 {+0.21/11
4:29m} 23... axb4 {
5:04m (Lxb4)} 24. Bxe6 {+0.46/11 2:32m} 24... Bd6 {1:56m (Txa4)} 25. Nd3 {
+0.39/11 5:35m}
25... Qf8 {4:16m (De8)} 26. Bc4 {+0.54/11 3:41m} 26... Rf3 { 3:47m (Te8)} 27.
Qe2 {+0.73/10
3:30m} 27... Re8 {1s (Tf5)}

{1 second ??? so Qe2 should have been in the pb ?? strange , my machine has Qd2
in PB, not
Qe2. Again a different PB behaviour. all main-lines during Rf3 computation had
second half-
move Qd2 and nowhere Qe2. strange. strange.
on qe2 - i have to give this move manually and cstal has no pb since it computed
about Qd2, my
cstal has always
Rf5 (from 0 seconds to 124) and in no main line Re8. So what has happened there
that cstal
played Re8 ? again a pb move that cannot be reproduced. harald, there is
something wrong with
the pb-behaviour in autoplay-mode !}

 28. Ne5 { +0.79/11 4:02m} 28... Re7 {1s (Lxe5)} 29. Rd3 {+1.23/10 3:00m} 29...
Rf5 { 2:43m
(Txd3)} 30. Re3 {+1.45/11 3:06m} 30... Bxe5 {1s} 31. dxe5 {+1.63/12 2:23m }
31... Bf3 {2:49m}
32. Qe1 {+1.55/12 2:12m} 32... c5 {5:09m} 33. e6 { +1.62/11 0s} 33... Qe8
{4:39m} 34. Qd2
{+1.91/11 0s} 34... Qb8

{You can imagine, i have problems to reproduce this move. Mine wants to play
Bc6, later Ba8,
but never in all the main lines it wants to play Qb8, which looks very weird to
me:
d8, 58s, -1.56 f3a8 d2d6 e8c6 ...
d8, 68s, -1.21 f3c6 e3d3 e8h5 d3d8 ...
d9, 135s, -1.35 f3c6 e3d3 f5f8 d3d6 ...
d10, 421s, -1.74 f3c6 d2d6 h7h6 c4b5 ...
d11, 578s, -1.56 f3a8 d2d6 e8c6 d6c6 ...

{6:25m (Lc6)} 35. Qd3 {+2.81/12 2:53m} 35... Qf8 {7:34m} 36. Qd6 {+3.14/11 0s}
36... Bh5 { 4:16m
(Te8)} 37. Rc2 {+3.64/11 3:57m} 37... Rf3 {1s (Kg8)} 38. Rxf3 { +3.73/12 2:43m}
38... Bxf3
{3:49m} 39. Rd2 {+3.90/12 0s} 39... h6 {7:34m} 40. Qxb6 {+4.00/12 0s} 40... Re8
{4:29m} 41. Qc7
{+4.23/11 0s} 41... Bg4 { 6:15m (Df5)} 42. Bb5 {+5.06/12 3:34m} 42... Qe7 {57s}
43. Qxe7
{+5.13/13 2:56m} 43... Rxe7 {3:09m} 44. Rd8+ {+5.55/15 0s} 44... Kh7 {2:04m} 45.
Rd7 { +5.84/16
1:50m} 45... Rxe6 {59s} 46. Bd3+ {+5.76/15 1:32m} 46... Kg8 {1:35m} 47. Bc4
{+6.16/16 33s}
47... Kf8 {2:47m (Kh7)} 48. Bxe6 {+7.03/16 2:25m} 1-0

ok, i have seen enough to get an idea.
you asked me about the very short game

[Event "180 Minuten/Partie"]
[Site "Auto232, Computerspeed = 511"]
[Date "2000.02.07"]
[Round "1"]
[White "Shredder 4.00"]
[Black "CSTal 2.03"]
[Result "1-0"]

1. e4 {Buch 0s} c5 {1s (Sc6)} 2. Nf3 {Buch 0s} d6 {1s}
3. d4 {Buch 0s} cxd4 {1s} 4. Nxd4 {Buch 0s} Nf6 {4s} 5. Nc3
{Buch 0s} a6 {2s} 6. Be3 {Buch 0s} e6 {1s} 7. f3 {Buch 0s}
b5 {1s} 8. Qd2 {Buch 0s} Nbd7 {1s} 9. g4 {Buch 0s} h6 {1s}
10. O-O-O {Buch 0s} Bb7 {1s} 11. Bd3 {Buch 0s} Ne5 {1s}
12. h4 {Buch 0s} b4 {1s} 13. Nce2 {Buch 0s} d5 {1s}
14. exd5 {Buch 0s} Nxd5 {1s} 15. Kb1 {-0.31/9 6:26m} Rc8
{11:52m
ok, you can guess it, i cannot reproduce Rc8. mine plays b3.
rc8 is in the main-line only at search 5 after 1 second and refuted
after 1 second and never occurs. strange. how can this be ? Maybe your
hashtables were to big and yours was swapping on the hd ? anyway,
b3 gets not refuted within 15 minutes.}
16. Nf4 {-0.18/10 25s} Nf6 {16:05m
Is cstal really black ? My version swaps between: Nxf4 / Qa5 /Qb6 and
will never play Nf6 ! I have waited 1451 seconds, that is arround 24 minutes.
Mine wantes to play Qb6 from 222s to 1451s.

(Sxf4)}
17. Rhe1 {+0.80/9 5:41m} Nfd7 {7:19m} 18. Ndxe6 {+1.58/11
49s} Qe7 {5:45m (fxe6)} 19. Bd4 {+6.22/10 3:54m} Nxf3 {1s}
20. Nxg7+ {+7.18/11 3:42m} 1-0

When i can make sure you use the same version, my cstal2.03 is from
24.05.99, 14h46', and it is 3143168 KBytes, my chess.sty is from
24.09.97 12h05' and is 156 bytes,
and we have made sure your pc is as fast as mine (yours is in fact a little
slower..)
i wonder why we have those different results.

the only different thing i see is:

you have played by autoplayer,
i have played manually !

So far to the autoplayer discussion, or any other points ??

Maybe we come nearer to the question:
why do some peole have so strange (autoplayer) results, not beeing
able to reproduce with manually outplayed software.

Any ideas or remarks from others ?

How can we trust that anything goes right while autoplaying ?
I do not trust.




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.