Author: Mike S.
Date: 19:04:24 12/06/02
Go up one level in this thread
On December 06, 2002 at 19:24:29, Bertil Eklund wrote:
>(...)
>You missed one point, when it sees itself in trouble it plays "Hiarcs-wise" and
>use almost all its time to survive, if it does it only has seconds left for the
>game or time-control.
I had checked for that, and indeed it used more time in a few games when it had
a bad position (but there were also at least one game when it had a bad position
and still kept ~1/3 reserve). Btw. (just in case you didn't it already anyway)
when you set the original time control of that tournament in Fritz 7 (40/1h +
20/30:00 + 20:00), then you can see the original clock standings of each game,
move by move, with the cbv data. I discovered that just recently when analysing
those games.
>Was "Novemberturnier" played with its own book? What do you think of the
>Ruffian-book?
You point to a small mistake I made during the setup for that tournament:
Mainly, it was meant to be played with the 6-plies "Kurzbuch.ctg" by all
engines, but Ruffian-UCI used it's own book in addition to that. 0...5 openings
moves came from that book after the Kurzbuch's moves, but I think it was not
decisive.
While some people have critizised the Ruffian book as a weak point, I had a very
good impression of it in a match against Deep Shredder 5, which used a very
large Necchi book. I didn't see much of an book advantage by Shredder in that
match. IIRC, I read somewhere no engine specific editing has been done, only
generated from games, but I think it's ok. I can only guess that the Ruffian
book may have weaknesses in side-variants and seldom openings, if any (?). It is
possible that "holes" became visible after a larger number of games than I
viewed.
40/40 P3/700 MHz rapid 2002
123456789012
1 Deep Shredder 5 101101½½0½½½ 6.5/12
2 Ruffian 1.0.0 010010½½1½½½ 5.5/12
In the following game, Ruffian's book varaint was even longer than the one from
Shredder's Necchi book. Shredder's last book move was 12...Bf5, Ruffians last
was 15.Rad1 (but that was an exception of course):
[Event "40/40"]
[Site "P3/700 MHz"]
[Date "2002.09.25"]
[Round "10"]
[White "Ruffian 1.0.0"]
[Black "Deep Shredder 5"]
[Result "1/2-1/2"]
[ECO "D20"]
[PlyCount "73"]
[EventDate "2002.??.??"]
1. d4 d5 2. c4 dxc4 3. e4 e5 4. Nf3 exd4 5. Bxc4 Bb4+ 6. Nbd2 Nc6 7. O-O Nf6 8.
e5 Nd5 9. Nb3 Nb6 10. Bg5 Be7 11. Bxe7 Nxe7 12. Bd3 Bf5 13. Nfxd4 Bxd3 14. Qxd3
O-O 15. Rad1 Re8 16. Rfe1 Ng6 17. Qg3 Qe7 18. e6 c6 19. Re4 Qf8 20. h4 Nd5 21.
h5 f5 22. Ree1 Ngf4 23. Qg5 a5 24. Re5 h6 25. Qh4 a4 26. Nc1 Qf6 27. Qxf6 Nxf6
28. Rde1 Rad8 29. Nxf5 Rxe6 30. Rxe6 Nxe6 31. Rxe6 Rd1+ 32. Kh2 Rxc1 33. Re7
Rc5 34. Nxg7 Rb5 35. a3 Rxb2 36. Nf5 Kf8 37. Re6 1/2-1/2
About the Ruffian engine itself, results are of course not always equally good
(as no engine can achieve), i.e. it is possible that CM9_Utz is it's
angstgegner:
CM9 utz12n_Ruf101 Ath 2002
123456789012345678901234567890
1 CM9000_Utz12n 1½½½½½½1½101½11011½10111½½½1½0 19.5/30
2 Ruffian 1.0.1 0½½½½½½0½010½00100½01000½½½0½1 10.5/30
AFAIK Kurt used remis.ctg for both, 40/40. Ruffian lost to the default CM9000
personality too (13.5-16.5).
http://www.skpfaeffikon.ch/Utzinger/Downloads/CM9utz12n_report.htm
Regards,
M.Scheidl
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.