Author: Robert Hyatt
Date: 14:26:00 01/02/98
Go up one level in this thread
On January 02, 1998 at 12:41:04, Howard Exner wrote:
>These two positions from Crafty games saw some bold moves made
>by crafty in the recent CSCC tournament. These are speculative moves
>that seemed to backfire in this case but may well work in other
>situations. Do other programs consider these moves?
>
>r2q1rk1/1bp2pp1/pb1p1n2/1p2N1p1/4P3/1BP5/PP1N1PPP/R2QR1K1 w - -
>id"Crafty 14.2 - Comet-A.75"; bm b3f7;
I ran this on my P5/233 notebook, which is about 15% slower than a
P6/200 running Crafty. Here's what I get with 14.3/14.4:
10 36.74 0.09 Nxf7 Rxf7 Bxf7+ Kxf7 a4 Qe8 a5 Bc5
Qf3 Kg8 Rad1
10-> 1:11 0.09 Nxf7 Rxf7 Bxf7+ Kxf7 a4 Qe8 a5 Bc5
Qf3 Kg8 Rad1
11 2:32 -0.05 Nxf7 Rxf7 Bxf7+ Kxf7 a4 Qe8 axb5
axb5
Rxa8 Bxa8 Qe2 Kf8 e5 Qxe5 Qxe5 dxe5
Rxe5
11 4:42 0.01 Bxf7+ Rxf7 Nxf7 Kxf7 a4 Qd7 axb5
axb5
Rxa8 Bxa8 Nf3 g4 Qb3+ Kf8 Ng5
time: 5:07 cpu:99% mat:0 n:21291910 nps:69565
>
>r2q1rk1/5ppp/p3bb2/1ppn4/8/1BPP2P1/PP1N1P1P/R1BQR1K1 b - - id"Comet-A.75
>- Crafty 14.2"; bm d5c3;
If I were annotating this game, at this point I would
probably comment that black has a strategically won game,
because of the big development lead and active pieces. But
to trade a knight for three pawns is a decision not to be made
lightly or quickly, because the resulting passed pawns have to
get mobilized quickly before the extra piece comes into play and
starts winning pawns back...
11-> 2:55 -0.89 Nxc3 bxc3 Bxc3 Rb1 Bf5 Bc2 Bxd3 Bxd3
Qxd3 Re3 Qd4 Re7
12 3:32 -0.86 Nxc3 bxc3 Bxc3 Rb1 Bf5 Bc2 Bxd3 Bxd3
Qxd3 Re3 Qd4 Qc2 Bb4
time: 4:24 cpu:99% mat:-1 n:17954205 nps:68342
>
>bm here means move made by Crafty. How does Crafty evaluate these
>positions on longer time controls? This active play style seems
>geared towards keeping human opponents on there toes.
>
>Anyway so as not to be perceived as picking on Crafty it won a nice
>endgame from this position:
>
>8/8/1b6/4kN2/P5KP/4p3/8/8 w - - id"Crafty 14.2 - M-Chess 7"; bm f5g3;
>
>Some programs might have played the poorer drawing move, Nxe3.
>
>Thanks for the tournament and posting of the games.
In any case, you are correct... I'm still working on "foxing these
humans" although I am not trying to make it lose two of two to Comet
or another program. :)
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.