Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: GM Suba sets a trap, several top engines fall straight in...

Author: Charles Milton Ling

Date: 14:15:33 05/27/99

Go up one level in this thread


On May 26, 1999 at 15:32:13, Francis Monkman wrote:

>Here's an interesting trap Suba set an opponent at Las Palmas, 1979:
>
>rq2r1k1/5pbp/4pnp1/2NbN2n/1p3P2/1P4PP/PB3QBK/2RR4 w - -
>
>If White plays g4?, as he did in the game, he loses immediately to Nxf4!
>Qxf4 Rxa2 and White played Rc2, but there's no defense.
>
>The only program that appeared not even to consider g4 was Genius 4,
>Genii 3 & 6 (under Genius 6 GUI) took about 8 seconds to prefer Bd4
>(which according to Suba is correct).
>
>Next were CM6000, which took about 14 seconds for g4 to go bad, played
>Ba1 (not a bad move?) after 32, Bd4 after 45, and,
>Rebel10, which took 28 seconds on g4, settled (after Ncd7 and Nc6) on Bd4
>at 1:06, allotting it a score of +.94 (?).
>
>Then, fourth equal, near enough, were LGG 2.0 and Junior 5, both finding
>Bd4 after about 3 minutes.
>
>Fritz 5.32 is still hung on g4 after *at least* 5 minutes. Maybe 10?...
>
>Hiarcs 7 plays g4 'happily', even awarding it +191 for two minutes, after which
>it drops to +159. At three minutes, it drops to -95, but it's not until 6:30
>that
>H7 finds an alternative, Nc6 (-33). This is still the choice at 17:45. It
>finally reaches Bd4 by 50:00 or so.
>
>Crafty 16.6 spotted the winning line after about 2 minutes, but as you see:
>
>depth=10 25/53 -0.17 1. g4 Nxf4 2. Qxf4 Rxa2 3. Rc2 Nxg4+ 4. hxg4 Rxb2
>5. Qxf7+ Kh8 6. Rxb2 Bxe5+ 7. Kg1 Bxb2 8. Bxd5 exd5 9. Rxd5
>Nodes: 37763067 NPS: 89276
>Time: 00:07:02.99
>
>Suba was expecting the *spurious* mate attack Qxf7+ etc., but it wasn't played!
>Soon after, Crafty gets it.
>
>depth=10 26/53 -0.14 1. Ba1 Bxg2 2. Kxg2 Ra3 3. Rc2 Qa8+ 4. Kg1 Ne4
>5. Nxe4 Qxe4 6. Nc4 Raa8 7. Bxg7 Nxg7
>Nodes: 40524144 NPS: 88569
>Time: 00:07:37.54 (these times are a few seconds late, grabbed with copy/paste)
>
>CSTal II played g4 until 10:02, then preferred Ba1.
>
>Interesting stuff.
>
>Francis

Interesting.  My Fritz 5.32 reacted differently (Pentium II 266 MHz, 16 MB hash
tables).
Liked g4 up to 1:18 (minutes/seconds), then Ned7
1:55 Nc6
4:22 Bxd5
5:45 Ba1
9:15 Bd4 (which it considers =).
Charley



This page took 0.01 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.