Author: Chuck
Date: 18:38:56 12/10/99
Go up one level in this thread
On December 10, 1999 at 21:33:36, Aloisio Ponti Lopes wrote:
>>[Event "ICC 3 0"]
>>[Result "1-0"]
>>[ICCResult "Black resigns"]
>>[WhiteElo "2235"]
>>[BlackElo "2377"]
>>[Opening "Anti-Borg (Desprez) opening"]
>>[ECO "A00"]
>>[Time "00:14:23"]
>>[TimeControl "180+0"]
>>
>>1. h4 e5 2. d3 d5 3. Bd2 Nf6 4. Na3 Bxa3 5. bxa3 O-O 6. g3 Nc6 7. Bh3 Qd6 8.
>>Kf1 Bxh3+ 9. Nxh3 Qxa3 10. Ng5 Qa4 11. Rh3 h6 12. e3 hxg5?? 13. hxg5 Nd7 14.
>>Qh5 Qh4 {Black resigns} 1-0
>>
>>WOW!
>>
>Hmm...I could not do that against Fritz 5.16... after a few seconds it considers
>12....e4 as the best move, and evaluates this position as a true win for
>black... (running on AMD-K62-300 MHz, 128 MB RAM, only 32 MB Hash tables).
>Fritz also didn't "like" 5....0-0 - it thought of developing pieces earlier.
>Did you test this line against other strong commercial softwares?
>
>A. Ponti
Certainly. Several of the moves would be avoided with enough time to think.
That's the point. In the actual game the computer thought for 5 seconds on
12..hxg5. The programs I checked all pick this move for a second or two at
least, but the point is more general. You shouldn't look at this as "does my
computer find this move in 5 seconds?", but rather just be aware that most
programs are susceptible to this type of play if you shorten their thinking time
enough. This is just an extreme example of anti-computer play on ICS.
Chuck
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.