Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Interesting position?

Author: Mogens Larsen

Date: 06:14:40 08/03/00

Go up one level in this thread


On August 03, 2000 at 07:48:33, Andrew Williams wrote:

>I like moves like this one. My program finds it but it takes 64 seconds on
>my PIII-450. There doesn't seem to be any huge tactical gain that my program
>can find. I ran it for 600 seconds, but it couldn't finish ply 11 in that time:
>
> 8= 32    20    906282   1. a3 e5 2. b3 Qc7 3. gxf7 Rxf7 4. Nde2 Qa5 5. Nb1
> 9> 32    40   1694404   1. b3 Qb4 2. Kb1 Nc6 3. Nxc6 Bxc6 4. Nd5 Qxd2 5. Nxe7
> 9b 37    64   2555066   1. Bxh6 b4 2. Bxg7 bxc3 3. Qh6
> 9= 42    70   2815029   1. Bxh6 b4 2. Bxg7 bxc3 3. Qh6 Bg5 4. Qxg5 cxb2 5. Kxb2
>10> 42    86   3380818   1. Bxh6 b4 2. Bxg7 bxc3 3. Qh6 Bg5 4. Qxg5 cxb2 5. Kxb2
>10= 81   242   9724783   1. Bxh6 b4 2. Bxg7 bxc3 3. Qh6 Bg5 4. Qxg5 cxb2 5. Kxb2
>11> 81   300  11821321   1. Bxh6 b4 2. Bxg7 bxc3 3. Qh6 Bg5 4. Qxg5 cxb2 5. Kxb2
>
>How long did Bringer take to see this? On what HW?

I don't how fast it found the move in the game, probably around 60-70 seconds,
but I tried running an analysis in the Bringer GUI. It had to use more than 2
minutes to change from gxf7 to Bxh6. I guess this tells you something about the
importance of storing results from previous searches. So if your program had
made the same 16 moves, then it would probably find the right move a lot faster.
Your program also considered b3, which might work as well because the forced
queen move can't protect against Bxh6.

The hardware was a 800MHz Athlon with 32Mb hash used.

Best wishes...
Mogens



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.