Author: John Stanback
Date: 13:07:11 12/01/99
Go up one level in this thread
On November 30, 1999 at 18:26:44, Howard Exner wrote:
>Recently their have been some interesting posts on the theme of
>a Bishop exchange for two kingside pawns. One way to test your program
>is to have a series of such positions where at times it is a good idea
>to make the exchange while other times it is not. Below are some that work,
>- a1(posted by Peter McKenzie),b1 (posted last week by yibling fan), c1 and d1.
>The other five positions (a2,a3,b2,c2, and d2) are modifications to
>the originals and for these the exchange of bishop for two pawns should be
>avoided or at least becomes not as certain. Posa3 is now very speculative.
>
>r3q1k1/ppp1rpp1/2n1b2p/8/2P2B2/3B4/PPPQ1RPP/5RK1 w - - id Posa1 - BXH6; bm f4h6;
>r3q1k1/ppp1rpp1/4b1np/8/2P2B2/3B4/PPPQ1RPP/5RK1 w - - id Posa2 - amBXH6;
>3rrnk1/ppp1qpp1/4b2p/8/2P2B2/3B4/PPPQ1RPP/5RK1 w - - id Posa3 - BXH6spec;
>3r1rk1/p2q1ppp/2pb1n2/5b2/1nB5/2N1BN1P/PP2QPP1/R4RK1 b - - id Posb1 - BXH3; bm
>f5h3;
>3r1rk1/p2q1ppp/2pb4/p4b2/1nB5/4BN1P/PP2QPP1/R4RK1 b - - id Posb2 - amBXH3;
>r1br2k1/p1q2pp1/4p1np/2ppP2Q/2n5/2PB1N2/2P2PPP/R1B1R1K1 w - - id Posc1 - bm
>Bxh6; bm c1h6;
>r1bb2k1/p1q2pp1/4p1np/2ppP2Q/8/2P2N2/2P2PPP/R1B1R1K1 w - - id Posc2 - amBxh6;
>1qr1b1k1/4bpp1/pn2p2p/1p1nN3/3P4/P2BBN1Q/1P3PPP/4R1K1 w - - id Posd1 - BXH3.EPD;
>bm e3h6;
>1qr1b1k1/4bpp1/p3p2p/1p1nN3/3P4/P3BN1Q/1P3PPP/4R1K1 w - - id Posd2 - amBXH3.EPD;
>
>Programming chess programs looks complicated. How much to weight this or that
>piece of
>code has an element of "rolling the dice". I can imagine alot of headheaches
>when one
>goes about trying to teach their program something new - will it be successful
>in the target
>area you were aiming for but at the same time hurt some other aspect of the
>program ? Like
>someone plugging up a hole in a boat when all at once two more leaks gush forth.
I tried these positions for 3 minutes each on Zarkov.
Posa1: Zarkov likes Bxh6 from depth 4 on, score is +1.04 after depth 9
Posa2: Zarkov likes Bxh6 at depth 3 (+0.49) and depth 8 (+0.61) but fails
at depth 9 and switches to Bxg6 (+0.24)
Posa3: Zarkov never likes Bxh6, picks Qa5 (+0.68)
Posb1: Zarkov did not find Bxh3 in 3 minutes, likes Nc2 (+0.30)
Posb2: Zarkov never likes Bxh3, picks Nd5 (+0.91)
Posc1: Zarkov did not find Bxh6, likes Bxg6 at all depths (-0.16)
Posc2: Zarkov never likes Bxh6, picks c4 (+0.23)
Posd1: Zarkov likes Bxh6 at all depths, +1.81 after depth 9
Posd2: Zarkov never likes Bxh6, picks Bd2 (-0.59)
John
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.