Author: Robert Hyatt
Date: 11:15:38 12/02/02
Go up one level in this thread
On December 02, 2002 at 13:45:57, Peter McKenzie wrote:
>I've been looking at some positions from the Warp-Diep game from the WCCC in
>Maastricht this year. Here are a few test positions:
>
>[D]7k/p3Qp1p/1p2n3/4nqP1/2P5/2P3R1/P5RP/3r2BK w - -
>
>Warp played Qe8+ followed by Qa4 here, which wasn't very good.
>Qf6+ was suggested and certainly looks like an improvement, although perhaps its
>not so easy to white to make progress there.
Crafty quickly goes for Qxa7 but that fails low a bit and it returns to Qf6+ and
it
sticks there. Here is the analysis on a quad 550mhz xeon:
7 0.22 2.55 1. Qf6+ Qxf6 2. gxf6 h5 3. Rg8+ Kh7
4. Re8 Rf1 5. c5 bxc5
7 0.56 2.78 1. Qxa7 Nf4 2. g6 Rxg1+ 3. Rxg1 Qe4+
4. R3g2 Nxg2 5. g7+ Kxg7 6. Rxg2+
7-> 0.56 2.78 1. Qxa7 Nf4 2. g6 Rxg1+ 3. Rxg1 Qe4+
4. R3g2 Nxg2 5. g7+ Kxg7 6. Rxg2+
8 1.58 2.92 1. Qxa7 Nxc4 2. Qa8+ Kg7 3. Qf3 Qd5
4. Qxd5 Rxd5 5. h3
8-> 1.69 2.92 1. Qxa7 Nxc4 2. Qa8+ Kg7 3. Qf3 Qd5
4. Qxd5 Rxd5 5. h3
9 3.01 -- 1. Qxa7
9 4.34 2.38 1. Qxa7 Nf4 2. g6 hxg6 3. Qb8+ Kg7
4. Rg5 Rxg1+ 5. Rxg1 Qe4+ 6. R1g2 Nxg2
7. Qxe5+ Qxe5 8. Rxe5
9 4.77 2.68 1. Qf6+ Qxf6 2. gxf6 Ng6 3. Re2 Ngf4
4. Re4 Nh5 5. Rf3 Rd2
9-> 5.36 2.68 1. Qf6+ Qxf6 2. gxf6 Ng6 3. Re2 Ngf4
4. Re4 Nh5 5. Rf3 Rd2 (s=3)
10 6.11 2.73 1. Qf6+ Qxf6 2. gxf6 Rd8 3. c5 bxc5
4. Rb2 h5 5. Be3 Rd1+ 6. Kg2 Kh7 (s=2)
10-> 10.60 2.73 1. Qf6+ Qxf6 2. gxf6 Rd8 3. c5 bxc5
4. Rb2 h5 5. Be3 Rd1+ 6. Kg2 Kh7 (s=3)
11 13.11 2.68 1. Qf6+ Qxf6 2. gxf6 h6 3. Rg8+ Kh7
4. Ra8 Nc6 5. Rgg8 h5 6. Rae8 Na5 7.
Rh8+ Kg6 8. Re7 Nxc4 9. Rxa7 Kxf6 10.
Rxh5 (s=2)
11-> 28.19 2.68 1. Qf6+ Qxf6 2. gxf6 h6 3. Rg8+ Kh7
4. Ra8 Nc6 5. Rgg8 h5 6. Rae8 Na5 7.
Rh8+ Kg6 8. Re7 Nxc4 9. Rxa7 Kxf6 10.
Rxh5 (s=6)
12 34.02 2.77 1. Qf6+ Qxf6 2. gxf6 Rd8 3. c5 bxc5
4. Rb2 h5 5. Be3 h4 6. Rh3 Rd1+ 7.
Kg2 Ng6 (s=5)
12-> 42.77 2.77 1. Qf6+ Qxf6 2. gxf6 Rd8 3. c5 bxc5
4. Rb2 h5 5. Be3 h4 6. Rh3 Rd1+ 7.
Kg2 Ng6 (s=2)
13 52.28 2.83 1. Qf6+ Qxf6 2. gxf6 Rd8 3. c5 bxc5
4. Rb2 h5 5. Rb7 a6 6. Be3 Rd7 7. Rb8+
Kh7 8. Rbg8
13-> 1:22 2.83 1. Qf6+ Qxf6 2. gxf6 Rd8 3. c5 bxc5
4. Rb2 h5 5. Rb7 a6 6. Be3 Rd7 7. Rb8+
Kh7 8. Rbg8 (s=3)
>
>The current Warp however prefers Qxa7, which also looks good. Personally I like
>Qxa7 better than Qf6.
>What do other engines play here?
>
>[D]8/p4pkp/1p2n3/2P1n1P1/8/1QP3R1/P5RP/2qr2BK b - -
>
>Here Diep played Nf4!! which forces a draw. How long do other engines take to
>find Nf4 ?
>Warp needs about 11 minutes on my PIII 1Ghz.
Here is Crafty's output. If it has a 1 minute search time target it will find
this because
it has to finish the current ply-1 move before exiting. It takes 1:25 to fail
high on Nf4
at at 3:31 the score comes back but not yet as a draw...
10-> 5.32 1.72 1. ... Nxc5 2. Qb4 Re1 3. Qh4 Qb1 4.
Qh3 Qg6 5. Re3 Ncd3 6. Rxe1 Nxe1 (s=2)
11 7.82 1.60 1. ... Nxc5 2. Qb4 Re1 3. Qh4 Qb1 4.
Qd4 Ne6 5. Qd2 Qd1 6. Rf2 a6
11-> 11.90 1.60 1. ... Nxc5 2. Qb4 Re1 3. Qh4 Qb1 4.
Qd4 Ne6 5. Qd2 Qd1 6. Rf2 a6 (s=2)
12 25.21 1.56 1. ... Nxc5 2. Qb4 Kg8 3. Rh3 Ng6 4.
Rf3 Kg7 5. Rf6 Rd3 6. Rgf2 Nh8 7. Qh4
Qxc3
12-> 44.87 1.56 1. ... Nxc5 2. Qb4 Kg8 3. Rh3 Ng6 4.
Rf3 Kg7 5. Rf6 Rd3 6. Rgf2 Nh8 7. Qh4
Qxc3 (s=3)
13 54.96 1.56 1. ... Nxc5 2. Qb4 Kg8 3. Rh3 Ng6 4.
Rf3 Kg7 5. Rf6 Rd3 6. Rgf2 Nh8 7. Qh4
Qxc3 (s=2)
13 1:25 ++ 1. ... Nf4!!
13 3:31 1.15 1. ... Nf4 2. Rc2 Rxg1+ 3. Rxg1 Qe3
4. Qb1 Qe4+ 5. Rcg2 Ned3 6. Qf1 Nxg2
7. Rxg2 Nf4 8. c6 Nxg2 9. Qxg2
13-> 3:41 1.15 1. ... Nf4 2. Rc2 Rxg1+ 3. Rxg1 Qe3
4. Qb1 Qe4+ 5. Rcg2 Ned3 6. Qf1 Nxg2
7. Rxg2 Nf4 8. c6 Nxg2 9. Qxg2
14 4:24 0.99 1. ... Nf4 2. Rc2 Rxg1+ 3. Rxg1 Qe3
4. Qb1 Qe4+ 5. Rcg2 Ned3 6. Qf1 Nxg2
7. Rxg2 Nf4 8. cxb6 axb6 9. Kg1 Nxg2
10. Qxg2
>
>[D]8/p4pkp/1p6/2P1n1P1/5n2/1QP1q3/P1R4P/6RK w - -
>
>This is the game after Nf4 Rc2 Rxg1 Rxg1 Qe3.
>How long for your engine to show a draw score here? Warp's score drops to +0.3
>at depth 12, in about 3.5 minutes. I'll let it continue to see when it sees the
>draw.
My score is down to +.10 after a few seconds. I get draw score at depth=14,
after 94
seconds.
12 15.47 0.10 1. Qb1 Qe4+ 2. Rcg2 Ned3 3. Qf1 Nxg2
4. Rxg2 Nf4 5. h3 Nxg2 6. Qxg2 Qe1+
7. Qg1 Qxc3 8. cxb6 axb6 (s=3)
14 2:33 0.00 1. Qb1 Qe4+ 2. Rcg2 Ned3 3. Qf1 Nxg2
4. Rxg2 Nf4 5. cxb6 axb6 6. Kg1 Nh3+
7. Kh1 Nf4
Those would be 50% faster on my xeon 700 box...
>
>thanks,
>Peter
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.