Author: Slater Wold
Date: 09:02:10 07/25/02
Go up one level in this thread
On July 25, 2002 at 10:12:45, Kolss wrote:
>On July 24, 2002 at 23:34:17, Slater Wold wrote:
>
>>[D]1N4br/3pkPp1/6PP/8/pp3P2/8/P4r1P/6K1 w - - 0 1'
>>
>>Every program I tested found the correct move in about 5 seconds.
>>
>>Every program I tested then failed low for 10-13 ply.
>>
>>The real test here is not "finding" the move. But finding that it's winning.
>>
>>How long does it take your program to find hxg7, and _know_ it's winning?
>
>Fantastic position! (Thanks for posting it!)
You are very welcome. This is for sure my favorite chess position. I just
found it about 12 hours before posting. (From 1993 r.g.c.c.)
>Here is what Ikarus thinks about it (K6/2-475MHz, 8MB Hash):
>
> 9 0 +4.28 1. hxg7 Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Nxd7 Kxg7 6. f5
> 10 2 +4.22 1. hxg7 Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Na6 Kxg7 6. Nxb4 Kf6
> 11 3 FailLow 1. hxg7? Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Na6 b3 6. axb3 axb3
> 11 4 +3.74 1. hxg7 Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Nxd7 b3 6. axb3 axb3 7. Ne5+ Kxg7
> 12 14 +3.72 1. hxg7 Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Nxd7 b3 6. g8=Q+ Kxg8 7. Nf6+ Kf7
> 13 26 +3.59 1. hxg7 Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Nxd7 b3 6. g8=Q+ Kxg8 7. axb3 axb3 8. Nf6+ Kf7 9. Nd5
> 14 51 FailLow 1. hxg7? Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Nxd7 b3 6. axb3 a3 7. g8=Q+ Kxg8 8. Nc5 a2
> 14 52 FailLow 1. hxg7? Rg2+ 2. Kxg2 Rxh2+ 3. Kxh2 Bxf7 4. gxf7 Kxf7 5.
>Nxd7 b3 6. axb3 a3 7. g8=Q+ Kxg8 8. Nc5 a2
> 14 83 +0.44 1. hxg7 Rg2+ 2. Kf1 Rf2+ 3. Ke1 Re2+ 4. Kd1 Rd2+ 5. Kc1
>Rc2+ 6. Kb1 Rb2+ 7. Ka1 Rxa2+ 8. Kb1 Rb2+ 9. Kc1 Rc2+ 10. Kd1 Rd2+ 11. Ke1 Re2+
>12. Kf1 Rf2+ 13. Kg1 Rg2+ 14. Kxg2 Rxh2+ 15. Kxh2 Bxf7
> 15 224 FailHigh 1. hxg7! Rg2+ 2. Kf1 Rf2+ 3. Ke1 Re2+ 4. Kd1 Rd2+ 5. Kc1
>Rc2+ 6. Kb1 Rb2+ 7. Ka1 Rxa2+ 8. Kb1 Rb2+ 9. Kc1 Rc2+ 10. Kd1 Rd2+ 11. Ke1 Re2+
>12. Kf1 Rf2+ 13. Kg1 Rg2+ 14. Kxg2 Rxh2+ 15. Kxh2
> 15 265 +0.92 1. hxg7 Rg2+ 2. Kf1 Rf2+ 3. Ke1 Re2+ 4. Kd1 Rd2+ 5. Kc1
>Rc2+ 6. Kb1 Rb2+ 7. Ka1 Rxa2+ 8. Kb1 Rb2+ 9. Kc1 Rc2+ 10. Kd1 Rd2+ 11. Ke1 Re2+
>12. Kf1 Rf2+ 13. Kg1 Rg2+ 14. Kxg2 Rxh2+ 15. Kxh2 Bxf7
> 16 513 +1.01 1. hxg7 Rg2+ 2. Kf1 Rf2+ 3. Ke1 Re2+ 4. Kd1 Rd2+ 5. Kc1
>Rc2+ 6. Kb1 Rb2+ 7. Ka1 Rxa2+ 8. Kb1 Rb2+ 9. Kc1 Rc2+ 10. Kd1 Rd2+ 11. Ke1 Re2+
>12. Kf1 Rf2+ 13. Kg1 Rg2+ 14. Kxg2 Rxh2+ 15. Kxh2 Bxf7 16. gxf7
That is an outstanding PV.
I think about 4, maybe 5 or 6 more ply and you'd have a *lot* better score. Of
course, that might take upwards to 2500+ seconds.
>So it gets the "real" fail-low at ply 14 and shows the correct move 2. Kf1 after
>83 seconds, including the entire correct line. The score goes up again in the
>following plies, although I admit that it is far from a convincing winning
>score.
>
>There are several programs which show 2. Kf1 very quickly (e.g. Goliath 3.6,
>also Yace), but with a 0.00 score and repetition of moves (Kf1-g1-f1); however,
>I did not let them calculate for very long... (Maybe they will find the
>continuation with positive score a little later.)
>By the way, I do not think that 2. Kxg2 is lost - in fact, it should be drawn.
>Only after 2. ... Rxh2+ 3. Kxh2 Bxf7, you have to find 4. Nxd7!:
Shredder was one that as soon as it found Kf1, it thought draw. It never got
that after the pawn capture, the king can come back safely. Which is actually a
pretty tough concept to grasp, let alone the depth it has to see it at.
>[D]1N6/3pkbP1/6P1/8/pp3P2/8/P6K/8 w - - 0 4
>
>Here is my analysis in PGN-format:
>
>[Event "?"]
>[Site "Notebook"]
>[Date "2002.07.25"]
>[Round "?"]
>[White "?"]
>[Black "?"]
>[Result "0-1"]
>[SetUp "1"]
>[FEN "1N4br/3pkPp1/6PP/8/pp3P2/8/P4r1P/6K1 w - - 0 1"]
>
>1. hxg7 Rg2+ 2. Kf1 $1 (2. Kxg2 $2 Rxh2+ 3. Kxh2 Bxf7 4. Nxd7 $1 (4. gxf7 $2
>Kxf7 5. Nxd7 b3 6. axb3 a3 $1 {0-1}) 4... Bxa2 5. Nc5 a3 (5... Kf6 6. Nxa4 Kxg6
>7. Nc5 Kxg7 {1/2-1/2}) 6. Nd3 Kf6 7. Nxb4 Bg8 8. Kg3 Kxg6 9. Kf3 Kxg7 10. Ke3
>{1/2-1/2}) 2... Rf2+ 3. Ke1 Re2+ 4. Kd1 Rd2+ 5. Kc1 Rc2+ 6. Kb1 Rb2+ 7. Ka1
>Rxa2+ 8. Kb1 Rb2+ 9. Kc1 Rc2+ 10. Kd1 Rd2+ 11. Ke1 Re2+ 12. Kf1 Rf2+ 13. Kg1
>Rg2+ 14. Kxg2 Rxh2+ 15. Kxh2 Bxf7 16. gxf7 Kxf7 17. Nxd7 Kxg7 18. Nc5 a3
>0-1
>
>Regards, Munjong.
Thanks for posting your results!
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.