Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Test Position (for computers or strong players) -- Can White Draw?

Author: Dieter Buerssner

Date: 12:03:33 04/19/02

Go up one level in this thread


On April 19, 2002 at 14:07:13, John Merlino wrote:

>This position came up in a game between myself and a friend:
>
>[D]6k1/5R2/7P/6K1/8/8/p5P1/6r1 w - - 0 1
>
>The question is: can White draw (after 1.h7+ Kxf7 2.h8=Q a1=Q)?

Yace on K6 2 475.
   5936192  47.364  -4.30 12.  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke7 4. Qe2+
                               Kd6 5. Qd3+ Kc5 6. Qf5+ Kd4 7. Qf6+ Kd3 8. Qf3+
                               Kc4 9. Qf7+ Kb4 10. Qf4+ Kb3 11. Qe3+ Kc2 12.
                               Qf2+ Kc3 13. Qe3+ Kb4 {-420}
   7924329  1:01.1  -3.90 13++ 1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke7 4. Qe2+
                               Kd6 5. Qd3+ Kc5 6. Qf5+ Kb6 7. Qe6+ Ka5 8. Qf5+
                               Qe5 9. Qxe5+ Ka6
   8221980  1:03.1  -2.90 13++ 1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke7 4. Qe2+
                               Kd6 5. Qd3+ Kc5 6. Qf5+ Kb6 7. Qe6+ Ka5 8. Qf5+
                               Qe5 9. Qxe5+ Ka6
  11779777  1:27.8  -0.01 13t  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+H Kc4H 10. Qe2+H Kc5H 11. Qe7+H Kd4H
                               12. Qd6+H Kc3H {-420}
  12752834  1:38.4  -0.01 13.  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qd6+ Kc3 {-420}
  22981860  2:50.5   0.00 14t  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qf6+ Kd3 13. Qf3+ Kc4 {-420}
  24535028  3:04.6   0.00 14.  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qf6+ Kd3 13. Qf3+ Kc4 {-420}
  34509541  4:15.8  -0.40 15-- 1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qf6+ Kd3 13. Qf5+ Kc4 14. Qd7 Rxg2+ 15. Kf4
                               {-500}
  44439463  5:26.4  -1.40 15-- 1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qf6+ Kd3 13. Qf5+H Kc4H 14. Qf3H {HT}
  63631558  7:44.4   0.00 15t  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qd6+ Kc3 13. Qc6+ Kd3 14. Qf3+ Kd2 15. Qf2+
                               Kd3H 16. Qf3+H Kd2H 17. Qf2+H Kd3H 18. Qf3+HR
                               {-420}
  67524711  8:26.1   0.00 15.  1. h7+ Kxf7 2. h8=Q a1=Q 3. Qh5+ Ke6 4. Qe8+
                               Kd5 5. Qd7+ Ke4 6. Qc6+ Ke3 7. Qf3+ Kd4 8. Qf2+
                               Kc3 9. Qf3+ Kc4 10. Qe2+ Kc5 11. Qe7+ Kd4 12.
                               Qd6+ Kc3 13. Qc6+ Kd3 14. Qf3+ Kd2 15. Qf2+ Kd3
                               16. Qf3+ Kd2 17. Qf2+ Kd3 18. Qf3+ {-420}


I believe (but I am no sure) that the strange fail low at depth 15 is due to the
"graph history interference" problem. Deep in the path, some positions are found
in the HT, that look better for black, due to a previous transposition. But at
that time, the position was reached via another route, and so repetion of
position logic behaves strange. One could check this, by disable HTs (or at
least use HTs only for move ordering), but it would need quite some patience to
reach the same search depth.

The Monarch result is impressive!

Regards,
Dieter



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.