Author: Simon Finn
Date: 12:12:18 07/01/00
Go up one level in this thread
On July 01, 2000 at 07:30:19, Paulo Soares wrote:
>On June 30, 2000 at 19:47:36, Simon Finn wrote:
>
>(sniped)
>
>>Black's King gets driven back to d5, so it's simplest for Black
>>to answer 1. Rc4+ with 1.. Kd5, giving the following position:
>>
>>[D]8/8/8/p2k4/2R5/1P6/1K6/3q4 w - - 0 1
>>
>>Now Black's winning chances consist of either
>>
>>(1) Exchanging pawns to reach a winning QvR ending, or
>>(2) Sacrificing the a-pawn to reach a winning QvR+P ending.
>>
>>To prevent plan (1), White should leave his Rook on c4, to anwer a4 with Rxa4.
>>This allows Black to try plan (2).
>>
>>2. Ka2 Qd2+ 3. Kb1 Qf2 4. Ka1 Qe2 5. Kb1 Qd2 6. Ka1
>>
>>[D]8/8/8/p2k4/2R5/1P6/3q4/K7 b - - 0 1
>
>(sniped)
>
>>
>>Simon
>
>I think that 6. Ka1 is a mistake, because after 6 ...a4 7. Rxa4 Qc2!(instead
>7...Qc3+) should win.
Good move! You're absolutely right - 7... Qc2! wins for Black.
After 7... Qc2 8. Ra5+ Kc6 9. Ra6+ Kb7 10. Ra3 Kb8
zugzwang forces white to play 11. b4 after which Black can advance his King
and eventually capture the pawn.
>I didn't analyze other possibilities for the sixth white move, maybe
>there are some better move.
Moving the Rook off the 4th rank loses instantly to 6... Qb4 followed by pushing
the a-pawn, so the only reasonable alternative is 6. Ra4.
Now Black can win by improving the position of his King
6. Ra4 Kc5 7. Rc4+ Kb5
giving the following position:
[D]8/8/8/pk6/2R5/1P6/3q4/1K6 b - - 0 1
Again, White has to choose between 8. Ra4 and 8. Ka1.
8. Ra4 Qc3 wins for Black after:
9. Ka2 Qc2+ 10. Ka1 (10. Ka3 Qb1 forces a quick mate) Qxb3 11. Rxa5+ Kc4
with an EGTB win.
9. Ra3 Qd3+ 10. Kc1 Kb4 11. Ra4+ Kc3! (not 11... Kxb3? 12. Ra3+ =)and
9. Ra3 Qd3+ 10. Kb2 Qd2+ 11. Kb1 Kb4 12. Ra2 Qd1+ both win for Black
9. Ra2 a4 10. bxa4+ Kb4 11. Rc2 (other moves lose the Rook) Qd3 12. Kb2 Kxa4
is another EGTB win.
8. Ka1 is the "theory" move, analyzed by Henkin. The main line is:
8. Ka1 a4 9. Rxa4 Qc3+ (not 9... Qc2? 10. Rb4+! =) 10. Ka2 Qc2+ 11. Ka3 Qb1
12. Rb4+ Kc5 13. Rc4+ Kd5 14 Kb4 Qa1
leading to a position analyzed by Cheron in 1952.
(Note that 14. Rc8 Qa1+ 15. Kb4 transposes to 14 Kb4 Qa1 15 Rc8,
which is examined below.)
[D]8/8/8/3k4/1KR5/1P6/8/q7 w - - 0 1
Unlike the similar position with the white pawn on b4 (and all the other pieces
moved up one rank), here Black can win by forcing White to let his King cross
the c-file.
The main line of Cheron's analysis (as quoted in Hooper's "A Pocket Guide to
Chess Endgames") is:
15. Rc5+ Kd4 16. Kb5 (hoping to play 17. b4=) Qf1+ 17. Rc4+ Ke3 18. b4 Qe2
19. Kc5 Kd3 20. Kb5 Qe8+ 21. Rc6 Kd4 22. Kb6 Qb8+ 23. Ka5 Qa7+
[D]8/q7/2R5/K7/1P1k4/8/8/8 w - - 0 1
24. Ra6 (24. Kb5 Qb7+ 25. Rb6 Qd5+ 26. Ka4 Kc3) Qc7+ 25. Ka4 (or 25. Kb5 Qd7+)
Kc3
[D]8/2q5/R7/8/KP6/2k5/8/8 w - - 0 1
and having crossed the barrier file, Black has broken the fortress and will soon
win.
Of course, this analysis in intricate, and there are lots of subsidiary
variations, for example:
15. Rc8 Kd4 16. Rd8+ Ke4 17. Re4+ Kd3 18. Rc8 Qa6 19. Rc3+ Kd2 20. Rc4 Qa2
[D]8/8/8/8/1KR5/1P6/q2k4/8 w - - 0 1
21. Rc5 Qa7 22. Rc4 (22. Kb5 Qb7+ 23. Ka4 Qa6+ is similar) Qa6 23 Rc5 Qb6+
24 Rb5 Qd4+ 25 Ka5 Kc3
[D]8/8/8/KR6/3q4/1Pk5/8/8 w - - 0 1
and Black again crosses the c-file with a winning position.
Of course this analysis can possibly be improved by referring to the 5-man
tablebases (which I don't have) but the verdict of a win for Black looks
pretty secure.
So it looks like the original position was also winning for Black - thanks
Paulo!
Simon
>
>Original Position:
>8/8/8/p7/4k3/1P6/1KR5/3q4 w - - 0 1
>
>1. Rc4+ Kd5 2. Ka2 Qd2+ 3. Kb1 Qf2 4. Ka1 Qe2 5. Kb1 Qd2 6. Ka1 a4 7. Rxa4 Qc2!
>8. Ra5+ Kc6 9. Ra6+ Kb7 10. Ra3 Kc7 11. Ra7+ (11. b4 Kb6 12. Ra5 Qc3+ 13. Ka2
>Qxb4 {-#16/9}) 11... Kb8 12. Ra3 Kb7 13. b4 Kb6 14. Ra5 (14. b5 Qc1+ 15. Ka2
>Kxb5 {-#18/10}) (14. Rg3 Kb5) 14... Qc3+ 15. Ka2 Qxb4 {-#16/10}
>
>Paulo Soares, from Brazil
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.