So what's the best move here from a drawn game?

[Hagen]

On Susan Polgar's site there's a diagram from a game played recently between Antoneta Stefanova against Tatiana Kosintseva. It was a draw...but surprisingly according to GM Susan Polgar the following position could have lead to a completely different outcome. White to move -

4r1k1/2R4R/7p/4P3/1p1P3P/3n2B1/r1p3PK/8 w - - 0 1

I've tried but can't figure out the right move here (I decided not to use either Stockfish or Komodo. That would have been too easy.) Any ideas?

[Eduard Nemeth]

I see only this line, but with a good advantage for white:

[Event "Testposition Polgar'"]
[Site "Privat"]
[Date "????.??.??"]
[Round "?"]
[White "Neue Partie"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "4r1k1/2R4R/7p/4P3/1p1P3P/3n2B1/r1p3PK/8 w - - 0 1"]
[PlyCount "9"]

1. e6 Rxe6 2. Rcg7+ Kf8 3. Rb7 Kg8 4. Rhc7 Ra8 5. Rxc2 $18 *

r5k1/1R6/4r2p/8/1p1P3P/3n2B1/2R3PK/8 b - - 0 5

White is here now with better position (+-).

ED.

[carbon]

black holds - below is a run out into TB data

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "New game"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "4r1k1/2R4R/7p/4P3/1p1P3P/3n2B1/r1p3PK/8 w - - 0 1"]
[PlyCount "52"]

1. Rcg7+ (1. e6 Rxe6 2. Rcg7+ Kf8 3. Rb7 Kg8 4. Rhc7 Ra8 5. Rxc2) 1... Kf8 2.
e6 Rxe6 3. Rb7 Kg8 4. Rhc7 Ra8 5. Rxc2 Rf6 6. d5 (6. Rcc7 Rd8 7. Be5 Nxe5 8.
Rg7+ Kh8 9. dxe5 Re6 10. Rh7+ Kg8 11. Rbg7+ Kf8 12. Rg4 Rd5 13. Rxb4 Kg8 14.
Ra7 Rdxe5 15. Rbb7) 6... Rf7 7. Rcc7 Rxc7 8. Bxc7 Ra4 9. Kg3 Kg7 10. Bf4+ Kg6
11. h5+ Kxh5 12. Rb6 Nxf4 13. Kxf4 b3+ 14. Ke5 Ra2 15. Rxb3 Kg6 16. Rb6+ Kg5
17. Rb4 Rxg2 18. d6 Re2+ 19. Re4 Rd2 20. Ke6 h5 21. Re5+ Kg4 22. Rd5 Re2+ 23.
Kf7 Rf2+ 24. Ke7 Ra2 25. d7 Ra8 26. d8=Q Rxd8 $11 *

regards
carbon

[Eduard Nemeth]

7. Rcc7 is a wrong line. I think that white is winning after 7. Rb5! ++-

Position after 7. Rb5!:

r5k1/5r2/7p/1R1P4/1p5P/3n2B1/2R3PK/8 b - - 0 1

I think, white is winning here.

ED.

[carbon]

7. Rcc7 is a wrong line. I think that white is winning after 7. Rb5! ++-

Position after 7. Rb5!:

r5k1/5r2/7p/1R1P4/1p5P/3n2B1/2R3PK/8 b - - 0 1

I think, white is winning here.

ED.

Yes! with the idea Rc4/Rd4 winning a pawn.
Stockfish did not find this move within 26 plys.
I did not see this move either.
But any engine reaching to tb from here should find it with ease.
thanks for the correction.

[Hagen]

This is interesting. If Stockfish couldn't find this move...could Rybka 4 find it? And if neither engine can find it...is it because theyr'e not using TBs? In other words - would Stockfish have found this move without any TB support had the endgame strength been addressed in this release? Would a Stockfish 1.8a Beta fix this problem?

[carbon]

This is interesting. If Stockfish couldn't find this move...could Rybka 4 find it? And if neither engine can find it...is it because theyr'e not using TBs? In other words - would Stockfish have found this move without any TB support had the endgame strength been addressed in this release? Would a Stockfish 1.8a Beta fix this problem?

Here are pv's with less than 1 min:

[Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "New game"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "4r1k1/2R4R/7p/4P3/1p1P3P/3n2B1/r1p3PK/8 w - - 0 1"]
[PlyCount "13"]

1. Rcg7+ Kf8 2. e6 Rxe6 3. Rb7 Kg8 4. Rhc7 Ra8 5. Rxc2 Rf6 6. d5 Rf7 7. Rb5

({ Deep Shredder 12 UCI:} 7. Rb6 Rd7 8. d6 Kf7 9. Rb5 Re8 10. Rc4 Kf6 11. Rd4 Re3 12. h5 Ne5 13. Rbxb4 Nc6 14. Rf4+ Ke6 15. Rbe4+ Rxe4 16. Rxe4+ Kd5 17. Re8 Nd4 18. Rc8 Nf5 19. Rc7 Ke6 20. Rxd7 Kxd7 {2.42/18})

({Deep Rybka 4 w32:} 7. Rb5 Rf5 8. Rc4 Kf7 9. Rd4 Ne5 10. Rb7+ Kf6 11. Rdxb4 Ra6 12. R7b6+ Rxb6 13. Rxb6+
Kg7 14. Bxe5+ Rxe5 15. d6 Rd5 16. Kg3 Kf7 17. d7 Rxd7 18. Rxh6 Kg7 19. Rb6 Rd2 20. Kf3 Rd3+ 21. Kg4 {2.54/18})

({Houdini 1.02 w32 2_CPU w32 2_CPU:} 7. Rb5 Rf5 8. Rc4 h5 9. Rd4 Ne5 10. Bxe5 Rxe5 11. Rbxb4 Ra7 12. d6 Rd7 13. g4 hxg4 14. Kg3 Kf7 15. Rd3 Re1 16. Rxg4 Kf6 17. Kf3 Re5 18. Rd1 Kf7 19. Rd2 Re6 20. Rgd4 Re5
21. Rg2 Rf5+ 22. Kg4 Kf6 23. Rg3 Rg7+ 24. Kh3 Rxg3+ 25. Kxg3 {1.53/24})

({ Fritz 12:} 7. Rcc7 Rxc7 8. Bxc7 Ra4 9. Kg3 Nc5 10. Rb5 Nd7 11. Bf4 h5 12. d6 Ra3+ 13. Kf2 b3 14. Rxh5 Ra2+ 15. Kf3 b2 16. Rb5 Kf7 17. h5 Ke6 18. h6 {2.11/22 })

({Stockfish 1.8 JA:} 7. Rb5 Rf5 8. Rc4 h5 9. Rd4 Ne5 10. Bxe5 Rxe5 11. Rdxb4 Kf7 12. Rb7+ Kf6 13. R4b6+ Kf5 14. Rh6 Ke4 15. d6 Rd5 16. d7 Rd8 17. Rh7 Rd6 18. Rb4+ Ke3 19. Rxh5 R8xd7 20. g4 Rd4 21. Rxd4 Rxd4 22. Kg3 Ke4 23. g5 Kf5 { 3.11/28}) *



shredder comes up with Rb6 instead of Rb5
but stockfish is fast reaching 28 plys in this position compared to R4 with 18 plys only.

[Eduard Nemeth]

I am very sure (100%) that I would win this position
a) against the best human player of the world.
b) against the best Engine of the world.

This position (after Rb5!) is 100% won for white.

Here an analysis by Shredder 10.1 UCI Single-Engine after 10 Minutes
on 2.66 GHz, with 256 MB Hash and 5-men EGTB:

r5k1/1R3r2/7p/3P4/1p5P/3n2B1/2R3PK/8 w - -

Engine: Shredder 10.1 UCI (256 MB)
von Stefan Meyer-Kahlen

16/32 0:15 +2.84 7.Tb5 Td7 8.d6 Kf7 9.Tc4 Ta6 (15.034.812) 984 TB:10

17/33 0:20 +2.83 7.Tb5 Td7 8.d6 Kf7 9.Tc4 Ta3 10.Tc7 Ke6
11.Tbb7 Td8 12.Th7 b3 13.Txh6+ Kf5
14.Tb5+ Ke4 15.Te6+ Kd4 (19.629.155) 976 TB:49

18/38 0:49 +3.08++ 7.Tb5 Td7 (47.752.702) 958 TB:477

18/40 1:13 +3.25 7.Tb5 Td7 8.d6 Kf7 9.Tc7 Ke6 10.Th5 Th8
11.Tc4 b3 12.Td4 Sc1 13.Td1 Tc8 (69.857.429) 945
TB:1.477

19/42 2:41 +3.44 7.Tb5 Td7 8.d6 Kf7 9.Tc4 Kf6 10.h5 Te8
11.Td4 (147.433.603) 910 TB:4.548

20/46 7:56 +3.66 7.Tb5 Tf5 8.Tc4 (427.688.158) 897
TB:16.189

ED.