Marathon #157

[AdminX]

The solution that is given is Bxh4!. However Rybka 3 gives Rf3!! a much higher eval. Should the solution be updated? Hmmm ...

[d]5rk1/p1p1br2/1p1p2qp/4p3/2PPn1pP/3Q2P1/PPN1P1K1/2RR2B1 b - - 0 1

[Mike S.]

This is an example for a test position which actually has two good, here similar continuations and the second one is missing (if it's missing; I didn't look up the Marathon set now). I was always trying to avoid such positions, but for example the "process testset" feature in Fritz can process alternative solutions if the are added as variations, in a CBH database.

I never liked the Marathon set much anyway... I think it also contains some doubles. I think it wasn't very carefully edited.

Analysis by Toga II 3.1.2SE, multi-pv(2) after depth 13:

1. -+ (-5.63): 24...Rf3 25.exf3 gxf3+ 26.Qxf3 Rxf3 27.Kxf3 Nxg3 28.Be3 Bxh4 29.Rg1 exd4 30.Nxd4 Qe4+ 31.Kf2 Kh7 32.Rxg3 Bxg3+ 33.Ke2 Be5 34.Nf3 Bxb2 35.Rh1 Qxc4+ 36.Kf2 Qc2+ 37.Kg3
2. -+ (-5.17): 24...Bxh4 25.gxh4 Rf3 26.Rf1 Rxd3 27.Rxf8+ Kxf8 28.exd3 Nd2 29.Ne1 Qf6 30.Bf2 exd4 31.h5 g3 32.Bxg3 Qf1+ 33.Kh2 Nf3+ 34.Nxf3 Qxc1 35.Kh3 Qxb2 36.Kg4


[Event "Skopje olm"]
[Site "Skopje"]
[Date "1972.??.??"]
[Round "?"]
[White "Hug, Werner"]
[Black "Hort, Vlastimil"]
[Result "0-1"]
[ECO "E18"]
[WhiteElo "2475"]
[BlackElo "2600"]
[PlyCount "54"]

1. c4 Nf6 2. Nf3 e6 3. g3 b6 4. Bg2 Bb7 5. O-O Be7 6. d4 O-O 7. Nc3 Ne4 8. Nxe4
Bxe4 9. Bf4 d6 10. Qd2 Nd7 11. Rfd1 h6 12. Ne1 f5 13. Bxe4 fxe4 14. Qc2 Nf6 15.
f3 g5 16. Be3 Qe8 17. Kg2 Qh5 18. h3 Qg6 19. Rac1 Rf7 20. fxe4 g4 21. h4 e5 22.
Qd3 Nxe4 23. Nc2 Raf8 24. Bg1 Bxh4 ({Toga II 3.1.2SE: 1)} 24... Rf3 25. exf3 (
25. Ne3 Rxg3+) 25... gxf3+ 26. Qxf3 Rxf3 27. Kxf3 Nxg3 28. Be3 Bxh4 29. Rg1
exd4 30. Nxd4 Qe4+ 31. Kf2 Kh7 32. Rxg3 Bxg3+ 33. Ke2 Be5 34. Nf3 Bxb2 35. Rh1
Qxc4+ 36. Kf2 Qc2+ 37. Kg3 {-5.63/13}) ({Toga II 3.1.2SE: 2)} 24... Bxh4 25.
gxh4 Rf3 26. Rf1 Rxd3 27. Rxf8+ Kxf8 28. exd3 Nd2 29. Ne1 Qf6 30. Bf2 exd4 31.
h5 g3 32. Bxg3 Qf1+ 33. Kh2 Nf3+ 34. Nxf3 Qxc1 35. Kh3 Qxb2 36. Kg4 {-5.17/13})
25. gxh4 Rf3 26. Ne3 (26. exf3 gxf3+ 27. Kf1 Qg2+ 28. Ke1 f2+ 29. Ke2 fxg1=Q#)
26... Rg3+ 27. Kh1 Rf2 0-1

[Tord Romstad]

The solution that is given is Bxh4!. However Rybka 3 gives Rf3!! a much higher eval. Should the solution be updated? Hmmm ...

[d]5rk1/p1p1br2/1p1p2qp/4p3/2PPn1pP/3Q2P1/PPN1P1K1/2RR2B1 b - - 0 1

I can't check at the moment, but if I remember correctly, that position is identical to ECMGCP 1495 (you know you have been doing computer chess for too long when you look at a chess position for half a second, and the first thought that pops up in your head is "ECMGCP 1495"). Rf3 has been known to be an alternative solution for many years. Most engines (including mine, the last time I checked) prefer Rf3, but both Rf3 and Bxh4 are clear wins.

Tord