« Se peut-il que les deux camps aient roqué plus tôt dans la partie ? »

diag.9

a b c d e f g h

Retro Mailing List, 18.7.1995

« Se peut-il que les deux camps aient roqué plus tôt dans la partie ? »

Yes. Inefficient proof game:

1. Nf3 Nf6 2. Nh4 Ne4 3. Nf5 Ng3 4. c3 Nxf1 5. Ng3 Ne3 6. a4 Nd5 7. O-O g6 8. Nh1 Bh6 9. g3 O-O 10. a5 Nc6 11. a6 Nb6 12. Na3 Rb8 13. Nc2 Na8 14. Nb4 b6 15. Na2 Bb7 16. Qa4 Na5 17. Qh4 Bc6 18. Qh3 Ba4 19. Kg2 Nb7 20. Kf3 Qc8 21. axb7 Kg7 22. bxc8=B Kf6 23. Bb7 Rh8 24. Qg2 Bf8 25. Be4 h5 26. Bd3 h4 27. Kg4 h3 28. Kf4 hxg2 29. Ke3 g1=N 30. Be4 Ke6 31. Kd3 Nh3 32. Ke3 Kd6 33. Bd3 Kc6 34. Kf3 Kb7 35. Bf5 Ng5+ 36. Kf4 Nh7 37. Bh3 Ka6 38. Kf3 Bc6+ 39. Ke3 Bb7 40. Rg1 Bc8 41. Bf1 Kb7 42. Kf3 Nf6 43. Kg2 Nh7

Retro-analysis:

Suppose both sides have castled.

First, if white has castled, he must have castled kingside (queenside is blocked by WB on c1). He must have castled after WBg1 left, and before the knight reached h1 (the rook was blocking it). The knight must have reached h1 before g2->g3. Therefore, the WB left f1 before g2 advanced. That can only occur if it was captured by a knight, and the current piece is promoted.

The WP that promoted could only have come from a2 (all others are still present), and needed to make at least two captures to promote on a white square. That accounts for all captures by white. It promoted after BP->b6, which was after BN->a8, so it promoted on c8.

If black castled queenside, then by the same reasoning as in paragraph 1, white must have captured the original BB with a knight. Since that didn't occur, black castled kingside.

The white pawn couldn't have captured the missing BP directly (too far away), so that BP promoted. The BP had only one capture available, so it promoted on g1 to a knight or queen (only pieces that could escape).

The captured black pieces were both Q or N. (There are two BR's & BB's on the board, and the BP couldn't have promoted to them.)

With all that, it's fairly straightforward to work out a proof game.

[Solution provided by Mark Jeffrey Tilford <tilford(at)ugcs.caltech.edu>]

__________

Pour revenir à la page d’accueil du site, rétro-sauter là.