You can download endgame tablebases suitable for Crafty from here:
ftp://ftp.cis.uab.edu/pub/hyatt/TB
I haven't used them recently, but last time I had them all downloaded they were very useful for working out how to play difficult endgames.
I'm just downloaded the King + 2 knights vs King and pawn tablebases now. If I can work out how I'll post the forced win in fifty odd moves you mentioned!
55.Nf3 Kb6 56.Nf2 Kc6 57.Ne4 Kb6 58.Kd5 Kc7 59.Nc5 Kb6 60.Ne6 Kb5 61.Nc7+ Kb6 62.Kd6 Kb7 63.Nb5 Kb6 64.Na3 Kb7 65.Nc4 Kc8 66.Na5 Kb8 67. Kc6 Ka7 68.Nb3 Ka6 69.Nc5+ Ka5 70.Nd3 Ka6 71.Nb4+ Ka5 72.Kc5 Ka4 73. Nc6 Ka3 74.Kd4 Kb3 75.Kd3 Ka3 76.Kc3 Ka4 77.Kc4 Ka3 78.Ncd4 Ka2 79.Kc3 Ka3 80.Nb3 Ka4 81.Kc4 Ka3 82.Nc5 Kb2 83.Nd3+ Kc2 84.Nf2 Kb2 85.Nd2 Ka3 86.Kb5 Kb2 87.Kb4 Kc2 88.Nde4 Kb2 89.Ka4 f3 90.Kb4 Kc2 91.Ka3 Kb1 92. Kb3 Kc1 93.Kc3 Kb1 94.Nd2+ Kc1 95.Nc4 Kb1 96.Kd2 Ka2 97.Kc2 Ka1 98.Kb3 Kb1 99.Nd3 f2 100.Na3+ Ka1 101.Nb4 f1=Q 102.Nbc2#
You would need perfect chess to do it in under 50 moves as even the optimal solution (assuming best defence of course but defence is easier than attack in this case) is 48 moves.
I have a suspicion that anyone who played the above set of moves in a game here would find their account quickly banned.
Originally posted by Dragon FireAfter 55. Nf3 we get
Would you be kind enough to post for those doubters (and me if I ever come across this again) the forced mate following move 56 Nf3.
[Event "?"]
[Site "?"]
[Date "2006.08.17"]
[Round "?"]
[White "Thud"]
[Black "Fritz 9"]
[Result "1-0"]
[Annotator ",Thud"]
[SetUp "1"]
[FEN "8/8/8/k7/2K2p2/5N2/8/3N4 b - - 0 0"]
[PlyCount "94"]
{159MB, Elect2002.ctg, HAL9000}
1... Kb6 {[%emt 0:00:00]} 2. Kd5 {[%emt 0:00:00]} Kc7 {[%emt 0:00:00]} 3. Nf2 {[%emt 0:00:00]} Kb6 {[%emt 0:00:00]} 4. Ne4 {[%emt 0:00:00]} Kc7 {[%emt 0:00:00]} 5. Nc5 {[%emt 0:00:00]} Kb6 {[%emt 0:00:00]} 6. Ne6 {[%emt 0:00:00]} Kb5 {[%emt 0:00:00]} 7. Nc7+ {[%emt 0:00:00]} Kb6 {[%emt 0:00:00]} 8. Kd6 {[%emt 0:00:00]} Kb7 {[%emt 0:00:00]} 9. Nb5 {[%emt 0:00:00]} Kb6 {[%emt 0:00:00]} 10. Na3 {[%emt 0:00:00]} Kb7 {[%emt 0:00:00]} 11. Nc4 {[%emt 0:00:00]} Kb8 {[%emt 0:00:00]} 12. Nb6 {[%emt 0:00:00]} Kb7 {[%emt 0:00:00]} 13. Nd7 {[%emt 0:00:00]} Ka6 {[%emt 0:00:00]} 14. Kc6 {[%emt 0:00:00]} Ka5 {[%emt 0:00:00]} 15. Nde5 {[%emt 0:00:00]} Ka4 {[%emt 0:00:00]} 16. Kc5 {[%emt 0:00:00]} Kb3 {[%emt 0:00:00]} 17. Nc6 {[%emt 0:00:00]} Kc3 {[%emt 0:00:00]} 18. Nb4 {[%emt 0:00:00]} Kb3 {[%emt 0:00:00]} 19. Nd5 {[%emt 0:00:00]} Ka3 {[%emt :00:00]} 20. Nc3 {[%emt 0:00:00]} Kb2 {[%emt 0:00:00]} 21. Kb4 {[%emt 0:00:00]} Kc1 {[%emt 0:00:00]} 22. Kc4 {[%emt 0:00:00]} Kc2 {[%emt 0:00:00]} 23. Nb5 {[%emt 0:00:00]} Kb2 {[%emt 0:00:00]} 24. Nbd4 {[%emt 0:00:00]} Ka2 {[%emt 0:00:00]} 25. Kc3 {[%emt 0:00:00]} Ka3 {[%emt 0:00:00]} 26. Nb3 {[%emt 0:00:00]} Ka4 {[%emt 0:00:00]} 27. Kc4 {[%emt 0:00:00]} Ka3 {[%emt 0:00:00]} 28. Nc5 {[%emt 0:00:00]} Kb2 {[%emt 0:00:00]} 29. Nd3+ {[%emt 0:00:00]} Kc2 {[%emt 0:00:00]} 30. Nf2 {[%emt 0:00:00]} Kb2 {[%emt 0:00:00]} 31. Nd2 {[%emt 0:00:00]} Ka3 {[%emt 0:00:00]} 32. Kb5 {[%emt 0:00:00]} Kb2 {[%emt 0:00:00]} 33. Kb4 {[%emt 0:00:00]} Kc2 {[%emt 0:00:00]} 34. Nde4 {[%emt 0:00:00]} Kb2 {[%emt 0:00:00]} 35. Ka4 {[%emt 0:00:00]} f3 {[%emt 0:00:00]} 36. Kb4 {[%emt 0:00:00]} Ka2 {[%emt 0:00:00]} 37. Kc3 {[%emt 0:00:00]} Ka3 {[%emt 0:00:00]} 38. Nc5 {[%emt 0:00:00]} Ka2 {[%emt 0:00:00]} 39. Ncd3 {[%emt 0:00:00]} Kb1 {[%emt 0:00:00]} 40. Ne5 {[%emt 0:00:00]} Ka2 {[%emt 0:00:00]} 41. Nc4 {[%emt 0:00:00]} Kb1 {[%emt 0:00:00]} 42. Kd2 {[%emt 0:00:00]} Ka2 {[%emt 0:00:00]} 43. Kc2 {[%emt 0:00:00]} Ka1 {[%emt 0:00:00]} 44. Kb3 {[%emt 0:00:00]} Kb1 {[%emt 0:00:00]} 45. Nd3 {[%emt 0:00:00]} Ka1 {[%emt 0:00:00]} 46. Na3 {[%emt 0:00:00]} f2 {[%emt 0:00:00]} 47. Nb4 {[%emt 0:00:00]} f1=Q {[%emt 0:00:00]} 48. Nbc2# {[%emt 0:00:00]} 1-0
Originally posted by XanthosNZI agree. No one could play this perfectly without a tablebase installed but the position is still theorectically interesting and it was worth trying. I was always happy for discussion of this game on the forum whilst still in progress because I always felt anything other than a draw was a practical impossibility.
You would need perfect chess to do it in under 50 moves as even the optimal solution (assuming best defence of course but defence is easier than attack in this case) is 48 moves.
I have a suspicion that anyone who played the above set of moves in a game here would find their account quickly banned.
As you say defense is easier than attack but an error from the defense could result in a forced mate under 48 moves.
In answer to your earlier query "what difference does the corner colour make" the answer is none expect that 3 N moves are required to mate on a1 or h8 and 4 to mate on a8 and because of the position of the black P only 3 moves are available so any attempt to mate on a8 would result in black Q-ing and winning.
Besides now knowing how it could have been done I have also discovered that the >50 move rule has been repealed which will stop me giving wrong decisions when I arbitrate in future.
Thank you to everyone that has contibuted here. I have learnt a great deal and that type of debate demonstrates everything that is good about RHP.
Originally posted by XanthosNZI am not a qualified arbiter but as team captain I am called upon to arbitrate in some league games and in our intra club matches.
It makes me weep that an arbiter could be so ignorant to the FIDE rules of chess.
I might have slipped up on the 50 move rule for K & 2N vs K & P (but it has never come up). The old rule (i.e. 100 moves) is clearly described in the (obviously obsolete) booklet on FIDE laws that I have. That booklet has now been consigned to the bin and I have downloaded a replacement.
We are all learning in the time and whilst you might have been right in this case there is no need to gloat. I can accept defeat gracefully, learn and move on. After all learning is why I joined RHP.
I am sure XNZ does not know everything! Yet!