CHALLENGE: maximum number of moves delaying a checkmate in one?

[Ajedrecista]

Hello:

I read a thread about Nakamura trolling an opponent who did not want to resign in a totally lost position and I came up with a slightly different troll concept: how much can a checkmate in one be delayed? Pretty much in the spirit of this Simpsons' clip ('We totally wasted his time... and ours!').

The requirements shall be:

- A legal position (i.e. not white pawns on a2, a3 and b2; or 10+ queens-a-side).
- There must be always an available checkmate in one move for the winning side (i.e. wKh6, wQe7, bKh8 and white moves Qd7 shall be discarded).

I composed the following (hopefully legal) position after some tunings:

[d]7k/K6q/7Q/6NB/p1p1p1p1/PpPpPpPp/1P1P1P1P/8 w - - 0 1

The plan is to move the king with caution to avoid draws by stalemate, threefold repetition and fifty-move repetition. Black can only move Kg8 and Kh8. Something like this:

Code: Select all

   1. Nxh7  // From here, there will be Nf6# at least
  51. Kxa4
 101. Kxc4
 151. Kxe4
 201. Kxg4
 251. Kxb3
 301. Kxd3
 351. Kxf3
 401. Kxh3
 451. g4
 501. g5
 551. g6 
 552. Bg4  // Once g6 is covered by f7, to allow a later h5
 601. h3
 651. h4
 701. h5
 751. f3
 801. f4
 851. f5
 901. f6
 951. e4
1001. e5
1051. e6
1101. e7
1151. e8=N  // Also e8=B
1201. d3
1251. d4
1301. d5
1351. d6
1401. d7
1451. d8=R  // Any promotion
1501. c4
1551. c5
1601. c6
1651. c7
1701. c8=N  // Any promotion
1751. b3
1801. b4
1851. b5
1901. b6
1951. b7
2001. b8=B  // Any promotion
2051. a4
2101. a5
2151. a6
2201. a7
2251. a8=Q
2301. CHECKMATE!

I got 2,301 moves (4,601 plies) if I am not wrong, after checking the first moves with SF under Arena GUI in order to make sure I did not mess anything (like 100 or 101?).

How much can you push the limits?

Regards from Spain.

Ajedrecista.

[Ajedrecista]

Hello:

Going backwards with the position of the OP, I can squeeze an additional ply with black to move:

[d]6qk/K7/7Q/6NB/p1p1p1p1/PpPpPpPp/1P1P1P1P/8 b - - 0 1

Code: Select all

   1. ... Qh7  // Qxh7# available
   2. Nxh7  // From here, there will be Nf6# at least
  52. Kxa4
 102. Kxc4
 152. Kxe4
 202. Kxg4
 252. Kxb3
 302. Kxd3
 352. Kxf3
 402. Kxh3
 452. g4
 502. g5
 552. g6 
 553. Bg4  // Once g6 is covered by f7, to allow a later h5
 602. h3
 652. h4
 702. h5
 752. f3
 802. f4
 852. f5
 902. f6
 952. e4
1002. e5
1052. e6
1102. e7
1152. e8=N  // Also e8=B
1202. d3
1252. d4
1302. d5
1352. d6
1402. d7
1452. d8=R  // Any promotion
1502. c4
1552. c5
1602. c6
1652. c7
1702. c8=N  // Any promotion
1752. b3
1802. b4
1852. b5
1902. b6
1952. b7
2002. b8=B  // Any promotion
2052. a4
2102. a5
2152. a6
2202. a7
2252. a8=Q
2302. CHECKMATE!

4,602 plies.

------------

BONUS: I checked the legality of the position with a proof game:

[pgn][Event "Proof game"]
[Site "Madrid (Spain)"]
[Date "2025.08.11"]
[White ""]
[Black ""]
[PlyCount "77"]
[Result "*"]

1. e3 e5 2. Qf3 b5 3. Qxa8 a5 4. Qxb8 d5 5. Qxc8 f5 6. Qe6+ Qe7 7. Qxg8 Qc5 8. Qxh8 g5 9. Nc3 Qxc3 10. Qxf8+ Kd7 11. Be2 h5 12. a3 h4 13. Bh5 h3 14. g3 Qd3 15. Nf3 g4 16. Kd1 e4 17. c3 Qf1+ 18. Kc2 Qxh1 19. Kb3 Qxc1 20. Ng5 Qxa1 21. Qa8 a4+ 22. Kb4 c5+ 23. Ka5 c4 24. Ka6 Qa2 25. Ka7 Qb3 26. Qb8 Qb4 27. Qa8 Qf8 28. Qb8 Ke7 29. Qa8 Kf6 30. Qb8 Kg7 31. Qb6 Kh8 32. Qc6 b4 33. Qd6 b3 34. Qc6 d4 35. Qd6 d3 36. Qc6 f4 37. Qd6 f3 38. Qc6 Qg8 39. Qh6+ *[/pgn]

Regards from Spain.

Ajedrecista.

[Uri Blass]

It is easy to get more moves

Here is the beginning of a possible game

[pgn][Event "Computer chess game"]
[Site "DESKTOP-7QE6S12"]
[Date "2025.08.09"]
[Round "?"]
[White "àåøé"]
[Black "àåøé"]
[Result "*"]
[BlackElo "2400"]
[Time "00:53:57"]
[WhiteElo "2400"]
[TimeControl "60"]
[Termination "unterminated"]
[PlyCount "167"]
[WhiteType "human"]
[BlackType "program"]

1. e3 h5 2. Ke2 h4 3. Kd3 h3 4. Be2 g5 5. Bh5 g4 6. Qf3 b5 7. Qc6 Nf6 8.
Qxf6 Bh6 9. Qxh6 Nc6 10. Qxc6 Ba6 11. Qxa6 Kf8 12. g3 Kg8 13. Nf3 Qf8 14.
c3 Rb8 15. a3 Rb6 16. Qxb6 Qg7 17. Ke4 b4 18. Kd5 d6 19. Kc6 d5 20. Kb7 d4
21. Ka6 d3 22. Ng5 Rh6 23. Qxh6 Qh7 24. Ne4 b3 25. Ng5 f5 26. Ne4 f4 27.
Ng5 f3 28. Ne4 e5 29. Ng5 c5 30. Ne4 c4 31. Ng5 e4 32. Kb5 a5 33. Ka6 a4
34. Nxh7 Kh8 35. Ka5 Kg8 36. Kb5 Kh8 37. Rg1 Kg8 38. Kc5 Kh8 39. Kd4 Kg8
40. Ke5 Kh8 41. Kf4 Kg8 42. Rf1 Kh8 43. Kg5 Kg8 44. Kf6 Kh8 45. Ke7 Kg8 46.
Kd7 Kh8 47. Kc7 Kg8 48. Kb6 Kh8 49. Re1 Kg8 50. Ka5 Kh8 51. Bg6 Kg8 52. Kb5
Kh8 53. Kc5 Kg8 54. Kd5 Kh8 55. Kd4 Kg8 56. Ke5 Kh8 57. Kf5 Kg8 58. Kg5 Kh8
59. Kh4 Kg8 60. Kh5 Kh8 61. Kg5 Kg8 62. Kf6 Kh8 63. Ke7 Kg8 64. Rf1 Kh8 65.
Kd7 Kg8 66. Kc7 Kh8 67. Kb6 Kg8 68. Ka6 Kh8 69. Rd1 Kg8 70. Ka5 Kh8 71. Kb4
Kg8 72. Rh1 Kh8 73. Kc5 Kg8 74. Kc6 Kh8 75. Kc7 Kg8 76. Kb8 Kh8 77. Ka8 Kg8
78. Ka7 Kh8 79. Ka6 Kg8 80. Ka5 Kh8 81. Re1 Kg8 82. Ka6 Kh8 83. Kb5 Kg8 84.
Kxa4 *
[/pgn]

[Ajedrecista]

Hello Uri:

In your example, there is at least a checkmate in one since 27.- Ng5. However, I see that I did not write a requirement that I had in mind, which was that the losing side tries its best once the checkmate/s in one is/are available. In my example, I set up a position with no choices for black, so 'try its best' was like 'try its worst' or simply 'keep moving' and finally forgot to write that requirement, my bad. OTOH, in your example, 27.- ..., Qxh6+ is best for black and white loses the possibility of checkmate in one move. With 34.- Nxh7 in your game, we are basically in the same setup since the fifty-move rule only applies to captures and pawn pushes.

So, good point helping me realising that I forgot a requirement that I had in mind. Thank you.

Regards from Spain.

Ajedrecista.

[Uri Blass]

Hello Uri:

In your example, there is at least a checkmate in one since 27.- Ng5. However, I see that I did not write a requirement that I had in mind, which was that the losing side tries its best once the checkmate/s in one is/are available. In my example, I set up a position with no choices for black, so 'try its best' was like 'try its worst' or simply 'keep moving' and finally forgot to write that requirement, my bad. OTOH, in your example, 27.- ..., Qxh6+ is best for black and white loses the possibility of checkmate in one move. With 34.- Nxh7 in your game, we are basically in the same setup since the fifty-move rule only applies to captures and pawn pushes.

So, good point helping me realising that I forgot a requirement that I had in mind. Thank you.

Regards from Spain.

Ajedrecista.

The same setup but white has more pieces on the board so you can get more plies relative to your diagrams because white can avoid the 50 move rule also by sacrificing the pieces(for example Kg8 Rg5+ Kh8 Rg8+)

[Ajedrecista]

Hello Uri:

The same setup but white has more pieces on the board so you can get more plies relative to your diagrams because white can avoid the 50 move rule also by sacrificing the pieces(for example Kg8 Rg5+ Kh8 Rg8+)

Good point! In your position:

• The rooks: before g5 is played, to allow ..., Kg8; Rg5+, Kh8; Rg8+, Kxg8.

• The DSB: with all the [a-e] pawns promoted, something like ..., Kh8; Bb2+, Kg8; Bh8, Kxh8.

• The LSB: after g6 is played, thus f7 escape square is covered, then ..., Kh8; Bf5, Kg8; Be6+, Kh8; Bg8, Kxg8.

• The b1-knight: with all the [a-e] pawns promoted, something like ..., Kg8; Na3, Kh8; Nb5, Kg8; Nd6, Kh8; Nf7+, Kg8; Nh8; Kxh8.

The five promoted pawns can follow the same plans as the rooks, the bishops or the knight, either being promoted to anything (QRBN, except the e-pawn only to B or N). I propose a new position:

[d]6qk/K7/7Q/6NB/p1p1p1p1/PpPpPpPp/1P1P1P1P/RNB4R b - - 0 1

Since some captures are by white and others by black, avoiding the fifty-move rule might take 100 plies or 99 at most, depending on the case if I am not wrong. Counting as bunches of 50 moves or 100 plies, we had 46 fifty-movers (~ 2,300 moves) in the former position and we add now other 10 fifty-movers (R+R+B+B+N+a'+b'+c'+d'+e', with a' being the a-pawn promoted and so on), that results in 56 fifty-movers (~ 2,800 moves, surely a few less).

Thinking a little more, the f-pawn can be promoted also to B or N in the right moment: ..., Kg8; f7+, Kh8; f8=B {also f8=N}. This gives other 2 fifty-movers (the promotion and the later capture, since the advance to f7 must be followed immediately with the promotion to avoid stalemating). With this manoeuvre, the e-pawn can be promoted to anything as long as the f-promoted pawn is the last to stay at the 8th rank. So, circa 2,900 moves!

2,900... more less like my Elo (wink, wink).

Regards from Spain.

Ajedrecista.