What is the longest fixed sequence of consecutive moves you can find in any legal position that

  1. guarantees checkmating the opponent regardless of their responses (whether by forcing particular responses or otherwise), AND
  2. is the shortest available fixed sequence of consecutive moves that checkmates the opponent in that position?


Have a look at the position below:

[White "Checkmater"]
[Black "Checkmatee"]
[FEN "r5r1/1pq1k1Bp/3pb1p1/1p1P1p2/8/5N2/PP2QPPP/R3K2R w - - 1 18"]

1.Qxe6+ Kd8 2. Bf6+ Qe7 3. Qxe7+ Kc8 4. Rc1+ Kb8 5. Qc7+ Ka7 6. Bd4+ Ka6 7. Qb6#

In theory White has M5: 1.Qxe6+ Kd8 2.Bf6+ Qe7 3.Bxe7+ Ke8 (3...Kc7 4.Qxd6+ Kc8 5.Rc1#) 4.Bf6+ Kf8 5.Qe7#. But White's third move has to be different to achieve these mates: no one move works in both cases of 2...Ke8 and 2...Kc7. (For example, try 3.Bxd6+: with this, 2...Ke8 3.Bxd6+ comes with M2, and 2...Kc7 3.Bxd6+ comes with M5, yet neither of the two mates can be achieved by a fixed sequence of moves.)

However, White has the fixed sequence of moves shown above which guarantees checkmate (satisfying condition 1): 1.Qxe6+ Kd8 2.Bf6+ Qe7 3.Qxe7+ Kc8 4.Rc1+ Kb8 5.Qc7+ Ka7 6.Bd4+ Ka6 7.Qb6#. At 7 checkmater moves long this also appears to be the shortest fixed sequence available in the position above which guarantees checkmate (satisfying condition 2) - for example, we can create longer but equally forced sequences that also satisfy condition 1, such as trivially extending the mate above for example with 3.Qxe7+ Kc8 4.Rc1+ Kb8 5.Qc7+ Ka7 and now instead of 5.Bd4+, play e.g., 5.Qa5+ Kb8 6.Qc7+ Ka7, etc., and thereby add a few moves to lengthen the sequence while keeping it fixed and mate certain at the end.

Note that there is no necessity (in principle at least) for the opponent's moves to be forced. Condition 1 is satisfied as long as - no matter what Black chooses, if they have any choice - the same continuation for White (the checkmater) still results in checkmate.

Above I showed a position where the shortest fixed consecutive series of moves by White that guarantees checkmating Black is 7 moves long. (Not that I claim this as any kind of record.) I would like to see positions where there exists a sequence that satisfies these criteria and where that sequence is as long as possible - I'm guessing 8,9 White moves is small fry for some of the people here.

  • @RewanDemontay I think the OP has made it clear what is meant by "fixed sequence". The phrase "regardless of their responses (whether by forcing particular responses or otherwise)" suggests that black's responses may vary and "fixed consecutive series of moves by White" suggests that white's moves may not vary. – Brian Towers May 1 at 8:31
  • @BrianTowers: I interpret it the same way. Still, I think it might be formulated clearer, I place the versions I can think of in different comments for clarity. This question: "Give the longest sequence of moves for White that can be played regardless of Black's answers, and ends in #." (Current record: see Rewan's answer) – Hauke Reddmann May 1 at 9:52
  • Variant question: "Give a sequence of moves for White that can be played regardless of Black's answers, and ends in #, where the number of possible subgames during this sequence is maximal." (Current record: my attempt) – Hauke Reddmann May 1 at 9:56
  • Another variant: "Give the longest sequence of moves for White that can be played regardless of Black's answers." (Answer: infinite, but this seems interesting if the 50 move rule or threefold repetition is considered.) – Hauke Reddmann May 1 at 9:58
  • 2
    I suspect the intent of the question is "what is the longest checkmate that can be pre-moved (but can't be premoved any faster)" – QuadmasterXLII May 1 at 20:34

In other words, you wish for the record length for a completely dual-free sequence. Every White move must be absolutely unique to any possible by Black. Shorter, dualed mating sequences do not count.

To that end, I remembered a mate in 61 from this page on SuperProblem.ru (a Russian based chess problem site). It is also in the Die Schwalbe Chess Problem Database.

[Title "Bosko Miloseski, SuperProblem.ru 1/24/2020, Dedicated to the participants of TT-235, Mate In 61"]
[FEN "4Q3/1ppr4/br1k4/1p1p4/1PpP1p2/KBp2p2/2P2P2/8 w - - 0 1"]

1. Qe5+! Kc6 2. Qe6+ Rd6 3. Qe8+ Rd7 4. Ba2 Kd6 5. Qe5+ Kc6 6. Qe6+ Rd6 7. Qe8+ Rd7 8. Bb1 Kd6 9. Qe5+ Kc6 10. Qe6+ Rd6 11. Qe8+ Rd7 12. Ka2 Kd6 13. Qe5+ Kc6 14. Qe6+ Rd6 15. Qe8+ Rd7 16. Ka1 Kd6 17. Qe5+ Kc6 18. Qe6+ Rd6 19. Qe8+ Rd7 20. Ba2 Kd6 21. Qe5+ Kc6 22. Qe6+ Rd6 23. Qe8+ Rd7 24. Kb1 Kd6 25. Qe5+ Kc6 26. Qe6+ Rd6 27. Qe8+ Rd7 28. Kc1 Kd6 29. Qe5+ Kc6 30. Qe6+ Rd6 31. Qe8+ Rd7 32. Kd1 Kd6 33. Qe5+ Kc6 34. Qe6+ Rd6 35. Qe8+ Rd7 36. Ke1 Kd6 37. Qe5+ Kc6 38. Qe6+ Rd6 39. Qe8+ Rd7 40. Kf1 Kd6 41. Qe5+ Kc6 42. Qe6+ Rd6 43. Qe8+ Rd7 44. Kg1 Kd6 45. Qe5+ Kc6 46. Qe6+ Rd6 47. Qe8+ Rd7 48. Kh2 Kd6 49. Qe5+ Kc6 50. Qe6+ Rd6 51. Qe8+ Rd7 52. Kh3 Kd6 53. Qe5+ Kc6 54. Qe6+ Rd6 55. Qe8+ Rd7 56. Kg4 Kd6 57. Qe5+ Kc6 58. Qe6+ Rd6 59. Qe8+ Rd7 60. Kf5 Kd6 61. Qe6#

This works since Black basically has only one legal move for the entire mating sequence. There are no alternative lines that might spoil it with White dual mates and such.

  • Very nice find. If I had a secondary interest (besides length of the fixed sequence), it is that the sequence should be as close as possible to the shortest possible mate (which allows multiple choices by Black - I think you call these 'duals'?). The sequence you provide coincides with that mate. I also note it is relatively easy to adjust, if desired, for the 50-move rule (you can just remove some of the White King's obstacles on its path to f5). – Mobeus Zoom May 2 at 19:44
  • 1
    This is really nice. I could maybe imagine a longer version where the king isn't sufficient to mate, but instead the king has to go all the way up to H7 to capture a pawn allowing its own pawn to promote to a knight, which then goes around the back to deliver mate in some convoluted spot. – Steve Bennett 2 days ago
  • How does the addendum work? I thought 6. ... Rc3+ is a valid move by black, which makes white next pre-move invalid? – justhalf 2 days ago
  • @justhalf "Pre-moves" were just an analogy. – Rewan Demontay 2 days ago
  • I mean, it has to be a sequence of fixed move by white, right? So regardless of black move, white needs to make the same set of moves and still achieve checkmate. In 6. ... Rc3+ case, white cannot make the same move as in 6. ... Rc2 case to achieve mate, since Qe1 is illegal. (and also, I believe this is the definition of pre-move, not sure what you mean by pre-move being only an analogy) – justhalf 2 days ago
[FEN "k4b2/P1p1p3/KpP1Pp2/1P3P2/3pBp2/3P1P2/8/8 w KQkq - 0 1"]

This may be more "in the spirit": White has a forced 12# by 1.Bd5 2.Bb3 3... 12.Bb7# and Black a ton of variants (didn't count...) but White doesn't care about Black's moves.

EDIT: To increase the number of subgames, all black pawns except c7/b6 could be bishops. (If they move too early, White has a shorter mate, but this is hardly relevant, since Black tries to avoid shorter mates in a problem by definition.)

  • Thank you, thematically this seems somewhat similar to RewanDemontay's solution (most of the board is frozen, and a piece has to reach the appropriate square from which it can deliver/assist with checkmate), except with fewer moves but more subgames. – Mobeus Zoom May 2 at 19:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.