6

What is the longest move sequence that ends with checkmate and every move is a check?

Disclaimer - every move by both players is a check.

  • 6
    I believe that even of the question were answered, the question and answer would be of no value. – Tony Ennis Apr 27 '14 at 14:29
  • 2
    @Tony Ennis: I disagree. Many people find these type of puzzles and constructs to have artistic value, if not practical chess value. – GrizzlyRawrz Apr 29 '14 at 4:48
  • 1
    Why has this question been marked off-topic when several other extremely similar questions have been considered fine and serious chess publications, such as the ones linked in my answer, have found problems with subject matter very similar to this worthy of publishing? I've made a meta post about this issue as well. – GrizzlyRawrz May 1 '14 at 20:30
7

I'm not aware of any resources documenting the specific situation you're asking about. The closest available is probably this page from Tim Krabbé's chess records which documents the longest known series of mutual checks, but not ending in checkmate.

Edit: after reviewing the source above again, one of the puzzles there could have ended in a checkmate. The puzzle by G.Leathem from 1938 could have had 14... Qxb7# instead of 14...Bxb7+, so this would be an example of a checkmate after 27 consecutive checks.

  • 1
    If the link dies this answer dies with it, or so it seems. – hkBst Jan 18 at 8:17
  • And it is not even forced at all, not even in any of ththe meanings of forced. – Rewan Demontay May 7 at 12:15
4

The question asked in the title is different from the question asked in the body.

If it is required that every best option is check, but it is allowed for the players to have non-checking alternatives, then I offer the following. Every best option for White, i.e. every White move which enables White to mate in time against any Black defence (not just checks), is check. White has other moves available which are not check, but they fail to mate in time. Every best option for Black, i.e. every Black move which gives White only one move-sequence which enables White to mate in time, is check. Black has other moves available which are not check, but they either enable White to mate sooner, or give White a choice.

[Title "J. Rotenberg, Problemiste 2012, 1st pr.; PDB 1298074. #11"]
[fen "3Bn1N1/qrrpbppn/1Pppp1p1/2QR1Nk1/3PKR2/8/b2p2P1/8 w KQkq - 0 1"]
[StartFlipped "0"]

1.Nfxe7+ exd5+ 2.Nxd5+ Nef6+ 3.Ndxf6+ d5+ 4.Nxd5+ Nf6+ 5.Ndxf6+ d5+ 6.Nxd5+ Re7+ 7.Ndxe7+ Bd5+ 8.Nxd5+ Re7+ 9.Ndxe7+ f5+ 10.Nxf5+ Qe7+ 11.Nfxe7#

Here's Rotenberg's original problem of which the above is a correction.

[Title "J. Rotenberg, Europe Echecs 1978; Morse, 1st ed., 835; PDB 1298046. #11"]
[fen "3Bn1N1/qrrpbppn/1Pppp1p1/2QR1Nk1/3PKR2/1b6/6P1/8 w KQkq - 0 1"]
[StartFlipped "0"]

1.Nfxe7+ exd5+ (1... cxd5+ 2.Nxd5+ Nef6+ 3.Ndxf6+ d5+ 4.Nxd5+ Nf6+ 5.Ndxf6+ d5+ 6.Nxd5+ Re7+ 7.Ndxe7+ f5+ 8.Nxf5+ Re7 9.Qxe7+ Qxe7 10.Bxe7+ ~ 11.Rh4#) (1... f5+ 2.Rdxf5+ exf5+ 3.Rxf5+ Kg4 4.Rf4+ ~ 5. Qc3+ Kh2 6.Rh4+) (1...e5 2.Rxe5+ dxe5 3.Qxe5+ f5+ 4.Rxf5+) 2.Nxd5+ Nef6+ (2... f6 3.Ndxf6+ {and mate in 9}) 3.Ndxf6+ d5+ 4.Nxd5+ Nf6+ 5.dNxf6+ d5+ 6.Nxd5+ Re7+ 7.Ndxe7+ Bd5+ 8.Nxd5+ Re7+ 9.Ndxe7+ f5+ 10.Nxf5+ Qe7+ 11.Ndxe7#

The sideline after 1... cxd5+ is not best for Black because it gives White the duals 9.Qxe7+ Qxe7 10.Bxe7+ and 9.Bxe7+ Qxe7 10.Qxe7+. And it also gives Black non-checking options which force White to take the full 11 moves.

The sideline after 2... f6 is not best for Black because it enables White to mate in 9.

The above observation, and the choice of sidelines, are by Morse.

Unfortunately this problem is cooked. PDB notes a shorter win in 9.


If it is required that every White move and every Black move is forced, I don't know the answer.

3

Rosie F’s answer is a wonderful one.

However, if you want forced in terms of returning a check with a check, and having to make that move by the laws of chess, I have some answers for you. The positions are legal of course.

This is the record with promoted pieces allowed-22 half-moves.

[Title "Alexey Khanyan, 2008"]
[FEN "4Q2Q/4r3/4n1n1/1bbK1krn/RR1RR1RR/2qn1R1n/4n1nN/Q3Q3 b - - 0 1"]

1... Ng2f4+ 2. Rfxf4+ N2xf4+ 3. Rgxf4+ Nh3xf4+ 4. Rhxf4+ Ndxf4+ 5. Rxf4+ Nhxf4+ 6. Rxf4+ Ngxf4+ 7. Rxf4+ Nxf4+ 8. Rxf4+ Kxf4+ 9. Qee5+ Qxe5+ 10. Qaxe5+ Rgxe5+ 11. Qxe5+ Rxe5+ 12. Qxe5#

Source: https://timkr.home.xs4all.nl/chess2/diary_14.htm, Journal Entry #267

Now, if you want it so each player has only one legal move, I present the following.

One Legal Move With Promoted Pieces-14-Half-Moves

[Title "Noam D. Elkies, 2017"]
[FEN "3b3k/qqqqq3/rr4NK/7R/5N1Q/7Q/B6Q/7R b - - 0 0"]

1... Rxg6+ 2. Nxg6+ Rxg6+ 3. Kxg6+ Qh7+ 4. Rxh7+ Qxh7+ 5. Qxh7+ Qxh7+ 6. Qxh7+ Qxh7+ 7. Qxh7+ Qxh7+ 8. Rxh7#

Here is the same concept but without promoted pieces, and this is what I could find. Please do comment if you can find better, and I will credit you, or if their already is a known better record.

Multiple Legal Moves-8 Half-Moves, Based On Alexey Khanyan’s 22 With Promoted Pieces

[Title "Rewan Demontay, 2019"]
[FEN "8/8/nRb5/1R2PP2/Brk1KP2/1n4r1/8/Q1N5 w - - 0 1"]

1. Rxc6+ Nac5+ 2. Rcxc5+ Nxc5+ 3. Rxc5+ Kxc5+ 4. Qd4+ Rxd4#

Only One Legal Move-6 Half-Moves

[Title "Rewan Demontay, 2019"]
[FEN "8/qB3R2/1b4R1/2p1k3/3P4/2P1K3/1B4r1/Qb3r2 b - - 0 1"]

1... cxd4+ 2. cxd4+ Bxd4+ 3. Bxd4+ Qxd4+ 4. Qxd4#

If you want to know the record for one legal move regarding discovered checks, I can give you this.

I found this funny example on Tim Krabe’s website (Journal Entry #265.)

He gives this series of 7 mutual discovered checks. All of the moves, minus the first, are forced, which is what makes it unique. This also ties for the only legal move, no promoted pieces, sequence.

[Title "V. Korolkov, 1940"]
[FEN "6B1/5Nb1/3p4/q2krP1R/Nn2p3/pPKnr3/Q1PB4/3R4 w - - 0 1"]

1. Nd8+ Re6+ 2. f6+ Ne5+ 3. Bxe3+ Nd3+ 4. b4#

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.