I recall seeing a number of chess games where a lot of pieces are involved in the final checkmate, and was curious to know what is the maximum number of pieces of one player A that have been involved in a checkmate of the other player B, where a piece is said to be involved if and only if removing it either causes B's king to be no longer in check or causes one of the squares around B's king to be no longer attacked by A's pieces. It is easy to construct a reachable position where 9 pieces are involved, and to prove that no more than 9 pieces are involved. It is also possible for all 6 types of pieces to be involved. For example:

[White "Checkmate"]
[Black "9 pieces involved"]
[FEN "6B1/7K/2N5/5kP1/2R1P2P/8/6N1/3Q4 w - - 0 1"]

But are there any actual competitive games that ended with such a checkmate (i.e. using 9 pieces)? If not, what is the maximum number ever used in a competitive game? Among these games, what is the maximum number of types of pieces used? I exclude non-competitive games merely to avoid contrived games.

Note: As John Coleman pointed out in the comments, this notion of "involved" is more like "critically involved" because it is possible that a checking piece is covering another piece that would also provide check if the first was removed, or that a piece next to the checkmated king is protected by more than one piece. One could consider a variant notion of "loosely involved" defined as follows: A piece X of A is loosely involved in a checkmate of B if and only if we can remove some pieces of A that attack B's king or a square next to B's king such that it is still a checkmate and X is (critically) involved. I think this should cover such cases, so I am fine with including examples that use this looser definition of "involved".

  • Looking at the position that you give, a good bet is that the answer will be the end-position of a successful king hunt. – John Coleman May 16 '20 at 12:05
  • 4
    That is a rather odd definition of "involved". Surely a piece belonging to A is involved if removing it would stop B being in checkmate? With this definition it is possible for more than 9 pieces to be involved, if some of them prevent B from interposing another piece. (I think the fact that it is possible to interpose means there will be at most 7 pieces controlling the squares around the king, so you would need at least 3 pieces involved in this way to beat 9, but that is certainly possible.) – Especially Lime May 16 '20 at 14:05
  • 1
    The problem with your definition is that if two pieces are blocking off the same retreat square in such a way that it stays covered if only one of those pieces is removed then under your definition neither of those pieces are involved. To me it seems more natural to say that both are involved. I wasn't sure if that was your actual intention. I see now that it is, but if so, I don't think that it is a very good definition. For one thing, you can set up a position where mate is delivered by double check in such a way that under your definition, no pieces are involved in the checkmate. – John Coleman May 16 '20 at 15:26
  • 1
    Perhaps "essentially involved" might be a term for the concept that you are defining. Whatever you call it, it is clearly an interesting notion. It seems straightforward to say that pieces can be involved in a checkmate without being essential to the mate. – John Coleman May 16 '20 at 15:28
  • 1
    The rook is not involved in your example. If your queen is on d1, however, then it will be (as the rook will be defending the pawn on e4). – Carl-Fredrik Nyberg Brodda May 16 '20 at 22:51

After going through a list of games with king hunts, I found game at last in which 6 pieces are involved in the mate!

[Title "Alexander Alekhine-NN, Simul 34b, Moedling Austria, 3/24/1936"]
[FEN ""]
[startply "33"]

1. e4 e5 2. d4 f6 3. dxe5 fxe5 4. Qh5+ Ke7 5. Qxe5+ Kf7 6. Bc4+ d5 7. Bxd5+ Kg6 8. Qg3+ Kh5 9. Bf7+ g6 10. h3 Qf6 11. Nf3 Be7 12. Qg4+ Bxg4 13. hxg4+ Kxg4 14. Nh2+ Kh4 15. Nf1+ Kg4 16. Be6+ Qxe6 17. f3#

The Bc1 prevents the Black king from moving to f4 and g5, the e4 pawn covers f5, the f3 pawn is giving mate, the Nf1 covers g3, the g2 pawn protects the mating piece and covers h3, and the Rh1 covers the h3, h4, and h5 squares.

Upon seeing this question, I remembered entry #355 of Tim Krabbe’s Chess Diary, which gives some examples of checkmates involving all four minor pieces. The best and simplest example is this game.

[Title "Vasily Borisovich Malinin-Viktor Savinov, Leningrad Russia, 1988"]
[FEN ""]
[startply "71"]

1. d4 Nf6 2. c4 c5 3. d5 b5 4. cxb5 a6 5. bxa6 Bxa6 6. Nc3 d6 7. Nf3 g6 8. g3 Bg7 9. h4 O-O 10. h5 Nbd7 11. hxg6 hxg6 12. Bh3 Re8 13.  Qc2 Rb8 14. Be3 Nxd5 15. Qxg6 fxg6 16. Be6+ Kf8 17. Nxd5 Rxb2 18. Ng5 Nf6 19. Nf4 Qa5+ 20. Kf1 Bxe2+ 21. Kg1 Reb8 22. Bf7 Rb1+ 23. Kh2 Ng4+ 24. Kh3 Rxh1+ 25. Rxh1 Nxf2+ 26. Bxf2 Bg4+ 27. Kxg4 Rb4 28. Kf3 Qa3+ 29. Be3 Qa8+ 30. Bd5 Qa5 31. Nfe6+ Kg8 32. Nc7+ e6 33. Bxe6+ Kf8 34. Nh7+ Ke7 35. Bg5+ Bf6 36. Bxf6#

This is easily deduced to be a pure mate, thus raising the known record to four pieces.

Also, to user @John Coleman from the comments, here’s a king hunt, with four pieces, for you. :-)

[Title "Roland Schmaltz-Ronen Har-Zvi, Internet Chess Club, 3/7/2001"]
[FEN ""]
[startply "48"]

1. e4 Nc6 2. Nc3 e5 3. g3 Bc5 4. Bg2 Nf6 5. Nge2 d6 6. h3 Bb6 7. O-O O-O 8. Kh2 Re8 9. f4 exf4 10. gxf4 Ng4+ 11. Kg3 Nf2 12. Rxf2 Qh4+ 13. Kxh4 Bxf2+ 14. Ng3 Re6 15. Kg4 Nd4 16. f5 Rg6+ 17. Kf4 Rxg3 18. Qf1 g5+ 19. fxg6 Ne6+ 20. Kf5 Ng7+ 21. Kf6 Rxg6+ 22. Ke7 Re6+ 23. Kd8 Bb6 24. Nd5 Re8#
  • 1
    Nice find! Thanks! =) – user21820 May 16 '20 at 14:31
  • 1
    Hey I managed to find a historical game with also 4 involved pieces, but of all different types! =) – user21820 May 16 '20 at 15:18
  • Nice 6-piece mate! Though it's arguably not a competitive game because most simul games are with complete amateurs haha. =) – user21820 May 16 '20 at 17:43

It took a while, but I found via youtube an old well-known game with a king hunt ending in checkmate with 5 involved pieces!

[Title "Josef Matschego − Ernst Falkbeer , Vienna (1853)"]
[FEN ""]
[startply "50"]

1. e4 e5 2. f4 exf4 3. Nf3 g5 4. h4 g4 5. Ne5 Nf6 6. Nc3 d6 7. Nc4 Be7 
8. d4 Nh5 9. Be2 Bh4 10. Kd2 Qg5 11. Kd3 Nc6 12. a3 Bf2 13. Nd5 Bd4 
14. Nc7 Kd8 15. Nd5 f5 16. Nd6 fxe4 17. Kc4 Qd5 18. Kd5 Nf6 19. Kc4 Be6 
20. Kb5 a6 21. Ka4 b5 22. Nb5 axb5 23. Kb5 Ra5 24. Kc6 Bd5 25. Kd6 Ne8#

Interestingly, from move 17 onwards it is a mate in 9, and the only way to achieve that is via sacrificing the queen first!

  • It is even five involved pieces, I believe, or am I double counting? – Vincent May 16 '20 at 22:42
  • @Vincent: Lol! I don't know how I miscounted that! – user21820 May 17 '20 at 3:23
  • Shocked to see Mr Falkbeer playing 2 ...exf4 ! – James Martin May 17 '20 at 9:56
  • @JamesMartin: Can you elaborate on why it is shocking to you? Stockfish 10+ on Lichess at depth 40 thinks it is in fact the best move. – user21820 May 18 '20 at 4:22
  • 1
    @JamesMartin: Ahh I see! So Wikipedia says that Falkbeer "played it in an 1851 game against Adolf Anderssen", so I guess 2 years later he figured out the 'right' move? =P – user21820 May 18 '20 at 8:30

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.