# Consecutive captures on the same square

What is the most number of consecutive captures that you can have on the same square in a standard game of chess? Assume that black and white alternate in taking turns. Good luck!

• by consecutive captures, do you mean consecutive moves as well or should just captures be on same square and other non capturing moves are not counted? Commented Dec 25, 2020 at 20:00

Since we are talking about a standard game of chess (although with both players co-operating), we know that there are four pieces that cannot possibly make a capture in the series:

• the two bishops on the wrong coloured squares
• one of the kings (the other can be the last to capture)
• the first piece that gets captured (cannot be any of the above).

Furthermore, getting the maximum number of pawns involved is tricky. It seems best to group the pawns in quadruplets (e.g. all four pawns on A and B files), and then sacrifice 1 pawn to promote the other 3. We can also feed the useless bishops from above to the lane-switching process, each sacrificed bishop allows 1 pawn pair to promote.

Since we can leave two pairs of pawns unpromoted and attacking the same square, and the useless bishops enable 2 more pawn pairs to promote, we will only have to sacrifice 2 pawns to promote the rest of them.

Therefore, the maximum number of consecutive captures on the same square is

26. (32 initial pieces minus the 2 bishops, the two unavoidably sacrificed pawns, a king, and the piece that was the first to be captured)

To confirm that this is possible, let's construct an actual game.

To get started, let's make the unavoidable sacrifices. The plan is to do the captures at f5.

1.  b4   a5
2.  bxa5 c5
3.  d4   cxd4
4.  g4   g6
5.  Bf4  f5
6.  Bg3  f4
7.  h4   fxg3
8.  h5   Bg7
9.  h6   e6
10. hxg7 g2
11. e4


In this position we can confirm that now every piece can eventually reach f5 without any further captures, so it's just a matter of doing the tedious arrangement work: (this could definitely be improved, in efficiency and coherence both)

11. -     Bb7
12. Rh5   g2
13. Rg5   h5
14. c4    h4
15. c5    h3
16. Nf3   h2
17. Nc3   d3
18. c6    h1=N
19. Nh4   Ng3
20. Kd2   g1=N
21. Ke3   b4
22. f4    b3
23. f5    b2
24. Kf4   Nf3
25. Bh3   d2
26. Qc2   Qf6
27. Nd5   Rh5
28. Ne3   d5
29. c7    d4
30. c8=R+ Kd7
31. Qc5   b1=R
32. Qe5   Rb5
33. Rc5   Ra6
34. a4    Rd6
35. a6    Rd5
36. a7    Nh6
37. g8=B  d1=B
38. Bh7   Bc2
39. a8=N  Bc8
40. Nc7   Nc6
41. Ne8   Ne7
42. Ng7   d3
43. a5    Nd4
44. a6    d2
45. a7    d1=Q
46. a8=N  Qd3
47. Nc7   Kd8
48. Nce8  Bd7
49. Nd6   Qf8
50. Ra5


Now all the ducks are properly lined up:

and it's finally time for the fireworks:

50. -     Qxf5+
51. Qxf5  gxf5
52. exf5  Qxf5+
53. gxf5  Bxf5
54. B3xf5 exf5
55. Rxf5  Rdxf5+
56. Nhxf5 Ngxf5
57. Rxf5  Bxf5
58. Ngxf5 Rbxf5+
59. Rxf5  Nhxf5
60. Ndxf5 Nexf5
61. Bxf5  Nxf5
62. Nxf5  Rxf5+
63. Kxf5


As a final note, it's worth noticing that doing black's last capture with a rook is important: trying to do it with a bishop or a knight comes with a nasty surprise: the white King won't be able to complete the final capture, because

the position would suddenly be a draw caused by the "insufficient material for checkmate" rule.

• Just curious Bass, are you a proffesional chess player?😋. Commented Dec 25, 2020 at 15:38
• @Smartest1here Almost! Only short by about a 1000 Elo points or so :-)
– Bass
Commented Dec 25, 2020 at 15:49
• This is very cool! Commented Dec 25, 2020 at 22:42

Note: This is from an actual game, and as such is not likely to be the longest possible.

Longest consecutive series of captures on one square: 12

Weiss - Burschowsky, Austria 1995
37.hxg4 hxg4 38.fxg4 Nhxg4 39.Nhxg4 Nxg4 40.Nxg4 Bxg4 41.Bxg4 Qxg4 42.Qxg4 Rxg4 and 8 moves later, White resigned.

• Some very interesting records there! Thanks for the link. Commented Dec 25, 2020 at 13:08

@Bass comprehensively beat me to it but my solution is significantly shorter, so I'd hate to let it go to waste... Number of captures is the same.

1. d4 e6 2. d5 Be7 3. d6 g5 4. e4 g4 5. c4 g3 6. b4 gxh2 7. g4 f6 8. g5 c6 9. g6 h5 10. Bg5 fxg5 11. b5 h4 12. b6 h3 13. bxa7 b5 14. a4 b4 15. a5 b3 16. a6 b2 17. Ra5 g4 18. Rc5 Qa5+ 19. Nc3 b1=B 20. dxe7 d5 21. f4 Kd7 22. f5 Ba2 23. f6 Kd6 24. e8=B Ne7 25. f7 Nd7 26. f8=N Nf6 27. Bf7 Rh5 28. Nd7 Rb8 29. g7 Rb5 30. Nb6 Bb7 31. a8=N Re5 32. Ne2 g3 33. Nf4 g2 34. Nc7 g1=N 35. a7 Nf3+ 36. Kf2 Ne1 37. Rg1 h1=B 38. Qh5 Bf3 39. Rg5 h2 40. g8=B h1=B 41. a8=B Nc2 42. Bg2 Ne3 43. exd5 exd5 44. cxd5 cxd5 45. Rxd5+ Rexd5 46. Rxd5+ Bfxd5 47. Bgxd5 Bbxd5 48. Baxd5 Baxd5 49. Bxd5 Bxd5 50. Bxd5 Rxd5 51. N3xd5 N3xd5 52. Nfxd5 Nfxd5 53. Nbxd5 Nxd5 54. Nxd5 Qxd5 55. Qxd5+ Kxd5 *

Replay

• Are you able to add a replay link? Commented Dec 25, 2020 at 22:43

Not really sophisticated, but (with promoted pieces) it's possible to do

27 consecutive captures on d5

in the following position:

The sequence of moves can be replayed here.

• Even though OP's comment is weird (it's very common for a game of chess to have multiple queens), this is not reachable from the starting position. Also, if it were, then you could capture once more by placing the black king better.
– Bass
Commented Dec 25, 2020 at 13:14
• Ah, right, there's a maximum to the number of promoted pieces for both sides. It will be quite tough to figure out what the theoretical maximum is, then. Commented Dec 25, 2020 at 13:51
• Yeah... Would not want to be the pawn on d5. @__@
– COTO
Commented Dec 27, 2020 at 18:44