Does someone know the shortest orthodox chess game such that a single square is directly controlled 16 times? After 1. Nf3 for instance, the square d2 is controlled 5 times. And 2. ... Qa5 pushes the counter to 6. I guess this game (White and Black collaborating) would be around 23-26 moves long?

  • I take it for granted that X-ray attacks don't count? (Because if the "intention" of the stipulation would be a following massacre on that square, X-ray would suffice but White and Black must additionally have the same number of pieces ogling at the square, which is not the case for the examples. Jan 28, 2021 at 12:22

1 Answer 1


Addendum #2: It turns out that a move can actually be saved further still! Michel Caillaud wrote on the Retro Mailing List that 19.5 moves is possible!

[Title "Michel Caillaud, Retro Mailing List 9/18/2020, Non-unique Proof Game in 19.5 moves"]
[FEN ""]
[startply "39"]

1. f4 h5 2. f5 h4 3. f6 h3 4. fxe7 hxg2 5. exf8=N Rh6 6. Nh7 a5 7. Nc3 a4 8. e4 a3 9. e5 axb2 10. e6 b1=N 11. e7 Nxd2 12. Nh3 Ne4 13. Nf2 f6 14. Ng4 Kf7 15. Qxd7 Ra6 16. Qd4 Nd7 17. e8=N gxf1=N 18. Nd5 Nfg3 19. Bg5 Nh5 20. Rf1

Addendum #1 After some back and forth emailing on The Retro Mailing List last night, Eric Angelini has devised a much shorter game than mine in 20.5 moves! This is surely optimal. Good job Eric!

[Title "Eric Angelini, Retro Mailing List 9/18/2020, Non-unique Proof Game in 20.5 moves"]
[FEN ""]
[startply "41"]

1. a4 d5 2. a5 d4 3. a6 d3 4. axb7 dxc2 5. bxc8=N cxd1=N 6. Nb6 Nf6 7. Na4 Ne4 8. h4 f5 9. h5 f4 10. h6 f3 11. hxg7 fxg2 12. Nf3 g1=N 13. g8=N Nxe2 14. Nxe7 Na6 15. Nd5 Nb4 16. Ra3 Na2 17. Rxh7 Qf6 18. Rxc7 Rh3 19. Rc2 Bb4 20. Nd4 Rc8 21. Nb5

For starters, there are a limited number of squares that can be attacked 16 times. A square can only be attacked eight times horizontally and vertically, or four times from each direction. The other eight attacks are from one knight each. A knight can only have its full reach on the central sixteen squares. Therefore, we are restricted to those squares only.

Furthermore, while we have plenty of horizontal and vertical moving pieces in the initial game array, only four knights are present. This means that four more knights must be promoted. The shortest way to do this, theoretically, is to promote 2 knights for both White and Black to split the workload. Promoting four knights already puts us at least 10+ moves in total in addition to the knights' moves to their attacking squares.

Here is my beginning bid of 25.5 moves with the target square of d6.

[FEN ""]
[startply "51"]

1. d4 e5 2. dxe5 d6 3. a4 b5 4. b4 a5 5. axb5 c5 6. Na3 c4 7. Nxc4 Ra6 8. bxa5 Rc6 9. b6 Nd7 10. b7 Nc5 11. a6 h5 12. a7 h4 13. a8=N h3 14. bxc8=N Nb7 15. Nc7+ Kd7 16. Nb5 Qc7 17. Ba3 hxg2 18. h4 Rh6 19. h5 Re6 20. h6 Nf6 21. h7 Ne8 22. h8=N gxf1=N 23. Nxf7 Nd2 24. Nf3 Ne4 25. Nh4 Be7 26. Nf5
  • I have 20.5 for the square c3 : 1. a4 (d5) 2. a5 (d4) 3. a6 (d3) 4. axb (dxc) 5. bxB=N (cxQ=N) 6. Nb6 (Nf6) 7. Na4 (Ne4) 8. h4 (f5) 9. h5 (f4) 10. h6 (f3) 11. hxg (fxg) 12. Nf3 (g1=N) 13. g8=N (Nxe2) 14. Nxe7 (Na6) 15. Nd5 (Nb4) 16. Ra3 (Na2) 17. Rxh (Qf6) 18. Rxc (Rh3) 19. Rc2 (Bb4) 20. Nd4 (Rc8) 21. Nb5 Sep 18, 2020 at 17:13
  • 1
    Michel Caillaud (the Usain Bolt of chess problems) writes (on the Retro mailing list): « 19,5 moves to control 16 times the square f6 is possible ». 2bqN1n1/1ppn1kpN/r4p1r/3N2Bn/3Qn1N1/8/P1P4P/R3KR2 1.f4 h5 2.f5 h4 3.f6 h3 4.fxe7 hxg2 5.exf8=N Rh6 6.Nh7 a5 7.Nc3 a4 8.e4 a3 9.e5 axb2 10.e6 b1=N 11.e7 Nxd2 12.Nh3 Ne4 13.Nf2 f6 14.Ng4 Kf7 15.Qxd7 Ra6 16.Qd4 Nd7 17.e8=N gxf1=N 18.Nd5 Nfg3 19.Bg5 Nh5 20.Rf1 Sep 27, 2020 at 18:53
  • 1
    No, I’m lost in SE rules, procedures, etc. Way much to complicated and user-unfriendly. This will be, I’m afraid, my last appearance — will only lurk from time to time. Thanks anyway James for the good job you’ve done. Sep 28, 2020 at 21:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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