36

The move e2e4 could be played on move 1. (One half-move)

The move e7xd6 could be played no earlier than by black on move 3 (6 half-moves).

What move(s) has the latest "earliest possible move it could be played"?

A move is distinguished by:

  • what type of piece is being moved
  • the square the piece moves from
  • the square the piece moves to
  • whether it is a capture
  • what piece it is being promoted to, if any (probably not actually pertinent, but still)

(And notably, not by whether it causes check or checkmate, nor which colour the piece is, nor which piece [if any] is being captured, nor if it's en passant.)

Rewan Demontay
  • 17,514
  • 4
  • 67
  • 113
Steve Bennett
  • 3,304
  • 16
  • 32
  • 5
    Does the colour matter? Is a white rook moving from a1 to a8 a different move than a black rook moving from a1 to a8? Or is it just "Ra1 - a8", regardless of colour? – Abigail Oct 28 '21 at 01:17
  • 2
    Colour does not matter. – Steve Bennett Oct 28 '21 at 04:23
  • 2
    I would suggest that moves like Bb1-h6 or Ne2-c4 can only happen after move 150 when both players are tired enough, but I suppose that's not the answer you are waiting for... – Evargalo Oct 28 '21 at 09:21
  • 1
    Does stalemate from both sides having insufficient material to mate count as a "move"? – Darrel Hoffman Oct 28 '21 at 15:18
  • I'm not really sure what you mean. I defined what a move is fairly precisely - it's not a condition such as stalemate. @DarrelHoffman – Steve Bennett Oct 28 '21 at 23:25
  • Is a move distinguished for you by en passant? – Mobeus Zoom Oct 29 '21 at 00:16
  • No, because either way the capturing pawn starts and finishes in the same places, and it's a capture. – Steve Bennett Oct 29 '21 at 00:17
  • Using algebraic notation, I think Bgf1 takes at least 19 ply :) – BlueRaja - Danny Pflughoeft Oct 29 '21 at 02:03
  • See the move definition - a move is defined by the piece being moved, where it moves from, and where it moves to. So your bishop move would just be something like Bg2f1, which can be achieved more easily. – Steve Bennett Oct 29 '21 at 04:45
  • 1
    The question is what it is (and it is already interesting enough), but I think adding a + or # to the move, which you explicitely excluded, would make the puzzle even more fun. – Hauke Reddmann Oct 29 '21 at 07:24
  • I think it's very likely that the winning move in that case would be a checkmate so I made a question just for that: https://chess.stackexchange.com/questions/37263/what-checkmating-move-takes-the-most-moves-to-set-up I think it's a much harder problem. – Steve Bennett Oct 29 '21 at 10:59
  • @Evargalo I can't imagine 150 moves as a minimum to reach that point. – Loren Pechtel Oct 30 '21 at 21:04

13 Answers13

16

Evargalo's Ra1-h1 for 15 plies is not correct, since this can be done in 13:

[FEN ""]
1. e3 h5 2. Qxh5 Rxh5 3. Bc4 Rxh2 4. Nf3 Rxh1+ 5. Ke2 Rxc1 6. Nc3 Rxc2 7. Rh1

However, it cannot be done in fewer than 13, making it a reasonable candidate. We can prove this as follows.

Suppose White can do it in 11 plies. This means seven White pieces get off the first rank (either by moving or being captured) in the first 10 plies. Since White only moves on five of those plies, at least two pieces get captured, and that means another ply is required to get the capturing piece off the first rank (or capture it). That is eight different plies accounted for.

Now Black cannot capture on the first rank with either of their first two moves. So all five of White's plies and the remaining three of Black's are needed to get pieces off the first rank. In particular, White makes no pawn moves. But if White makes no pawn moves, Black can't capture on the first rank with any of their first three moves, and now there aren't enough moves left.

Next, suppose Black can achieve it in 12 plies. Now we have one extra White piece to get off the first rank, but the capturing Black piece doesn't have to move off the first rank (as long as it is a rook). So in order to get everything clear in time, we need White to make at most one move which is not taking a piece off the first rank, but Black captures on the back rank with a rook on ply 6. This isn't possible - there's certainly not enough time to capture anywhere other than a1 or h1, and there's no way to get rid of both h pawns in the time (note that capturing one of them doesn't help, since the capturing piece still has to get off the file, and e.g. 1 g4 h5 2 h3 hg 3 hg Rxh1 isn't valid because White has made more than one pawn move).

[edit] In response to Sandro's comment, Rh1-a1 can also be done in 13 plies (and this is optimal for the same reasons).

[FEN ""]
1. b3 e6 2. e3 Qf6 3. Qf3 Qxa1 4. Bd3 Qxb1 5. Nh3 Qxc1+ 6. Ke2 Qb2 7. Ra1

[edit] As pointed out by corsiKa, Ra8-h8 and Rh8-a8 both require 14 plies. The analysis is similar to the above, except for Black to complete in 12 plies requires Black to move at least 4 pieces of the back rank first, with White capturing the other three and then moving off (this has to be done with the queen, since a bishop can't make two consecutive captures on the back rank, and the rooks can't get there in time). This means Black can't make more than one pawn move. That has to be the d, e or f pawn to allow the black king to get off the back rank in one move, and it has to be a light-squared pawn to allow the white queen to capture on the back rank on ply 5. However, if it is the f pawn White must eventually capture the f bishop (it can't move), but this forces the king to move at least twice since it can't stay on f7. So suppose Black moves the d pawn instead. Now White must play 3. Qxc8. If the black queen has already moved, this is check and the king can't move to d7. If not, the black queen is pinned and eventually White will play Qxd8, which forces a second king move. So this can't be done. (It's possible for Black to do either move in 14 plies - just swap sides in the examples above, with White playing any additional first move that doesn't get in the way.)

White can't do it in 13 plies. This would require either Black to make exactly one pawn move with White capturing with a rook on the back rank on ply 7, or two with White getting a rook to the back rank on ply 5, neither of which can be done.

Especially Lime
  • 1,488
  • 9
  • 8
  • Since this uses the queen to make way for the rook, could Rh1-a1 take more than 13 plies? – Sandro Oct 28 '21 at 16:32
  • 1
    @Sandro Good question, but no it doesn't - see edit. – Especially Lime Oct 29 '21 at 07:44
  • What about without any captures or checks? – Simd Oct 30 '21 at 08:16
  • 1
    @Anush without captures or checks White has to move seven other pieces of the first rank themselves. That also requires at least two pawn moves to get the bishops out, and either another pawn move so that the other rook can move forwards, or two moves to get the other rook off the first rank. So the earliest White can make this move is on their 11th move, ie 21 plies. Black can't do it any sooner, because that would require White to move both rooks of the first rank, taking two extra white moves. It's easy enough to do it in 21 plies. – Especially Lime Oct 30 '21 at 08:34
  • 1
    Does it follow that the Black version of these must require 14 plies? – corsiKa Oct 31 '21 at 01:46
  • @corsiKa it doesn't immediately follow (since if Black is trying to do it in 12, instead of White in 11, there's chance for an extra capture), but I believe it is true. I'll edit once I have a convincing argument. – Especially Lime Oct 31 '21 at 09:59
  • Thanks for your answer. One additional suggestion would be Re1-e8, since that requires moving the rook to e1 and getting rid of the kings and pawns on the e-file. I wasn't able to do that in less than 15 plies, can you? – Sandro Nov 01 '21 at 00:25
  • 1
    @Sandro yes: 1 e4 f5 2 Nf3 e6 3 exf5 exf5 4 Bc4 d5 5 O-O Kf7 6 Re1 Nf6 7 Re8 – Especially Lime Nov 01 '21 at 07:18
6

Ra1-h1

This is a similar idea to Rosie's, but using a rook instead of queen makes it longer for the far away side to achieve this move.

I believe it takes 15 plies (edit: actually 13 plies, see comments) for White to play this move :

1.Na3 e6 2.e4 Qf6 3.Qh5 Qxb2 4.Ba6 Qxb1 5.Ke2 Qxc1 6.h4 Qxg1 7.Rh3 Qh2 8.Ra1-h1

[FEN ""]
1.Na3 e6 2.e4 Qf6 3.Qh5 Qxb2 4.Ba6 Qxb1 5.Ke2 Qxc1 6.h4 Qxg1 7.Rh3 Qh2 8.Ra1h1

As a side note, on the other side of the board, Ra8-h8 seems to take only 14 plies :

1.e3 Nh6 2.Qf3 e5 3.Qxb7 Ba3 4.Qxb8 Qh4 5.Qxc8 Ke7 6.Qxh8 a6 7.Qxh7 Ra8-h8

[FEN ""]
1.e3 Nh6 2.Qf3 e5 3.Qxb7 Ba3 4.Qxb8 Qh4 5.Qxc8 Ke7 6.Qxh8 a6 7.Qxh7 Ra8h8
Evargalo
  • 15,979
  • 47
  • 63
  • 8 moves? that can't be right. surely promoting a particular pawn requires more moves. unless i'm misunderstanding something big about the question. – foerno Oct 28 '21 at 18:27
  • nevermind, turns out i was wrong about the pawn. i thought something like fxe8=Q would take longer but it's actually just 5 moves: 1. a4 (1. Na3 g6) (1. f4 d6 (1... g6) 2. f5 Kd7) (1. e4 f6 2. e5 Kf7 3. exf6 Ke6 4. f7 Qe8 5. fxe8=Q) 1... b6 2. a5 Nc6 3. axb6 Rb8 4. bxa7 Bb7 5. a8=Q – foerno Oct 28 '21 at 18:42
  • but wouldn't something like Rb1-h1 require at least 1 extra move? – foerno Oct 28 '21 at 18:48
  • @foerno No: 6... Qxg1 7. Rh3 Qh2 8. Rh1 – Steve Bennett Oct 28 '21 at 23:29
  • 1
    I think this does it in 13 plies 1. d4 e6 2. Na3 Qg5 3. Nh3 Qxg2 4. Bh6 Qxh1 5. Qd3 Qxf1+ 6. Kd2 Qxh3 7. Rh1 – xnor Oct 29 '21 at 02:05
6

Rosie F proposed the move Qh8-a8. I can reach this move in 13 plies (White's 7th move):

[fen ""]
  1. c4 e5 2. Qa4 Qh4 3. Qxa7 Ba3 4. Qxb8 Nf6 5. Qxc8+ Ke7 6. Qxh8 Ra4 7. Qa8

The key idea to is ensure that Black's 6th move is useful, which is accomplished by having the white queen open the way from the black queen's rook to escape.

isaacg
  • 292
  • 1
  • 7
6

Ka8-b8 (14 plies)

Earliest I could find requires 14 half-moves.

[FEN ""]
1. c3 d6 2. Qb3 Bh3 3. Qxb7 Kd7 4. Qxa8 Kc8 5. Qxb8+ Kxb8 6. c4 Ka8 7. c5 Kb8

Other king moves from a8 seem to work out the same.

Ka8xa7 (14 plies)

[FEN ""]
1. Nc3 d6 2. Nb5 Kd7 3. Nxc7 Kc6 4. Nxa8 b6 5. Nc7 Kb7 6. Nb5 Ka8 7. Nxa7 Kxa7

I couldn't prove that less moves is impossible. However, it requires starting with the pawn at d7, then reaching a8 by moving the Black king 4 times. Seems difficult even with help from White pieces.

Steve Bennett
  • 3,304
  • 16
  • 32
murat
  • 176
  • 3
  • That looks pretty good! It definitely has to be at least 12 because 10 for the king moves, and the king can't move on move 1. – Steve Bennett Oct 30 '21 at 21:57
  • 2
    Ka8a7 might actually be one better? – Steve Bennett Oct 30 '21 at 22:12
  • You might be right. Trying it now and could not find a way with 14 yet. – murat Oct 30 '21 at 22:19
  • Found solutions with 14 half-moves for that one as well. Even one that captures a piece in the last move: 1. Nc3 d6 2. Nb5 Kd7 3. Nxc7 Kc6 4. Nxa8 b6 5. Nc7 Kb7 6. Nb5 Ka8 7. Nxa7 Kxa7 – murat Oct 30 '21 at 22:53
  • I'm surprised that the longest series of moves required is to move a the king onto your own side of the board, rather than your opponent's, since moving your king into check isn't allowed, so your opponent would need to clear out their side of the board firs to let you. – nick012000 Oct 31 '21 at 03:18
  • 2
    @nick012000 the point is that it doesn't matter whose king carries out the move (it counts either way), so the quickest way is normally going to involve whichever king started closer to the departure square. – Especially Lime Oct 31 '21 at 10:56
  • 2
    @SteveBennett Ka8a7 can be done in the same number of moves via 1.e3 b6 2.Qf3 d6 3.Qxa8 Kd7 4.Qxa7 Kc6 5.Qb7+ Kxb7 6.Ke2 Ka8 7.Kd3 Ka7. – Rebecca J. Stones Oct 31 '21 at 12:59
  • I suspect any king move starting from a8 similarly requires 14 half-moves. Added one of the examples I found to the answer. – murat Oct 31 '21 at 16:14
  • @murat Revisiting this, did you try Ka8xb7 (the third possible Ka8 capture)? I can't see how to do it in 14 plies. Particularly since the normal path is for white's capturing piece to capture the rook at a8, then move into position to be captured. How could a piece capture a8 on move 4, then get to b7 on move 7, without checking b7 on move 5 or 7? – Steve Bennett Jun 03 '22 at 06:05
  • Oh never mind - doesn't have to be the same piece. Also 14 plies: 1. Nc3 d6 2. Nb5 Kd7 3. Nxc7 Kc6 4. Nxa8 b6 5. e3 Kb7 6. Ba6+ Kxa8 7. Bb7+ Kxb7 – Steve Bennett Jun 03 '22 at 06:13
4

Some more candidates that couldn't break the 13 mark, but may've equaled it, until someone with sharper eyes comes along:

13 Re1-e8: 1. Nf3 e5 2. d4 e4 3. Kd2 exf3 4. exf3 Ke7 5. Bc4 Kf6 6. Re1 Kf5 7. Re8
13 Re2xe8: 1. f4 e5 2. g3 Ke7 3. h4 exf4 4. Rh2 f3 5. exf3 Kd6 6. Re2 Qe8 7. Rxe8
13 Rd2-d8: 1. Nc3 e5 2. d4 exd4 3. Bf4 dxc3 4. Qxd7+ Qxd7 5. Rd1 Qg4 6. Rd2 Qxg2 7. Rd8+
11 Re2-e8: 1. f4 e5 2. g3 Qe7 3. h4 exf4 4. Rh2 Qxe2+ 5. Rxe2+ Kd8 6. Re8+

  • All this reminds me far too much of 3 Check or Giveaway chess or such on chess.com. That would be some variant... play a specific move before your opponent (or perhaps even you each have a specific move to accomplish that is secret???) – JeopardyTempest Oct 29 '21 at 11:01
3

I thought Rb2-h2 could be better than Rb1 etc, but JeopardyTempest found a 12 ply variant: 1. d4 h6 2. Bxh6 Rxh6 3. e4 Rc6 4. f4 Rxc2 5. g4 Rxb2 6. h4 Rh2 (and scuttled other ideas I had)

So, it seems 13 is the fastest possible.

Some analysis: Pawns reach any place in 5 moves (10 ply) pretty easily. Knight takes 4 (8 ply). Bishop seems to need at most 12 ply (Bg1-a7, Be1-Ba5). King does K(x)a8 in 12 ply (white uses knight to remove rook, black moves b7 and d7 pawns and king 4 times). Castling is 5th move (9 ply). So, the only remaining pieces are queen and rook, which have been attempted in many other answers with the maximum of 13 ply. Out of unmentioned possibilities, Qa2-h2 is doable in 12 and I can't think of any other rook or queen move that is a contender for 13+.

Zizy Archer
  • 291
  • 1
  • 7
  • Your 15 actually still has an issue? 5 ... Qxf2 = check? So barring another path, it could be even longer? – JeopardyTempest Oct 29 '21 at 09:36
  • Also if b2-h2 is hard, you would think moves like c2-h2 would be just as hard? – JeopardyTempest Oct 29 '21 at 09:37
  • Though it's still 15 for white: 1.a4 e5; 2.Ra2 Qh4; 3.b4 Qxh2; 4.Rb2 Qxg2; 5.c4 Qxf2; 6.d4 Qxe2; 7.a5 Qh5; 8. Rh2. – JeopardyTempest Oct 29 '21 at 09:41
  • 2
    But hate to tell you, I've got 12 for black to do it, regardless of whether it's a take: 1. d4 h6 2. Bxh6 Rxh6 3. e4 Rc6 4. f4 Rxc2 5. g4 Rxb2 6. h4 Rh2 – JeopardyTempest Oct 29 '21 at 09:45
  • 2
    @JeopardyTempest Oh, I forgot about checks, because I saw faster 14 ply for black so I didn't pay too much attention to white's moves. But thanks for 12 ply one, so mine doesn't work either. I will delete that part of the answer. Hopefully the Re2-e8 isn't broken too. – Zizy Archer Oct 29 '21 at 11:03
  • I've got 13 in my answer for Re2-e8 (it doesn't matter if it's a take or not as black is wasting moves... which is also concerning). But we'll hope together and perhaps fail together on that (and similar d moves)... glad to see someone thinking the same avenue as me at least. – JeopardyTempest Oct 29 '21 at 11:07
  • Does Rb2-h2 actually take 14 plies minimum? If not, finish editing your answer so it doesn't start by saying so. It's not good enough to just have the next paragraph say something else; this question has lots of answers and people are going to want to skim through them. Comments and question history are there if you want to mention that people should look at them for earlier mis-steps. – Peter Cordes Oct 29 '21 at 13:39
3

What about Kh1xh2?

15 plies is achievable:

[FEN ""]
1. h4 d6 2. f3 Qd7 3. Nh3 Qxh3 4. Kf2 Qxh1 5. h5 Qh4+ 6. Kg1 Qg3 7. Kh1 Qh2+ 8. Kxh2

Similarly I also only found a 15 half move solution for Ka1xa2:

[FEN ""]
1. d3 Nc6 2. Bh6 Nb4 3. Kd2 Nxc2 4. Kc1 Nxa1 5. Nd2 Nc2 6. Kb1 Nb4 7. Ka1 Nxa2 8. Kxa2

Let's consider the latter and see if we can do better. If we could, white would only have 7 moves.

The obvious one first: It needs to be the white king. The black king would need 8 moves just to do Ka1a2 on an otherwise empty board.

If we ignore castling for now, the white king has to move 4 times to reach a1 and once more to capture on a2.That only leaves 2 of white's moves to spare. On its way to a1 at least four pieces - the a1 rook and the piece or pawn on the b, c, and d file - either have to move out of the way or have to be captured. Since we established white has only 2 non-king moves this means black has to capture at least two of them and additionally land on a2, all within 6 moves. Can black actually use 2 pieces for that? I don't think so. If black used 2 different pieces, one has to survive and make it to a2. But the fastest black can land any piece on a2 is in 3 moves, namely with the h2 knight. In the remaining 3 moves black can only do one capture tho, since the distance is too large for knights and for queen/bishop one move has to be spent to open up the piece, one move to line up the diagonal (since neither bishop or the queen target anything on the a-d files), which leaves only one capture which we have determined is not enough to help white.

So we determined that black can only infiltrate with one piece.

Further, notice that if white were to clear a1 by themself they need to be careful not to block the path of their King again, i.e., not leave the rook on the back rank which they've spent a move to clear. But also moving the rook along the a-file takes two moves since the a-pawn would have to move first. There is the possibility that black clears the a-pawn, then moves out of the way, white plays Ra2 and black captures on e2 but this simply takes too much time since black needs 3 moves at least to capture on a2 in the first place.

This means the one black piece doing the infiltration must capture the rook and end on a2 at the same time. The only piece that can do this is the queen.

Notice that the first capture (that clears the way for the king) can only occur on black's third move. Also white may not capture the queen afterwards since we already determined one piece has to do all the infiltration and land on a2 at the end. This essentially still blocks the square the queen captured on and even worse, it also blocks all adjacent squares. Lastly, black's queen must be on a2 after black's 6th move but somehow has to allow the white king to reach a1 on white's 6th move. It's quite clear that this implies that black's 5th move can not be in white's queenside since otherwise the queen would block the king's access to a1.

So to summarize (without casling):

  • White has to move the king on all but 2 moves
  • Black must land the queen on a2 on their 6th move
  • Black must retreat the queen on their 5th move to allow the king to reach a1
  • Black can not capture on their first 3 moves

The first and last statement imply that within white's first 3 moves they have to move their king and since black cannot capture they have to make way themself. This leaves only a d-pawn move into Kd2. As soon as the king is on d2 the a3 bishop may no longer be captured since the king would be forced to take black's queen (or run from his target square). This means white has to move the bishop out, essentially locking down the first 3 moves: d3 (or d4) - bh6 (or g5/f4/e3) - kd2. From here white would have to move the king on every move and remember that at this point black still would not have had the chance to capture. It's not too hard to see that this move sequence will not work.


So this leaves castling as our only option. But that does not work fast enough either. After castling, the king needs 3 to make more moves (including Ka1xa2) which leaves 3 moves to clear the back rank. However, clearing the back rank certainly requires 4 moves: d4 - qd3 - bd2 - nc3. Black can also not be of any help since even if they captured on the back rank (which they couldn't even do fast enough) and moved out, they'd cover the square on which they took, disallowing white to castle.


So long story short: I'm pretty sure Ka1xa2 is only possible as the 15th half move and not earlier.

PattuX
  • 131
  • 3
2

I thought I had an answer which was 18 half-moves, but actually only 11 half-moves.

Ra7xh7. Assuming it's black's rook, I think this takes 7 turns to move all the pawns out of the way, one to move the rook, then one more for the capture. Meanwhile, white has to get a pawn up to h7. (Making it a capture prevents white from helping by clearing out one of the other pawns). So, 18 half-moves.

However, if it's white's rook: black moves pawns on c, d, e, f and g. Meanwhile, white plays a4, ra3, rb3, rxb7, rxa7. 11 half-moves.

[fen ""]
  1. a4 c6 2. Ra3 d6 3. Rb3 e6 4. Rxb7 f6 5. Rxa7 g6 6. Rxh7

Steve Bennett
  • 3,304
  • 16
  • 32
2

Qh8-a8.

As a White move this can happen on White's 9th i.e. the 17th ply:

1 e3 h5 2 Qxh5 Nf6 3 Qxh8 d6 4 Qxf8+ Kd7 5 Qxd8+ Kc6 6 Qxc8 Na6 7 Qxa8 g6 8 Qh8 g5 9 Qa8.

It seems not to matter much whether the queen goes from a to h or h to a.

Rosie F
  • 6,305
  • 2
  • 19
  • 36
  • Black can play Qa8h8 on their 10th move, maybe sooner: 1. e4 a5 2. Qh5 b6 3. Qxh7 Ra6 4. Qxh8 Nc6 5. Qxg8 Bb7 6. Qxf8+ Kxf8 7. h4 g6 8. h5 Kg7 9. Nf3 Qa8 10. Bd3 Qh8 (Ah, you changed your answer just as I posted this :)) – Steve Bennett Oct 28 '21 at 06:59
  • Black can do it in 8 moves (16 plies): 1. e4 a5 2. Qh5 d6 3. Qxh7 Bh3 4. Qxh8 Nc6 5. Qxg8 Ra6 6. Qxf8+ Kd7 7. Qxg7 Qa8 8. gxh3 Qh8 – Steve Bennett Oct 28 '21 at 07:01
  • 3
    White can do it even earlier (15 plies): 1 e3 d5 2 Qf3 Qd6 3 Qxf7+ Kd7 4 Qxf8 Na6 5 Qxc8+ Kc6 6 Qxg8 Rd8 7 Qxh8 Rd7 8 Qa8 – Especially Lime Oct 28 '21 at 08:52
1

... Ba1xPb2 entails a black pawn getting to a2, then a1=B. This needs 12 plies. For example:

  1. Nc3 b5
  2. Rb1 b4
  3. ~ b3
  4. ~ bxa2
  5. ~ a1=B
  6. ~ Bxb2
Rosie F
  • 6,305
  • 2
  • 19
  • 36
  • Ooh, that's a clever approach. I think I have an answer that is 17 plies, however. – Steve Bennett Oct 28 '21 at 05:50
  • It does, however, rely on the unit captured being a pawn. But the OP has now clarified that it doesn't matter what got captured, only whether the move is a capture or not. This also makes my next try less good -- this was going to be Ba1xNh8. But Ba1xh8 can happen sooner if it's BxR. – Rosie F Oct 28 '21 at 06:05
  • Yeah, maybe the puzzle is more interesting if the colour and/or the captured piece do matter. – Steve Bennett Oct 28 '21 at 06:11
1

Bbg3

This is what I came up with after seeing the other answers and wondering if promotion could make this take longer and while I haven't explored it too much, I think that this bishop under-promotion idea is definitely one to consider.

[FEN ""]
1. d4 c5 2. dxc5 Nh6 3. c6 Ng8 4. c7 Nh6 5. cxb8=B Ng8 6. Bcf4 Nh6 7. h4 Ng8 8. Bh2 Nh6 9. Bbg3
Steve Bennett
  • 3,304
  • 16
  • 32
JellySword
  • 11
  • 2
  • Yeah, it's a nice thought. The way the problem is worded though, the move is just "any bishop from b8 to g3 without capturing" which can be achieved in fewer moves, without the promotion. Probably 10 half-moves for black, 7 half-moves for white. – Steve Bennett Oct 29 '21 at 00:08
  • Oh ok, I wasn't quite sure since you didn't say if it mattered whether or not there are multiple pieces that can move to the same place, just that the piece being captured (and the color moving) doesn't matter. – JellySword Oct 29 '21 at 00:34
  • See the "from where" and "to where" parts of "a move is distinguished by". – Steve Bennett Oct 29 '21 at 04:40
1

Great question!

Here's another try: bxa1=B+

Edit: Boo! Didn't read the passus about + being irrelevant!

So here's a cheat that needs 16 plies.

[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"]
  1. b4 c5 2. c3 cxb4 3. Qa4 Qb6 4. Na3 bxa3 5. Bb2 Nf6 6. Kd1 h6 7. Kc2 axb2 8. Kb1 bxa1=B+

Peter Fischer
  • 871
  • 3
  • 5
0

15 half-moves are needed for d7xe8Q.

[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"]
  1. d4 d5 2. Nc3 f6 3. Nxd5 Kf7 4. Nxc7 e6 5. d5 Qe8 6. d6 Be7 7. d7 e5 8. dxe8=Q#

Rewan Demontay
  • 17,514
  • 4
  • 67
  • 113
ts.
  • 191
  • 4