Is there a double-forced mate position?

A usual forced mate position is where the winning player can play in a way that ensures checkmate, no matter what (legal) moves the opponent makes (unless the game ends due to a non-play reason, i.e. resignation, draw accepted, time ran out.)

What I'm asking about here is what I'll call a double-forced mate: A position where checkmate is ensured no matter what (legal) moves either player makes.

With this rule alone we can find a position such as this, where white has only one legal move, resulting in checkmate. Let's call this type 1.

Type 1:

``````[FEN "K1k5/1q6/P7/B3N3/8/8/8/8 w - - 0 1"]

1. axb7#
``````

The only option is `axb7#`. However, this isn't too satisfying as a "double" forced mate, as there was only a single possible move to continue. We can explore positions that have multiple possibilities, but still ensure checkmate absolutely.

Type 2:

``````[FEN "k7/2QQQQQQ/KQ1QQQQQ/QQQ1QQQQ/QQQQ1QQQ/QQQQQ1QQ/QQQQQQ1Q/QQQQQQQ1 w - - 0 1"]
``````

White has many legal moves, but every one of them results in an immediate checkmate. Still, I feel this question isn't quite answered, as in a true double-forced checkmate, both players should play a role. (Also, this position clearly isn't possible from a standard board start. And even if it were, it doesn't seem that it could be achieved without first creating a stalemate.) Let's finally look to type 3.

Type 3:

The true double-forced mate: A position where checkmate is ensured no matter what (legal) moves either player makes, however there is at least some line where both players make a move. Meaning not every legal ply in the start position is an immediate checkmating move.

Do you know of (or can you find) any such chess positions of (type 3) double-forced mate? Or can you prove that this is not possible?

Also, share if you have any interesting or more plausible examples of type 1 and 2.

Indeed, there are many possible positions in which all mates are forced on the first ply. Since you ask not for that, I shall provide examples with more than one move by both sides.

Here is the known, and overall, length record for both sides having multiple legal moves since you said that a single line "isn't too satisfying."

``````[Title "Alexey Khanyan, Tim Krabbe's Website Diary Entry #267 2008, Mate In 11 Moves/22 Plies"]
[FEN "4Q2Q/4r3/4n1n1/1bbK1krn/RR1RR1RR/2qn1R1n/4n1nN/Q3Q3 b - - 0 1"]
[startflipped ""]

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: Diary Entry #267

Additionally, I know of this simple problem that illustrates it with either player giving mate.

``````[Title "Eugene B. Cook, The Chess Amateur 1926, Mate In For White And Black"]
[FEN "Bk6/1P6/1P3p2/8/8/2P3p1/6p1/6Kb w - - 0 1"]
``````

Lastly, be sure to check out a funny mutual zugzwang position in this CSE question by the user @Peter, in which both sides are forced to let the other mate them!

• I think this solves the specific wording but not the intention of the OPs question. They may be longer chains, and there are arbitrary choices about which particular knight/rook/queen to move next in places, but both of them are still forced in the same way as the Type 1 example the OP said they wanted to avoid.
– Kaz
Jan 27, 2021 at 18:01
• You could make that first example satisfy the criteria simply by starting one ply earlier - start with the f3 Rook on e3 or g3. Mate no longer guaranteed on the first ply. Jan 27, 2021 at 20:06
• The chess diary entry #267 is brilliant! I took some attempts at this problem, and I had doubted such a long and varied solution was even possible. Jan 28, 2021 at 18:05

Rewan Demontay already has a great answer. However, I would also like to add this 'self-solving' chess problem from Tim Krabbé's Chess Diary #267. Here, all moves are forced and will lead to a checkmate. The composer's full name comes from chess problem databases.

``````[Title "Vilhelm Röpke, Skakbladet 1942, Mate In 6"]
[FEN "K1k5/P1Pp4/1p1P4/8/p7/P2P4/8/8 w - - 0 1"]

1. d4 b5 2. d5 b4 3. axb4 a3 4. b5 a2 5. b6 a1=Q 6. b7#
``````

Although it does not completely satisfy your type 3 requirement that checkmate is not guaranteed in the first ply, I think it is a nice little puzzle that has some similarities with your question.

Inspired by Thijs van Ede's answer, I've modified the pattern from Chess Diary #267

``````[FEN "K1k5/P1Pp4/3P4/1P6/1P6/8/5p2/8 w - - 0 1"]

1. b6 f1=B 2. b7# (2. b5 Bg2+ 3. b7+ Bxb7#)
``````

This position features something rather strange. Still according to the rules, checkmate is ensured no matter what legal moves either player makes. However, it is not guaranteed which player shall deliver checkmate. The outcome depends on what piece black promotes to.

In any case, white can just push `b6` and `b7` to mate before black can respond. However if white delays with `b6` followed by `b5`, black can save themself and checkmate white by choosing the queen or bishop. See the example line or try it yourself on lichess.

A position where checkmate is ensured no matter what (legal) moves either player makes.

This definition does not specify that in all variations, the same side delivers checkmate. If the only requirement is that checkmate by one side or the other is inevitable, and that a drawn outcome is impossible, then the position below fits the bill.

``````[FEN "8/pp1p1p1p/kp6/8/KP6/PP6/3P1P1P/8 w - - 0 1"]
``````

This position was originally designed by Noam Elkies in response to a different question. Imagine a position in which I have a mate in 1, but instead I play a different move that gives my opponent a mate in 1. Call such a move a worst-possible blunder (WPB). Problem: Design a position and a long sequence of consecutive WPBs.

• Here's a 9.5 move sequence: BBk5/P1P2pp1/2P5/8/3Pp3/5p2/1PP2p1p/5Kbb w - - 0 1 Jan 20 at 9:11