Finding the Invalid Move

Don’t worry, you’re not dumb! You can logically deduce the illegal move well with good reasoning. The trap with not seeing that black can move was there just to trick people.

a8=Q+, b8=R+, c8=B+, d8=N+

Which move is invalid?

put a black king on b7 white pawns on a7 c7 and d7 and the white king on.. h1 say

b8=R+ is invalid

1) if a8=Q+ were invalid

then if b8=R+ were a direct check, the king is on the b file

c8=B+ would then be invalid because it cant be discovered (the diagonal would be on d8, blocking d8=N+)

and to be direct it would attack b7, which has whites pawn!

if the king is on the 8th rank, it would be at least on f8

then c8=B cant be direct (geometry) and not discovered either (again, how the pieces move)

2) if c8=B were invalid

a8=Q is direct (cant be discovered, theres nothing to the left of a8 and b8 only attacks a7), king is on a file

b8=R must be discovered, specifically by a diagonal on c8 attacking a6

then d8=N cant be direct (too far) or discovered (geometry)

king is on 8th rank again atleast f8 (d8 blocks d8=N)

and d8=N isnt direct or discovered, though chess with a nightrider would enable this

3) if d8=N were invalid

a8=Q is direct king is on a file

b8=R has to be discovered via a diagonal on c8, which blocks c8=B

king is on 8th rank

atleast e8

c8=B cant be direct and again we need a nightrider for it to be discovered

whew, that was tedious

Excellent analysis as always. This one was about linking all the 4 promotion types together, and I thought this puzzle would work well, despite the long analysis.

i feel flattered lol

give me a list of 10 moves and my brain would collapse

i find a solution by tinkering then i disprove the others on the fly

but yeah, that was way too long and also tedious given how similar they are (the proofs)

but honestly the queen couldve been a rook and it would have been the same

It needed to be a8=Q+ so that it could check the bKb7. Anyway the idea was to have each type of promotion, and I thought these moves were a nice way of illustrating it, although the logic is not as elegant.

I have seen the previous posts of this thread and some of them seem to be a lot more complicated than mine. For example #74 has like 17 moves and includes fairy pieces. If only the previous members of this thread could return…

aah ofcourse it is a check

and yeah, that stuff was what i was talking about "making my brain collapse" and we definitely need more people here no matter how good their compositions are (at your level, that level, anything inbetween)

this thread deserves to be larger btw

time to re-resurrect this thread again lol

This time it’s a hard one. White to play, 1 invalid:

Ba1#, Ba2#, Ba3#, Ba4#, Ba5#, Ba6#, Ba7#, Ba8#, R1c7, R1c4

No Ba8#

But, move Ra8 to a7 - no Ba7. Seems cooked.

If I counted right, the position is legal: "only" 8 promoted units.

Huh - didn't expect one of the original members of this thread to return!

However, that position technically doesn't work, according to strict algebraic notation. For disambiguating moves, the order of preference for disambiguation is file, rank, then file + rank (see posts 62 and 63). So for R1c4, there are two rooks which can move to c4 but cannot be disambiguated by file (i.e. both rooks are on the c-file). Similar idea for R1c7. In your position, your "R1c7" should be written as Rcc7 and your "R1c4" should be written as Rcc4.

If that still doesn't work for you, then replace R1c4 + R1c7 with Qc4 and there should be 1 unique invalid move.

This one works for the original stipulation. It's only missing 1.Ba6#

It doesn't work for the stipulation with Qc4, so it may not be the intended position.

Nice job: Ba6# is the invalid move! That was the position (particularly with the bishops and Pa6) I had in mind. I’m curious to know: what was your strategy for solving this puzzle?

As for Qc4, it appears I confused you with this: my apologies. Qc4 was just a variant of #109 (with the bishop a-file mates) in case you thought I was being too pedantic on algebraic notation. Next time I will try to be clear with what I mean by R1c4 (and similar moves).

This Qc4 variant is essentially a different puzzle altogether. The moves for that would be “Ba1#, Ba2#, Ba3#, … , Ba8#, Qc4”, again with one of them invalid. I personally like this “Ba1#, Ba2#, Ba3#, etc.” idea as it restricts the seemingly various options available on the chess board to only a few; other moves (such as Qc4) can then influence the invalid move.

Something like this works for Qc4. Again, 1.Ba6# is not possible.

That is correct again! As for the solving strategy, I think there is a proof, but it would be too long for this thread, and there is a lot of experimenting. I suppose trial and error must have taken you hours!

There is in fact only 6 places for the black king, if at most one of “Ba1#, Ba2#, …” is invalid. What I like is that at least one of them is invalid, since the best that could be done is 7 mates but 1 check.

Here are two puzzles:

1. 0-0, Kxd1, Kxf2, Kxe3

2. 0-0, Kxe3, Kxf2, Kxf1

1. Any black unit on f2 either stops castling or protects d1, so no 0-0.

2. If 0-0 is possible, then Kxe3 and Kxf1 cannot be. However, if the King is on e2, a position like this is an illegal double check:

Any substitution of a checking piece protects something. 2. seems impossible.

ive seen this trick before

Indeed you have. I remember introducing this trick at #99 to confuse you for a while. For this position, O-O is invalid because the black king is at e2!