Retro Challenge #3

Honestly I not sure what you mean by economical/cages might explaining this to a patzer to retro challenges?

@Tacticoco: the objective here is not to construct a game which meets the requirements. Rather, it is to create a single position, from which it can be proven that [whatever the challenge was] in any legal game leading to that position. For an example, see my first position on the first page (No rooks on board, neither side in check, promotion to R) and shoopi's analysis of it slightly further down - it is provable that any legal game leading to that position must have included a rook promotion.

Likewise, for your solution, you may have constructed a game where all 4 rooks are promoted (it's not too hard), but I can reach the same position in a legal game without any promoted rooks, so it doesn't solve the challenge.

Oh so the objective not only to meet the requirments but, to make it look good?

Not look good; it is logically provable that the stipulation is met. i.e. the only possible legal games leading to the position meet the condition.

Ohh so the postion can only be reached in a certain order?

Take my position for "no bishops on the board, kings not in check, bishop promotion must have happened".

I can prove from this position alone that a bishop promotion must have occurred. What was the last move? Not the black knight from e3, as that would have already been checking white's king. Not the white queen from g1 or h1 as it would have been checking black's king. Not the white king from g1 or g3 as it would have been next to black's king. Therefore black made the last move with his king.

Where did black's king come from? Not from g1 or g3 as that's next to white's king. Not from h3, as it would have been in check from the pawn on g2, which started from that square and has never moved. So the black king came from h1, where it was in check from white's queen.

What was white's move before that? The queen cannot have moved to deliver that check, so it must have been a discovered check. Not by the white king from g1 as it would have come from next to black's king again. Therefore black's last move would have been to capture a white bishop on h2, which came from g1 to deliver a discovered check; a white bishop is the only piece that can do so.

This white bishop is a dark squared bishop, which couldn't have left c1 because of pawns on b2 and d2. Therefore it must be a promoted dark bishop - therefore in reaching the position, a promotion to bishop must have occurred. We want positions like that, where the position allows for a chain of logic like this to meet the requirements.

I am not sure this answers the task. Retract 0...Kd1xQe1, but can't that wQ be the original one? I see no reason why not.

Edit: nevermind, I just saw the reason why not. :)

No. The position may be reached via hundreds of different legal games. But every single one of those hundreds must have the specified promotion in it. If there is even one possible legal game that does NOT have the required promotion, then the task has NOT been met.

Perhaps a clearer way to say this is: "prove by logical deduction from the rules of chess that white must have promoted a pawn to Q." (Or whatever the task may be.)

Very simple example:

Task: construct a position in which it is logically deducible that it is black's move, and white has promoted a pawn to a Rook.

One way of meeting the task:

It is easy to prove that it's black's turn, since he is in check. The rules of chess don't allow for black to be in check with white to move.

It is also easy to prove that the white R is promoted. A regular Rook could not have moved to a8 to deliver the check. Where would he have come from?

I can even prove that white must have captured something on his last turn. The pawn had to capture during promotion.

If I had not included the a7 pawn, then my position would not fully meet the task, because the white R could have been one of the original ones, and simply moved from (say) a1 to a8 to deliver the check.

Last challenge was a nice one.

Can you construct a position where one side has his king on it's original square, yet it must have made at least 10 moves?

Satoshi Hashimoto will see your 10 and raise you 8:

PG in 24

[This doesn't meet the challenge, but it's a very cool problem and the challenge reminded me of it.]

Fantastic. Of course, you practically gave the solution, but this is a very nice PG indeed.

EDIT: I just realised there's a cleaner position.

The following position is both an answer to the challenge and an exercise in redundancy (x.x). The various tries might highlight a few themes to new solvers.

So, how about a position where one side has a bishop on a home square, but it must have been promoted? (1x Pronkin B; also nice promoted rook shoopi! Clean position.)

As impressive as ever Remellion. It is not at all obvious that the king must have gone all the way to c8, especially in the second position.

As for the latest challenge, it could be done using the same method I used with the promoted rook:

So to make it a little more difficult, could you follow the same challenge, only with the other side still having 8 pawns on the board? This time I have a solution.

I sincerely hope this works. Pretty sure it's not whatever you have on your side.

I also used a retrocage. However, here there is a way to release without needing to promote:

Perhaps a slightly different structure would fix it? Either way, it seems a little more complicated than my own structure.

Changing wRb7->wBb7 is a very artificial way to fix, I believe. I'm still thinking on this one - can't see it.

Yeah that seems to be solid, and it meets the requirements.

g3 must have come from g2. So white played hxg to promote a bishop, capturing the only available black piece, which is a black knight. The black knight is required to finalize the retrocages, which is why when hxg is played there's already a bishop on b7, therefore f1 is the promoted one.

To be more precise, the only way to retract the north-east corner is wRc8 - bNxc8 with the knight getting out, likewise the south-east corner can be released with a bNb1-Na3 etc. Since the other black knight is part of the retrocage, only the knight which is captured can finalize the retrocage, and so on.

I'll go ahead and post my position - there's nothing spectacular here. In fact, I tried to avoid over-complicated retrocages and nuances. I had an idea and went with it, luckily it worked out:

As for the next challenge, I don't have one just yet, I might post one later. Feel free to do so if you have one.