I think it's because you can't legally unpin a piece, but if you block check and put your opponent's king in check, it's legal.
Pinning in Atomic: Confusing Circumstances
Sort:
but technically you should've been able to take because you kill the king before they can
What does lichess analysis board say? legal or not?
The lichess analysis board allows for both capturing the king or the pinning piece. I tried it out at
https://lichess.org/analysis/atomic
at the FEN's
rnbqk2r/ppp2ppp/5n2/8/1b1P4/2N1p3/PPPBQPPP/R3KBNR b KQkq - 1 7
and an artificially created position which could happen from the earlier position, for the sake of testing a position where the king does not explode in that move.
r1bqk2r/1pp2ppp/p1n2n2/8/1b1P4/2N1p3/PPPBQPPP/1K1R1BNR b kq - 1 9
On Lichess, both of these moves are fine, in both of the circumstances, but here, neither in (I'm just assuming for the latter on here) either of them. Lichess is not the definitive judge of the variant, yes, but doing it anyway different from what Lichess does in these cases (although if I'm given the chance, I might try to make this a universal for chess sites) does not make much sense to me.
Dec 2, 2020
0
#7
@Dave
>why was i not checkmated with qd7?
https://www.chess.com/variants/atomic/game/11228035/13/1
because the opponent has a legal move. king captures/explosions are more direct than checkmates which lead to king capture only on the next turn (by which time white will already be dead in this case).
What does li have here? If it's different, I'll say I prefer our rules in this case (we have the same rule in 4player chess)
Lichess allows for exploding your opponent's king out of check, just like here. That rule sometimes is a bit confusing, like when your opponent plays the move, and you don't see the capture that explodes their king, but as it is the only legal move, you'll find it.
Dec 2, 2020
0
#9
From @Hest1805:
I found one case that looks like a bug. In this position: https://www.chess.com/variants/atomic/game/11210376/94/1, white can't move next to the black king, apparently because the black queen is covering b3, b4, b5. But if the black queen didn't cover those squares, white could move next to the black king, leading to a draw. As far as I can see, there is no difference between the two. Notably also, if the two kings are already next to each other, they can step into check from other pieces as long as they remain right next to each other.
What do you think? What does li think? 
Lichess allows for moving your king next to your opponent's king, even if the king would be attacked from a third piece (in this example, the queen). Atomic only-king endgames are confusing, but I think black could still win that game even after Kb3-5. Envision it like this, BK on h1, BB on h2, WK on g2, white to move. This position is Zugzwang for white. White moves his king to f3 or h3(as the king crucially can not move to f2 because of the queen), and black responds with Qg2. After separating the kings, it's an easy win.
edit: I forgot to mention that the black queen would be on g1
Should a move that would normally be illegal, but because the pinning piece would explode after that move be legal?
Shouldn't you be able to take the king with a pinned piece? I thought that was a settled matter, a pinned piece can still give check (that is, threaten to take the king), so why shouldn't be able to capture/explode the king?
The attached game has an example of both of these on the same move, but the move(s) in question (exf2/exe2) could not be played. Is this a bug because of unforeseen circumstances, or is it one of those "features"?
https://www.chess.com/variants/atomic/game/11217687