Chess puzzles?

that will cost you either the bishop or the knight.

Sorry, best I can do is a quarter knight.

No correct suggestions so far. But then again this is the most difficult chess puzzle I know of and even engines won't find the right move.

Here, have another difficult one.

(white to move)

Is there really a path without variance in OP pic? I think i could win this but i can't seem to find a 100% certain way of doing so.

There is one specific move in my first puzzle that turns things around for white. If you tell me the very first move you'd make I can tell you if you are correct.

Pawn to g3

My best guess so far is either Pawn to g4, threatening Bishop c2, or a King move.

Bishop to C2

eh g4 i mean. silly me.

DING DING DING! We have a winner! Well done!

I can't claim full credit for it, because I've seen the puzzle before, but IIRC, the solution is

Bc2+. If ... Kxh5,
d8=Q, Nf7+
Ke6, Nxd8+,
Kf5.

And black is pretty much fucked. The king is stuck ,and sooner or later white gets his bishop to either the d1-h5 diagonal or the e8-h5 one, either of which is mate. You can stall for a surprisingly long while with pawn advances and knight moves, but it's just delaying the inevitable. Other initial king moves or not playing Nf7+ to get rid of the queen means white just kills you with it.

I think the solution is

bg, h1=Q
gq=Q+, Bb8
Rxe5,

and while black can delay you by playing checks, you can march your king towards the center, and eventually force either a trade of queens or for your rook to come crashing down to join in the pin crushing Black's king. There might be a better line of defense for black I'm missing though.

pawn to a7

do chess engines really fuck that one up? Bc2 seems like the only real logical move

Those initial pawn moves came out fucked up somehow. For the starting moves of the second puzzle, I'm saying

g7, h1=Q, b8=Q+

I'm moderately dyslexic and sometimes read the board position backwards, and initially wrote in b7, a1, b8, and then when fixing it on more careful diagramming from the bottom right square's color, I deleted the number instead of the letter because I'm also an idiot.

They won't usually see the full continuation no, and then usually play Nf4+ and then lose. And it's pretty tricky to see the full continuation from the start. Most people try to come up with a way to salvage the queening pawn on d7.

Yup. Even when searching with depth 30 Stockfish 8 won't find the move, while at the same time evaluating that black is at a huge advantage. Other engines aren't any different. But once that move is made engine suddenly realize that white is at a huge advantage.

Still no correct move for the second puzzle.

Is there some defense for black I'm not seeing? Because b7 and the resulting queening looks pretty convincing from where I'm standing.

Wouldn't surprise me, Engines still operate by mostly calculating advantage over time, especially in the late game. Sacrificing a knight and a queen seems like an awful lot, for not much gain in the short run, for the amount of time white needs to win in the long run.
There are some surprisingly simple chess problems that engines still can't solve.

If white plays b7 white gets ripped apart by black.

Pawn g7

Nope.

Rook e2.

Oh, I see it, although I still think that black's best is a draw, not a win. But you can shut down all those check shenanigans right off the bat with

Re1. That keeps the queen off the checking diagonal for another turn since he has to move his knight away. So you get something like

Re1, Nxe1,
g7, h1=Q
g8=Q+, Bb8
And now you can play a7 without fears of spoiling checks.

That is, black's best continuation against the suboptimal b7 play is a draw, not a black win.

Re1 is correct! It prevents the black queen from being a pain in the ass and buys you time.

White playing b7 is definitely a win for black.

I don't see how. The continuation I gave in forcibly leads to white having a king, queen, rook, and 3 pawns against a King, Queen, knight, and Bishop. Unless there's a VERY elaborate forced pin that I can't see leading to black winning the queen, you've got a draw, not a win. Even then, you're going to need damn near perfect play to pull it off.

After b7 black plays Ka7 and now white can't do anything while black builds up momentum.

a7 is covered by the pawn on b6, the only legal king move black has is Kb8, and after which you lead to b8=Q+, Bf8, Qxf8#.

And I am again fucking retarded, not seeing the difference between b and g because retardation. Yeah, b7 (near the king) probably leads to a black win, I didn't even really consider it. I was talking b7, which I still maintain draws, if not wins.

Friendo if you move your pawn to b7 then a7 is no longer covered.

g7. Brb, killing myself.

lol but yes, g7 would indeed lead to a draw.

just get the fucking queen, trade your officer if he tries to get through to, and then just chek him until mate

If d8=Q, then black plays Nf7+ and takes the queen after you have to move your king. That leaves you each with a knight and a bishop, only he has 4 pawns to your 1. Plus, his threat of pawn to e2 is not something you can easily ignore.