# Confused by chess.com puzzle

Can someone explain to me why White even puts themselves in the position of possibly being checkmated by not trading queens and moving the rook?

Then after black's rook grabs the free knight and threatens the queen - why would white take the rook and give up checkmate? Why wouldn't they move their rook to g3, pinning the queen to the king and putting up a fight that as far as I can see leaves them about a Knight down instead of being checkmated?

I'm fairly new - so I must be missing something. But it seems weird to have a puzzle that makes moves that look really bad. When I first took the Knight with my rook, I didn't dream the queen would just take back and give me checkmate. It just doesn't make any sense to me.

So in this puzzle, the White queen is the only piece which is preventing Black from immediately playing Qxf2#.  By taking the knight, Black threatens to take the White queen.  This attack on White's queen crumbles the position.  The queen must stay protecting f2 or there is immediate checkmate.  Even if the White queen attempts to save herself, there is now an unstoppable checkmate threat.

The first move of a puzzle is often a blunder. The blunder is intentionally made by the puzzle program because that's how it works most of the time in real chess games. Unless someone makes a serious error a chess game is likely to end in a draw. Your task is to recognize when an error has been made and to punish it!

Alramech wrote:

So in this puzzle, the White queen is the only piece which is preventing Black from immediately playing Qxf2#.  By taking the knight, Black threatens to take the White queen.  This attack on White's queen crumbles the position.  The queen must stay protecting f2 or there is immediate checkmate.  Even if the White queen attempts to save herself, there is now an unstoppable checkmate threat.

Right - but why not just Rg3 instead of moving the queen. Give it up and force a queen trade and play on. The puzzle's answer seems like the worst answer to me, so I'm confused by why it would play queen takes or moves.

Arisktotle wrote:

The first move of a puzzle is often a blunder. The blunder is intentionally made by the puzzle program because that's how it works most of the time in real chess games. Unless someone makes a serious error a chess game is likely to end in a draw. Your task is to recognize when an error has been made and to punish it!

Got it. That at least makes sense of the first rook move.

As for the rest of the puzzle. It usually seems like the 'opponent' makes the best move they can. Usually when I try to analyze different paths, their move just leads to the same outcome of losing pieces or being mated. But in this one, just moving Rg3 in response to having your Knight taken stops mate and forces an exchange of a bunch of pieces. I thought that would be the puzzle for sure, but instead it was the opponent just giving up mate. That hit me weird, because I thought puzzles were supposed to play the best lines to force you to think it out.

I see what you mean now in your original post.  Yes, Rg3 would be the best move, but it is still losing for Black.  You would be a piece up after all the trades, and you could've used this as a good justification to play the first move, Rxd4.

Like real chess, puzzles don't necessarily involve your opponent playing the "best" moves, but your responses need to be the clearly best moves.  Note also that if you go to the analysis board from the original puzzle that it does have some names and year listed which indicates this puzzle was generated from an actual game in a database.  It is likely that the White played the response Qxd4 in that game.

MrWizzykin wrote:

..... But in this one, just moving Rg3 in response to having your Knight taken stops mate and forces an exchange of a bunch of pieces. I thought that would be the puzzle for sure, but instead it was the opponent just giving up mate. That hit me weird, because I thought puzzles were supposed to play the best lines to force you to think it out.

Yes! The opening blunder is one part of the puzzle logic. Another part is that you must play the best moves, but the puzzle program needs not. It plays the moves which are instructive on the rating level of the puzzle. Many players (sometimes even GMs) overlook mates in one under time pressure and if you miss it you will lose the game. Objectively, the game is won once you captured the knight with 2. Rxd4 and the program chooses to take a risk rather than be tamely slaughtered in the endgame. It's a good tactic for players in lost positions in real games - not just for puzzle machines.