For some reason mates not on the back rank take a little longer for me to see. Might have been something I would have missed in blitz, but this one took more like a minute or two.

Perhaps I should examine checks before captures in my position evaluation?>

You're analysis is always good so I think you're just having an "off" day. Generally checks should be examined first because they are USUALLY more forcing than captures or threats but not always. I think it was Bobby Fischer who said something like "Check, check, check, mate!" And then there is "Patzer sees check, patzer gives check" (no pun intended) but that was referring to moves made...not moves considered.

<eblunt> <I guess I'm just too scared of a big sac which I might miscalculate when I can see a clear win without the sac>

I'm not sure that would help. In some positions, the key move is a check and in others the key move is a capture. If you make a firm rule of exhausting all the checks first then you will waste time when the answer is a capture, and vice versa.

And then there are the positions when the answer is a threat or even a retreating move. And for these we might find that it doesn't matter whether we look at captures or checks first - both would be a waste of time.

So I guess it really doesn't matter which you look at first. I tend to be less structured in the way that I approach these. I don't have a procedure about checks-captures-threats. Instead I look at the move that I think the position needs. And instead of the Kotovian idea of analysing each variation to a conclusion before considering the next, I tend to hop around from variation to variation as I learn more about the position.

There are a couple of things about today's POTD which are a little unusual, and which would help to explain why we might not see the solution straight away.

The first thing is that the puzzle position has several of what Vukovic would term "focal points" for the attack. The square f7 is one clear focal point as we have two white pieces trained on it. Likewise, the g file screams out to be exploited, whether this is on g8, g7 or g6. And finally we have that lonely h6 pawn just waiting to be gobbled up.

With so many focal points it is not surprising that we might follow a couple of false leads before finding the solution.

The second unusual thing about this puzzle is that mates are rare in the middle of the board. If the enemy king is in the corner you need to control four squares for a mate - the three flight squares plus the square he is standing on. And that can be done by K+Q, R+R, or even the dreaded K+B+N.

If the enemy king is on the edge of the board, then we need to cover six squares - five flight squares plus the checking squares. And we often get the assistance of the enemy king's own pawn cover to help. That is why a lone rook can mate on the back rank, if the enemy pawns provide a blocking screen.

But a king in the middle of the board needs all 9 squares to be covered - eight flight squares and the king's own square. And that is why mates away from the back rank are rare and why we can find it hard to spot them. The mating patterns are, almost by definition, unique so we cannot rely on pattern recognition.

Absolutely.

King safety and tactical shots are the main ways to win games. A check is the only truly forcing move. Every check must be looked at in any position.

Also, analyzing checks first is easier. A check must be dealt with. The opponent must take the checking piece, move the King, or interpose. The branches of the analysis tree are easier to deal with when the opponent has few legal moves.

How about 17...Nh8(?) (lol)

However I think that one of the key point to the game was the delay of developpement of the WKN. This allows 23.Nh3 to be a strong move (D)

click for larger view

The obvious point was to get the only good piece on the board. White choose h3 probably because he was not sure of the destination square for the N at this point: f4 or g5?

But W wanted more than only one good piece. He wanted several good pieces. So: 23...Kg7 24.Be3 Rh8 25.Qd2 Ng8 26.g4 (trying to improve the LSB) h6 27.gxf5 gxg5 28.g6! (D)

click for larger view

White waits to occupy f4 with the N in order to do it with maximum of advantage. He sacrifices a pawn and as compensation gets activity for his other pieces AND wins an important tempo whether Black captures or not. If Black does not take the pawn then W plays Nf4 any way. After that a plausible idea (I think!) could be h5, Bf3, Nxd5, Bxd5 and the f5-pawn would eventually fall, as W gets a protected passed pawn as compensation anyway.

28...Kxg6 29.Nf4 Kg7 30.Bh5 (D)

click for larger view

I think that the key point to any of the combinations is that they work due to piece coordination on one side vs. a lack of piece of coordination on the other side.

Thus If we look at the position of the white pieces on the kingside, it is clear that the key to developing a powerful attack is the g file. Unfortunately, black has some control over all of the squares g8-g4! Thus we would seek to remove some defenders. Notice in particular that the g7 square would be ideal for a build up, so we would want the rook out of the way, but to affect that we would need to be on g6 or g8 Thus the key piece to the black position is the knight. on e7.

We could develop a plan to remove that piece, but if we take a cursory glance at black's pieces, it is clear that they are just sitting around waiting to do nothing. Black has no activity and his pieces are getting in each other's way. Thus we can start looking for sacrifices, since black effectively is already down material (note the queen and rook doing nothing!). Thus we can give the material back.