I have little experience in tournament chess, but here is a strategy drawn from other experience (mainly medical). Adopt the following attitude: <Assume there is a combination until proven otherwise.> If the position really is positional, first convince yourself of that fact.

Instead, White can hold the position quite nicely with 32. Rd5!, when play might continue 32...f6 33. hxg5 h4 34. gxf3 hxg3+ 35. fxg3 c6 36. Ra5 . Here, with three pawns for the piece, White actually has winning chances -- and it is Black who must fight for the draw.>

As I have already pointed out to Wouldpusher above, who posted this same line before, in this line Black wins with the simple 35. ...f2, Queening the pawn and winning.

White's 32nd move was an attempt to save the knight from loss by means of a pin at h4-of course if hxg5 h4 does the trick.

In which 35...f2 is clearly not an option.>

Thanks. I transposed the f3 and f6 pawns.

(This happens to me on occasion, as I learned Descriptive Notation before Algebraic, and still tend to think of each side as having a rank "3", i.e. KB3, etc. When I record a game, it's still easier for me to do it 1. P-K4, N-QB3, etc.)

I first looked at 32 ...Rh1 quickly but then thought that capturing the knight with the bishop leads to mate - until I realized of course that white can capture the bishop with the king instead of the f2 pawn.

So switching back to 32 ...Rh1; 33. Nxh1 g4#

In a jocular spirit, you wrote: "Over the board, on an average day, I would have found gxh3, winning a piece....."

After 32...gxh4, the pawn threatens to take the Knight.

If the Knight moves to safety, Rh1+ would force the King away and allow the f-pawn to promote.

If Kxh4, then the f-pawn promotes.

1: Solve lots of tactical puzzles. Books like 1,001 Winning Sacrifices and Combinations by Reinfeld are good. Yusupov recently wrote a book with lots of exercises for the student.

2: Play over games by attacking players. Tal, Alekhine, Kasparov, Topalov, Polgar, Morphy. That helps you see how positions evolve and how the tactically minded GMs create tactical conditions.

3: I must disagree slightly with JohnLSpouge here. I hear what he is saying, but searching for combinations at every move is a sure way to land in time pressure! Instead, recognize the conditions that must be met for combinations to be possible. And if those conditions are present, allow your imagination to try out many different ways of exploiting them.

Silman lists as conditions needed for combination as:

1: Open or weakened King (including stalemated King)

2: Undefended pieces

3: Inadequately defended pieces.

Hi, <AccDrag>. I defer to your practical experience, which I lack. I agree wholeheartedly that an exhaustive search looking for a combination is a bad idea. My meaning was to have a state of mind that a combination is present, and practice going through some active process to rule it out, which can be honed to become instantaneous, i.e., "intuition".

I second the notion that tactical practice is useful. The emrald tactics chess server is a good way to develop a "preliminary routine" for determining the conditions of a combination. With some practice on emrald, I now scan for legal K moves, loose pieces, and protecting pieces automatically, within the 3 secs before the "opponent" moves. With more practice, I plan to discern the geometric patterns for pins and skewers within the same time. The emrald server complements the puzzles here nicely, precisely because you are heavily penalized for taking time.

<Silman lists as conditions needed for combination as: (1) open or weakened King (including stalemated King); (2) Undefended pieces; (3) Inadequately defended pieces.>

Obviously, I learned most of Silman's conditions on my own. In fact, in CG puzzles, Silman's Condition (1) is woefully inadequate. One needs to add: a local superiority on the K-side (either present or possible). In CG puzzles, local superiority is often the tip-off to throw pieces at the K, before his cavalry arrives. The geometric patterns I mention above are necessary to catch simple combinations that Silman's conditions would miss. I assume his book is for advanced students, however.

I would be interested in further conditions to indicate combinations. Some people, e.g., see geometric patterns indicating a N fork in the air, which I do not fully understand.

To summarize a long-winded post, active development of a preliminary routine slows you at first, but in any activity, good technique requires the development of good foundations.

Hi, <WickedPawn>. Chess is really a specialized problem-solving domain to me, so half my advice is ignorance, but you are welcome to the other half, if you can find it.

<Regarding <johnlspouge>, it's good to hear that thinking like a Grand Master can save a baby's life.>

It actually made no difference to that particular baby how I thought (though I appreciate the compliment). His murmur was an "innocent murmur" (an S3 murmur, to be more precise), which disappears naturally with age, but in the medical profession, certainly, good habits are paramount.

Yeah then I plays the game thinking like black and I'm with him and Rb1+, yes saw it, and f3+, yeah would have played it and, h5, then g5 looking good, nervous about where is that killer move and white goes kh3...yeah, uh,

okay, 32 B x N! I says!

I's wrong.