So, today I inadvertently gave myself an interesting problem.

Sometimes I print the diagram to work on away from the computer. However, when I print just the diagram I lose all supporting details such as whose game it is, what move number it is, and who actually is on the move. Today, I printed the diagram without paying attention to who had the move... resulting in a -- "whose puzzle is it, anyway?" puzzle.

I muffed today's solution, but I am awarding myself the fuzzy-lollipop consolation prize for successfully figuring out it was White to move and draw.

Opps! Ok Shamkovich blew the win. *sigh*

White’s position looks so desperate that I suddenly imagined Taimanov mocking Shamkovich after playing 29.Bb7 and saying ‘You don’t catch me, you don’t catch me’ and then Shamkovich taking furiously the bishop just to see 30.Re8#. So, what about 29.Bb7?

A) 29... b2 30.a5 b1=Q 31.Rxb1 Nxb1 32.a6 followed by 33.a7 with a possibly drawn endgame B+2P vs N+3P.

However, after

B) 29... g6 30.a5 Rxb7 31.a6 Ra7 White is lost.

The problem with 29.Bb7 is that it gives much election to Black. Therefore, let’s change the move order to force line A) above with 29.a5:

A) 29... b2 30.Bb7

A.1) 30... b1=Q 31.Rxb1 as the previous A) line.

A.2) 30... g6 31.a6 followed by a7.

B) 29... g6 30.a6 as in line A.2).

Yes, you do! And surely that is all that matters.

Then I correctly figured out that I would somehow take advantage of black's back-rank vulnerability (probably by putting my bishop on the b-file at some point).

White has the initiative and a move in hand since Black must protect his king. White will play a5! and eventually, the pawns and the rooks will leave the table (barring a blunder) leaving a drawn position.

These kinds of positions have been heavily analyzed in the past. With each side having only one minor piece and the pawns facing each other, the strategy of the player with fewer pawns is to exchange them off, and then finally give up his piece for the last pawn.

That is, in fact, exactly what I eventually did. Needless to say, I was well aware that this was not the solution to the problem. I saw the possibility of Bb7 (Black cannot, of course, take the bishop) and figured that White was probably playing for a draw, but I was not aware of the following:

* That after 29 Bb7 b2, White can get out for only one pawn down with 30 a5 b1=Q 31 Rxb1 Nxb1 32 a6 Nc3 33 a7 Rd8 34 a8=Q Rxa8 35 Bxa8, a successful outcome given the bleak outlook.

* That after 29 Bb7? Black does not push the pawn but simply plays 29 ... h6 and wins

* That White can enforce the above idea by starting with 29 a5! (if 29 ... h6 30 a6 and White has tied up the score)

* That after the combination, i.e., 29 a5 b2 30 Bb7 b1=Q 31 Rxb1 Nxb1 32 a6 Nc3 33 a7 Rd8 34 a8=Q Rxa8 35 Bxa8, despite being a pawn behind, White has a theoretical draw (the actual game reaches this setup by a more roundabout course).