Generally, I prefer fairly natural studies in positions that look as though they might have arisen in a game. Today's is an exception. Composed by the Russian Leopold Mitrofanov, I could have consulted any of a number of sources to refresh my memory but in fact used Endgame Magic, by John Beasley and Timothy Whitworth (BT Batsford, pounds 9.99).
When solving a study like this, it's essential to do so in sympathy with the composer. Unfortunately, the strain put on the board by tactical sequences such as what follows often leads to errors; indeed this is a correction of a previously unsound setting. However, the manic moves - line clearances and obstructions, apparently absurd sacrifices - come straight from (to mix my analysts) the "collective chess id" and should be picked out almost "subconsciously" from it.
Obviously White starts with 1 b6+ when the Black king must retreat 1...Ka8 to retain the defence ...Bb8. If now 2 g7? h1Q 3 g8Q+ Bb8 4 a7 Qa1+. So we get the first "surprise" 2 Re1! to force a blockage on the first rank. The next few moves are natural: 2 Re1! Nxe1 3 g7 h1Q 4 g8Q+ Bb8 5 a7.
Happily the mundane 5...Nd7 is refuted by 6 Qe6 forcing a quick and cheerful mate in all lines eg: 6...Nxb6 7 Qc6+ Kxa7 8 Qxb6+ Ka8 9 Qc6+ Ka7 10 Qa6 mate. Instead, Black must play the much more aesthetic 5...Nc6+ 6 dxc6 Qxh5+.
And now the effect which the composer spent perhaps hundreds of hours to achieve. If 7 Ka6 Qe2+ and so...7 Qg5!! deflecting the white queen 7...Qxg5+ 8 Ka6. With the queen on g5, there is no check and so: 8...Bxa7 9 c7 Qa5+ 10 Kxa5 Bxb6+ 11 Kxb6 and wins - but, without the knight stranded on e1, this would be stalemate. A masterpiece!