Since there is nothing more forcing than a double-check, one's first thought should be 1 . . . Rg3+ 2. Kh4 g5+ 3. Kh5, but what next? There is no way to get at the white king, and the black rook is now attacked. Pausing briefly to realise that 1 . . . Rh4+ 2. Kxh4 g5+ 3. Kh5 is also not mate, the next try should be 1 . . . Rg6+ when 2. Kh4 Rh6 is mate. Since 2. Nd7+ allows 2 . . . Rxd7] when 3. cxd7 Bxd7+ forces mate again, this looks promising. But it's not the answer. After 1 . . . Rg6+, White plays 2. g4] when 2 . . . fxg3+ can be met by either Nd7+ or Kg2.
At this stage, there is a strong temptation to curse one's luck and take the knight back; after all, White's king cannot escape the discovered check. It is very easy to play 1 . . . Kxe5 and leave White to solve his problems, but that would be too lazy. After 1 . . . Kxe5 2. Rhe1+ Kf6 3. Be6]] Bxe6 4. Rxe6+] Kxe6 5. Kxg4, White emerges with the better endgame.
Pinter found a brilliant way to close the net around the cornered king, playing 1 . . . Kg5]] Now 2. Nxg4 Bxg4 is mate, so Portisch forked one rook, then pinned the other with 2. Nf7+ Kh5 3. Be2. Now came the real point of the combination: 3 . . . Rd3+] 4. g3 (4. Bxd3 allows Rh4 or Rg3 mate, while 4. Bf3 loses to Rxf3+) 4 . . . f3] Now 5. Bxd3 Rh4 is mate, while 5. Bxf3 Rxf3 6. Kg2 Rxf7 leaves Black a piece ahead. The game concluded with a fine attempt at the consecutive check world record: 5. Rc5+ Rg5+ 6. g4+ Bxg4+ 7. Kg3 fxe2+ and White resigned.
The late Mikhail Tal could often be seen glaring at a position, almost intimidating it into revealing a forced win for him, but that is only one side of the problem. The hardest work, as the above example shows, can be in finding your opponent's defences to the tempting lines that do not work.
(Graphic omitted)Reuse content