Dann Corbit wrote:
Unfortunately, there are two other clearly winning moves, and so it is not a great test position. On the other hand, it was really fun to analyze it.
Yes, this was from a Tal's game.
So let's make it a really great test position where computers can't solve it (look at each variation i give, it has truly marvelous different combinations each one):
Let's just move the King to a3. That means:
White to play 1.Rf6!! and win. Now other moves are not winning easily as before....
[d]r1b1q1k1/1p2rppn/p3p2p/3pP2Q/3P4/K1PBN2P/PP4P1/R4R2 w - - 0 1
Now again 1.Rf6!! wins since we have:
1...gxf6 2.Bxh7 Kxh7 3.Ng4 Qf8 4.Rf1 and now:
--4...f5 5.Nf6+ Kg7 6.g4! Re8+(6...Rc7+ 7.Kb3 and now on 7...Qh8 we have 8.g5 1-0 and on 7...Bd7 8.Rg1 +-) 7.Kb3 Qh8 8.Rg1 fxg4 9.hxg4! +-
--4...Re8+ 5.Kb3 f5 6.Nf6+ Kg7 7.g4 +-
1...Nxf6 2.exf6 Rc7 3.Ng4 Qf8+ and now:
--3...e5 4.fxg7 and now:
----4...Qe7+ 5.Kb3 Rc6 6.Nxh6+ Kxg7 7.Ng8 Qc7(7...Qe6 8.Qg5+ Kf8 9.Rf1 e4 10.Rf6 Qe8 11.Be2 Bd7 12.Bh5+ Be6 13.Qh6+ +-) 8.Qh7+ Kf8 9.Qh8 Be6 10.Nf6+ Ke7 11.Qxa8 e4 12.Rf1 Rb6+ 13.Ka3 Qd6+ 14.b4 Qc6 15.Kb2 exd3 16.Rf2 black is almost zugzwang and it's 1-0(for example 16...a5 17.Ng8+ Kd7 18.Qxa5 Ra6 19.Qc5 1-0)
----4...Rc6 5.Nxe5 Rf6 6.Re1 a5 7.Kb3 a4+ 8.Kc2 Kxg7 9.g4 ±
----4...Bxg4 5.Qxh6 f5 6.hxg4 e4(6...Qe7+ 7.Kb3 Qxg7 8.Qe6+ Qf7 9.Qxe5 Re7 10.Qg3 +-) 7.Qh8+ Kf7 8.Qh7 Qg8 9.Qxf5+ Ke8 10.Qg6+ Kd7 11.Rf1 Rd8 12.Qf5+ Kc6 13.Bxe4 Rxg7(13...dxe4 leads to a nice endgame win) 14.Bf3 Rg5 15.Qf6+ ±
--3...Qf8+ 4.Kb3 Bd7 5.Rf1 Bb5(5...Rc4 6.Kc2! wins immediately.) 6.Bxb5 axb5 7.fxg7 Qxg7 8.Rf6+ +-
1...Qd8 2.Ng4 Nf6 3.exf6 Rc7 4.Kb3! 1-0
1...Qf8 2.Bxh7! Kxh7 3.Raf1 a5(3...Bd7 4.Ng4 g6(4...Bb5 5.Rxh6 1-0) 5.Rxg6 fxg6 6.Rxf8 gxh5 7.Nf6+ Kg6 8.Rxa8 +-) 4.Ng4! a4 5.b4 axb3 6.Kxb3 Rc7 7.a4 1-0
1...Rc7 2.Bxh7! Kxh7 3.Raf1 g6 4.Qh4 +-