In Geller-Smyslov (Zurich Candidates, 1953), the following position was reached after 16...f7-f5.
Here, White played 17. Qa2, obtaining a passive position that he eventually lost.
In his book on the tournament (translated by Oscar D. Freedman), David Bronstein suggests 17. exf6 Nxf6 18. Bg5 Bxc4 19. Bxf6 Bxd3 20. Rxd3 Rxf6 21. c4, "keeping the knight out of c4 and creating strong pressure on the queenside. Black's win would be far from simple in that case; for instance, 21...Raf8 22. f3 Nb7 23. Rfd1, with Ng3-e4 in view." (See next diagram for this sequence, preceded by each side's 16th moves.)
This PGN could not be parsed.
- move 1 could not be played: Invalid PGN : Error converting move Qc2
[fen "r3nrk1/p4ppp/bpq1p3/n1p1P3/Q1P5/P1PBB3/4NPPP/3R1RK1 w - - - -"]16. Qc2 f5 17. exf6 Nxf6 18. Bg5 Bxc4 19. Bxf6 Bxd3 20. Rxd3 Rxf6 21. c4 Raf8 22. f3 Nb7 23. Rfd1
But White's weak back rank allows Black to impede this regrouping by 22...Qe4! (instead of 22...Nb7):
The queen can't be taken because of mate, and 23. Rc3 Qxc2 24. Rxc2 Rd8 would give Black the open file to go with his extra pawn. So, 23. Rc1 seems forced. Then after 23...Qe5, doubling with 24. Rcd1 would lose material to 24...Nxc4, because of 25. Qxc4 Qxe2 or 25. f4 Ne3! 26. Qd2 Nxd1!. (See next diagram for the latter sequence.)
This PGN could not be parsed.
- move 1 could not be played: Invalid PGN : Error converting move Rc1
[fen "5rk1/p5pp/1p2pr2/n1p5/2P1q3/P2R1P2/2Q1N1PP/5RK1 w - - - -"] 23. Rc1 Qe5 24. Rcd1 Nxc4 25. f4 Ne3 26. Qd2 Nxd1
So, it seems that White should play 24. Ng3, after which 24...Rf4 25. Ne4 Nc6 gives Black an upcoming ...Nd4 and retains control. (See next diagram for the sequence from 23. Rc1 to 25...Nc6.)
This PGN could not be parsed.
- move 1 could not be played: Invalid PGN : Error converting move Rc1
[fen "5rk1/p5pp/1p2pr2/n1p5/2P1q3/P2R1P2/2Q1N1PP/5RK1 w - - - -"] 23. Rc1 Qe5 24. Ng3 Rf4 25. Ne4 Nc6