Guessing once he saw...f6, he remembered that if the now weaker e6 can be exploited it's probably very good.

I don't give a MONKEYS

This match was in 1971 .

Petrosian lost to Bobby ...

One small problem. Fischer won in Buenos Aires in 1970, (Buenos Aires (1970)), before he'd played Taimanov and Larsen. Fischer did beat Petrosian in Buenos Aires in 1971, of course, but that doesn't fit with Schonberg's narrative.

"we started our first game and around the 10th move he threw something at me. He figured to catch me by surprise. But when I looked at the position, I remembered it was something that Steinitz had tried against Lasker in the 1894 championship match. If I hadn't known that position, I might have spent a lot of time figuring it out and maybe I couldn't even have done it on my clock. But once I saw the position, I remembered that I had once analyzed it, and I knew Larsen was dead. When I played the right move, Larsen knew that I knew, and he lost the game and also the next five."

E.g. Schoenberg goes on to relate a 15 minute game he played with Fischer:

<'How about a game?'' I asked. He was amused. I grabbed the white pieces, not even giving him the chance to draw for color - what the hell, he was Bobby Fischer - and played a Queen's Gambit. It was the best game I ever have played. I held out for about 30 moves, and when I resigned, it was with flags flying and bands playing ''The Stars and Stripes Forever.'' I went down with honors. The game took about 15 minutes, of which 14 were mine. He would move instantly, with a bored look on his face.

''Know what was interesting about this game?'' he asked. No, I didn't. ''Up to the 19th move, it was an exact duplicate of a game I played against Mecking in Brazil nine years ago. You and Mecking both played the same 19th move, and it looks natural, but it loses in all variations. Let me show you.'' Fischer swept the board clean, instantly set up the complicated 19th-move position and showed me six variations in a row proving why White must lose.>

Maybe. But Fischer's recollection isn't going to be about some Winawer move order. Likely going to be much more the concept. Which I believe is that black's ...Qc7...c4...f6...Ng6 is aimed at dominating the dark squares in the center. Namely e5. The problem is that it severly weakens the white squares especially e6.

In the Lasker-Steinitz game 3 linked above. Black does get a bind on the dark squares in the center with a nice outpost on e5. Unfortunately, seems the weakness on e6 allows white to get a much better outpost.

Here, this looks like the game is mostly about light squares for white and dark squares for black. Where black does get some play. Noticeable that the black queen only lands on a light square once. With a pawn snatch late in the game. But as in Lasker-Steinitz, the light squares come out on top.

Steinitz vs Lasker, 1896

Position after 12th move of White.

click for larger view

click for larger view

Fisher misses 21. Bd6!, plays 21. Bf3? Loses all advantage, which was overwhelming, all edge now gone.

21...Ne5 22. Qd4 Kg6 23. Rxe5 Qxe5 24. Qxd7 Rad8 25. Qxb7 Qe3+ 26. Kf1 Rd2 27. Qc6+ Re6 28. Bc5 Rf2+ 29. Kg2 Rxg2+ 30. Kxg2 Qd2+ 31. Kh1 Rxc6 32. Bxc6 Qxc3 33. Rg1+

click for larger view

For all that has happened, White has no edge here.

33...Kf6 34. Bxa7

click for larger view

And Black plays 34...g5?

click for larger view