click for larger view

<"Both Fischer and I employed this line in earlier days when there was less reliable analysis available to our opponents. Progress shattered this advantage, and ultimately we had to abandon the variation for reasons given in the text of this game.

"With the development of his bishop on e7 instead of the fianchetto, Black's light squares become vulnerable. When the queen bishop goes to b7, obviously it cannot defend both d5 and f5. Consequently, Black must acquiesce to a weakening of the dark squares around his king (...g6) to guard against the intrusion of enemy pieces on f5. All told, Black's task is impossible, which accounts for the line's loss of popularity.">

Other revealing comments after <13.Nh4>:

click for larger view

<"White wastes no time zeroing in on f5. Theory formerly dictated a3 first (timing is paramount) to prevent or at least delay ...b4 driving White's knight from its view of d5. Modern theory arms White with the knowledge that this precaution is a waste of time.

"In passing, we observe that Black's ...Qc7 serves no useful function. Had he substituted ...Nbd7, with the queen on her home base the maneuver of the text could have been prevented (...Nxe4!)."

13...b4?!

"The time gained attacking the knight is illusory. The knight will have no trouble finding a new and equally effective post, and the pawn merely becomes a convenient target for the White queen. Black should have prevented Nf5 with ...g6. Now White has his own way on the light squares.">

I agree with the criticism of the queen on c7, first of all.

And while ...g6 (say 14...g6) might be better, the played 14...Kh8 should hold as well.

The move 15...Ng8 is just wrong, and causes Black to start tripping over his own feet. And it weakens g7 since it denies the g8 square to the rook, which is one of the points of ...Kh8.

(The f6-knight will get help via ...Nbd7, if need be, and the bishop should just go to c5 immediately)

White's 19-23 are fantastic, but 24.Rf3 is overly ambitious - though it does pay off (see diag).

After 23...Nd7-f6

click for larger view

24.Rf3 allows Black to get the queen's off with 24...Qg7!, and Black has approx equality, with compensation for the material. Instead, White really needs to just consolidate with 24.Be6 and the two pawns win, especially given Black's vulnerable king with the queens still on the board.

However, Black played differently, and we see the results.

After Black, already in hot water, plays the ill advised 30...Qxc2 - not so much to grab a pawn, I suspect, as to transfer the queen to the K-side after 31.Rc1 Qf5.

Ah, but Tarjan has a nice M5:

click for larger view