Probably played to avoid 10 Nd2, c4 as introduced by Fischer (vs. Pomar) in the Havana, 1966 event. The text is consistent with White's objective in this opening, i.e., a central pawn advance to disrupt Black's normal flow of development.

In Viktor Korchnoi's Best Games (McKay, 1978, Game 6) Korchnoi states "After 10 ... dxe5; 11 fxe5, Ng4; 12 Bf4, Nxe5; 13 Nxe5, Bxe5;14 Bxe5, Rxe5; 15 0-0 (a pawn sacrifice suggested by Alekhine) the active positioning of White's pieces gives him a lengthy initiative."

In 1967, "book" for average players was The King's Indian Defense by Barden, Hartston & Keene, which gave 11 exd6, a6; 12 a4, Nf6; 13 0-0, Bg4; 14 Ne5, Bxe2; 15 Qxe2, Qxd6 (Lehmann-Toran, Munich 1954) with White slightly better. Rob and his friend Elliot Winslow, a future USCF Master, were "into" the latest opening theory and undoubtedly had access to Schachmatny Bulletin and related sources. The text is a well-conceived positional sacrifice, rendering 9 ... Re8 a waste of time and hemming in Black's Queen's Bishop.

White didn't expect Black would part with the King's Bishop, and Black should have given it more thought!

Black picks up a second pawn and seems to be on the verge of consolidating his position. But development lags (only one piece off the back rank) and Black has seriously weakened his kingside. White wastes no time going on the offensive. Ironically, it is Black who occupies the center but White now demonstrates who controls it!

With his last two moves, White has cleared the way for all his pieces to flow to the kingside.

White is willing to sacrifice additional material to open lines to the Black King.

Black is hard pressed to find good moves. 19 ... Nc6 is met by 20 Ne5, Qd5; 21 Bf4, Rf8; 22 Qh5 and mate soon follows.

If 20 ... Qd3; 21 Qe5+, Kxh7; 22 Bf6, Qd5; 23 Ng5+, Kh6; 24 Qe3, Rf6; 25 Nxe6+ etc. with an overwhelming position. With the text, Black stumbles into a mating net.

If 21 ... Rxf1+; 22 Rxf1, Qxh7; 23 Qg4, Qg7; 24 Bf6, Qxf6; 25 Rxf6, Kh7; 26 Rf7+, Kh6; 27 Qg6 mate.

According to many opening manuals, Fischer's 10 ... c4 (clearing c5 for a knight) is stronger. Queenside expansion by ... a6 and ... b5 is easily thwarted by a4. Another idea is ... b6 without ...a6, allowing ... Ba6.

Does Black intend a Rook lift ( ... Ra7-e7)? For 11 ... Nbd7; 12 0-0, c4; 13 Bf3, Nc5 etc. see Game 4.

Black manages to get all his minor pieces off the back rank, but fails to attack and break-up White's pawn center. With no easy way to open lines for his pieces he is left with a passive postion.

Evidently Black sees something wrong with his plan to double Rooks on the e-file, but this hands White a free move.

Black is desperate for "breathing space" but White's set-up is quite flexible and he can easily shift to Queenside operations.

The text is played to control the a-file, especially a8, and the long diagonal after the center pawns advance. The unopposed white-square Bishop will dominate the position.

White had several choices. 26 Qa2 is a finesse move, as White can follow with Rf3, threatening Rb3 (trapping the Queen) and Ra3, threatening Ra7 or Ra8. 26 f5 shifts the attack back to the Kingside. Of course, 26 Bxf7 wins another pawn. The text, forking Queen and Rook, forces Black to part with the Queen or retreat ... Qb8 and surrender the queenside to White.

Rather than waiting to be crushed on the back ranks, Black sacrifices more material to gain some freedom for his pieces. The endgame that follows is no less difficult, however, as White has an extra piece and a more advanced passed pawn.

White threatens Nxf7+, winning the f-pawn with gain of tempo. If Black captures the knight on e5, White gains a protected passed pawn plus the half-open f-file. White does not rush to play Rb7, as he may want to play Rf1 should the f-file be opened through exchanges.

White blockades the outside passer plus maintaining the threat of a discovered check by the knight. While not a true zugzwang position, Black has just about run out of constructive moves.

Many players will not go into a Benoni unless White has played an early Nf3, i.e., there is no possibility of his adopting the Four Pawn Attack. Black evidently does not fear this line.

Thirty-five years later, the jury is still out on 10 e5 (Game 2.) If White doesn't wish to try this ultra-sharp move, then the text is the best way to protect e4. After Bf3 and 0-0, White can further support the e-pawn with Re1 and Qc2 (or Qe2,) i.e., Nimzovich's theory of over-protection. The QB is shut-in only temporarily; the Knight is headed for c4, after which White will complete his development. Black, on the other hand, does not have such an easy time developing his queenside.

Against Pomar at Havana, 1966, Fischer played 10 ... c4 and steadily took control after 11 Bf3, Nbd7; 12 0-0, b5; 13 Kh1, a6; 14 a4, Rb8; 15 axb5, axb5; 16 e5, dxe5; 17 Nde4, Nfxe4; 18 Nxe4, Nf6; 19 d6, Be6; 20 Nc5, e4! 0-1 in 41 moves. Opening manuals give ... c4! and often note the move was made "to clear c5 for a knight." In the Havana game, Black's knight never lands on c5. Although he won the game, to the best of my knowledge Fischer never played this variation again.

If you go through amateur games on the internet, players with White still mimick Pomar's moves, playing the unnecessarily timid Kh1, a4 too late and/or e5 without adequate preparation.

Black adopts the Fischer strategy with a slightly different move order. White's next several moves are designed to expand control of the center and simultaneously restrict Black's forces. Nimzovich calls this restraint in My System. Moves like 12 Bxc4 or 12 Nxc4 lead to early exchanges that will help Black free his cramped position.

This seriously weakens the queenside. See Game 5 for a practical example of 13 ... Bh6; 14 Qc2 (14 Nxc4, Nfxe4; 15 Nxe4, Nxe4 is unclear-ed) Nd3; 15 Nxc4, Nxc1; 16 Qxc1, Bg4; 17 Qd1 +- analysis by Nunn, as given by Lev Psakhis in The Complete Benoni, 1995. Alternatives such as 13... Qc7 or 13 ... b6 have not produced convincing results.

Creating the most problems by seizing the weak dark squares and disrupting Black's development. Capturing the immobile Black Rook is not that far-fetched, i.e., d6, Bc7, Nb6 etc. Nineteen moves into the game and Black still hasn't solved the problem of developing his queenside pieces.

Black should try to maintain the knight (outpost) by 20 ... f5, even though it hems in the QB. After the text, White dominates the center.

A recurring theme in these games ... Black sacrifices material to free his pieces.

Black must keep his Queen to have any chances for counter-play. The exchange of Queens transitions to an endgame where White holds the advantage.

Thanks to Black's 20th, 22nd and 24th moves, White has the superior endgame. Try to resist thinking in terms of a "won ending," as it implies the stronger side cannot lose and this lulls many players into complacency and ... disaster. If we mentally remove all the pieces except Kings and pawns, even with Black to move he cannot stop a pawn from reaching the queening square. Books on the endgame admonish the superior side to exchange pieces, not pawns. The chess terminology for this is simplification, the theory being the closer the superior side gets to a pure King and pawn ending, the more certain is the victory (and the less chance for error!) But one must exercise some judgment when playing to exchange pieces, i.e., opposite-colored Bishop endgames and many Rook and Pawn endings can be extremely difficult to win.

White's knight on c4 holds everything together. Understandably, Black seeks to eliminate this piece but it costs him the "two Bishops."

Careless. Correct was 28 b4, Bxc4; 29 Bxc4, Nxd6; 30 Bd5 and the Rook is behind the passed pawns where it belongs. Now the Rook impedes their advance, and White will have to waste several moves correcting the situation. In a critical endgame, those wasted tempi could turn a sure win into a draw. Psychologically, this is the kind of mistake that encourages the opponent to fight on rather than resign.

Although Black has managed to recoup one of the sacrificed pawns, White has outside connected passed pawns and, with the two Bishops a greater command of the board. White will try to advance those pawns, taking care that they not become dispersed or blockaded and (eventually) lost. If they reach the 7th or 8th rank, Black will have to sacrifice a piece but White should not allow Black to get both pawns for the piece if at all possible. The knight, an excellent blockader of pawns, is not able to cover both sides of the board efficiently and so Black should retain his long-range pieces. White, of course, is willing to exchange rooks and/or dark-square bishops.

After 34 Rb3?, Nxb6; 35 Rxb6, Ra5 Black has opposite-colored Bishops and White's task is much more difficult. The text forces the desirable (from White's viewpoint) kind of simplification.

On a6, the pawn is pretty much immune from capture and it's threat to advance (Nimzovich calls this the passed pawn's "lust to expand") will restrict one of the Black pieces, which must keep a7 under constant surveillance. Black's only real hope for counterplay is to advance his kingside majority to create a passed pawn. White's strategy is to support the advance of his passed pawns while restraining the advance of the enemy pawns. This will be accomplished by centralizing the King and the two Bishops so they attack as many squares as possible on both sides of the board. By creating threats on both wings, White intends to strain Black's defensive resources to the breaking point. Once the White King reaches the fifth and sixth ranks, either wing, more material will be lost for Black. Controlling the center is no less important in the endgame!

42 ... f5 attempts to create a barrier of squares (d4, e4) which the White King cannot cross. White, however, can simply out maneuver Black's forces. One example would be: 43 Bd5, Kf6; 44 Kc4, Bb6; 45 Bb7, Ke6; 46 Bd2, Kd7; 47 Bb4, Kc7; 48 Bc5, g5; 49 Bxb6+, Kxb6; 50 Kd5, f4; 51 Ke4, Nb5; 52 h4!, gxh4; 53 Kxf4, Nd6; 54 Kg4, Nc4; 55 b4!, Nd6; 56 Kxh4 wins.

Black follows the line analyzed by Nunn in The Complete Benoni. For 13 ... a6? see Game 4 above.

The text is "book," but is it correct to exchange a knight that has moved three times for a bishop sitting on its original square? If Black cannot overwhelm the e-pawn, then it's hard to justify the pawn sacrifice implied by 11 ... c4. Black is cramped and needs to develop counterplay quickly.

The first original move of the game. White achieves slightly better piece placement than with Nunn's suggested 17 Qc1.

After exf5, gxf5 the f-pawn will be very weak, not to mention creating a hole at e6 which White can exploit.

White's overwhelming position indicates the line commencing with 13 ... Bh6 is suspect. Black is burdened with a spectator bishop that just cannot get into the game.

White dominates the central white squares with a clear initiative, so it's no surprise the position is rife with tactical possibilities:
(a) 27 ... Kf8; 28 Nxd6, Bxd6; 29 Rxf6+
(b) 27 ... Rc7; 28 Qxa7, Nxd5; 29 Nxd5, Qxe6; 30 Nxc7, Qe1+; 31 Kh2, Bh4; 32 Qe3 etc.
(c) 27 ... Nxd5; 28 Nxd5, Rxd5; 29 Rxd7.

The "safe" continuation is 28. Qd3, winning the f-pawn.

Of course not 30 ... Nxd7?; 31 d6, winning another piece.

31. Re1, Nh5 followed by ... Ng7 and White has difficulty making progress because the Bishop will get into the game via f6. The text is the real point of Qxc5: White offers to return the exchange, eliminating the troublesome Bishop and setting-up connected passed pawns. It's an offer Black can hardly refuse, else he remains the exchange down and loses the f-pawn for nothing.

White immediately takes steps to centralize his king for the endgame ahead.

Black drives away the defender of c3, anticipating winning White's a-pawn to set-up an outside passed pawn for himself. The problem with this strategy is that it will take many moves to advance the pawn to a1. Black's f-pawn is lost and White's connected passed pawns, far advanced, will strike first. White also has additional resources in that his King or Knight can always catch the a-pawn, and his 2:1 kingside majority can create an additional passed pawn if necessary.

The Petrosian Variation, a positional system with a drop of poison! Petrosian's original concept from the 1940's was to induce ... h6 and .. g5, with play on the weakened white squares. The sharp line with h4, opening the h-file for a direct attack against the enemy king, would come later.

Theoreticians quickly realized 11 ... Nxg3; 12 fxg3, g4 followed by ... h5, keeping the kingside closed for awhile, offers Black better prospects for equality. Refinements for both sides followed and the whole line fell out of favor.

This cannot be sound-Black is opening the g-file directly to his own King. From White's perspective this is well worth a pawn. The loss of castling is not significant, as White's King is relatively secure with the center locked.

Once more we see Black sacrificing material in hopes of obtaining counter-play.

The win is lengthy but not difficult to work out, i.e., 27 ... cxd5; 28 cxd5, Kf7; 29 Rxf5+, Ke8; 30 Qh7, Rh8; 31 Qxg7, Rh7; 32 Nf6+, Kd8; 33 Nxh7, Kd7; 34 Qxe7+, Kxe7; 35 Rg7+, Ke8; 36 Rf8 checkmate.