This is the King's Indian Defense (1. d4 Nf6 2. c4 g6 3. Nc3 Bg7), Samisch Variation where White plays for a strong center with f3 & e4. The early Queen trade (7. Qxd8+ Kxd8) suggests White was playing for a draw, whereas the exchange sacrifice 19. Rd6!? is playing for the win. Typically Black's King is OK in the center after Queens are off, and then the centralized King can be a huge benefit. Aganalian vs Petrosian, 1945 Here Black played 13 ... e6 & 14 ... Ke7 to complete development. Fritz 7 says White has better compensation for the exchange with 22. Bxc5 instead of 22. Bd7 and Fritz is particularly harsh with "?" and "??" yet rarely "!" and almost never "!!" for us humans. Black had potential improvements with 26 ... Rxd5 and 32 ... b5 but the two extra Pawns would have won anyway.

<Do you want ME, of all people, to comment on this game?> Yes. =)

An interesting game ...

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.f3 c5!?;

This is a well-known gambit, Fischer used it against Spassky, (their second match); and Piket used it to defeat Karpov in 1992 in a rapid game. (In a similar line, dozens of GM's have used this gambit.)

Apparently, Torre is a specialist in this line ...

[The cousin of this gambit is:
5...0-0 6.Be3 c5 7.dxc5 dxc5
8.Bxc5 Nc6 9.Qxd8 Rxd8; <"comp"> GM A. Karpov (2651) -
GM H. Nakamura (2704); [E81]
Cap d'Agde CCAS Trophee KO
(R# 2.3), 31,10,2008.
(Black won a nice game, 0-1 in just
under 40 moves here.) ]

6.dxc5 dxc5 7.Qxd8+ Kxd8 8.Be3,

click for larger view

This is one line, it seems to be acceptable for White.

[The "Power-Book" gives the line: 8.Bg5 Nc6 9.0-0-0+ Nd7 10.Nd5 b6 11.Ne2, and White has the better game.]

8...Nfd7!;
With this odd-looking move, Black threatens to wreck White's Q-side with a capture on c3.

9.Nge2 ,
White stops Black's main threat.

[The move, 9.Rd1, is also good for White.]

Both sides continue to develop from here ...

9...Nc6 10.0-0-0 b6 11.f4,

This gains space and perhaps threatens e4-e5, to shut out Black's DSB.

[The (main) book line is: 11.b3 Bb7 12.g3 h6 13.f4 e6!? 14.Nb5 Ke7 15.Nd6, when White is clearly on top.]

11...Bb7 12.g3 Na5 13.b3 e6 14.Bh3,

This looks OK, but is perhaps imprecise.

[Probably better was: 14.Bg2 Ke7 15.Rhg1,
when White remains a pawn ahead, although Black has good play.]

14...Ke7 15.f5?!,
This advance was hasty and wrong, better was the "Power-Book's" suggestion of 15.Rhf1! (White would still be better.)

15...Be5 16.fxe6 fxe6 17.Bf4 Nc6;

click for larger view

Suddenly, Fritz 12 (and other engines) show that White's edge has virtually disappeared. (His pieces rest on odd squares.)

18.Nb5 Nf6 19.Rd6?!, ('?')
This is a deep exchange sack - but it ultimately proves to be unsound. (19.Nd6 looks to be better.)

19...Bxd6 20.Bxd6+ Kf7 21.Rf1 e5 22.Bd7?,
A fun move, but the engines show this to be the losing move here. (All the boxes agree, after this move, Black is winning.)

[Probably better was: 22.Nec3 a6 23.Nd5 axb5 24.Rxf6+ Kg7 25.Be6, "<=>" with incredible counterplay for White.]

22...Nb4 23.Nc7 Rad8 24.Be6+ Kg7 25.Be7 Nxe4 26.Bd5?!,

Another move that is clearly inferior ... or less than best.

[White had to play: 26.Bxd8 Rxd8 27.Rf7+ Kh6 28.Nd5, although Black should still win.]

26...Bxd5 27.cxd5 Rxd5 28.g4?, ('??')

White missed a cool resource here.

[Better was: 28.Ne6+! Kh6 29.Nc3!! Nxc3!? 30.g4!]

The rest is a mop-up operation for Black.

click for larger view

This looks to be pretty drawish. The big girls have had a Bad Blood falling out in the style of Taylor Swift and Katy Perry. Both have harrumphed off the board in a strop. We are left with all eight minors, a symmetrical pawn structure and a quarter of rooks who are probably going to nullify each other through mass exchanges along the open d file.

Portisch evidently doesn't want a draw so he tries to spice things up with a speculative sacrifice ... 19. Rd6?!

We can scoff in these silicon assisted days, but this is a valiant attempt to create imbalances and grab space. Because a few moves later we get to this position (after 23. Nc7):

click for larger view

In return for his exchange, Portisch has much greater space. His pieces have infiltrated the black position, the Nf6 is pinned, the Black e5 pawn is en prise. Is that enough to justify the exchange sacrifice? Fritzie says no, but Portisch was having to rely on instinct and judgement.

Unfortunately for White, he can't do much with his infiltrated pieces. There is too much fresh air around the Black king for White to be able to construct a mating net. White might be able to win the exchange back by harassing the Black rooks, but that would only restore the material balance, it wouldn't be enough to win.

In fact, White finds that his far advanced pieces become targets in their own right. Black carries out the patented strategy of "exchanging everything off", wins a couple of pawns in the process and gives back the exchange.