Making A Plus In The Petroff
For a long time, the Petroff Defense (also known as the Russian Defense) has been a very solid weapon in situations where a draw is acceptable for Black. Through the end of the 20th century it featured prominently in top level tournaments, especially in the games of Anatoly Karpov and Boris Gelfand.
More recently, that slot (the solid opening against 1.e4) has been taken over by the Berlin Defense, but the Petroff is still quite viable.
1.e4 e5 2.Nf3 Nf6
Against such a solid opening, White cannot expect to achieve a huge advantage by force. What counts as success for White against the Petroff consists of achieving:
1. A rich position.
2. A position with clear plans.
3. The makings of some sort of advantage, i.e. a small space advantage, slightly more active pieces, or an enemy weakness that might be exploitable later.
One cannot expect more than achieving these three things.
I have not met the Petroff very often in my practice. Perhaps this is because I play in open tournaments where everyone is spoiling for a fight. Even when I play against lower-rated players (which is almost all of the time) who would be happy to make a draw, they simply don't have the Petroff in their repertoire because they also are usually fighting for a win in their other games.
On the occasions when I have met the Petroff, I have generally used some obscure sidelines that I had hardly studied and did not fully believe in, such as 3.Nxe5 d6 4.Nc4!?
In a recent tournament, however, I faced the Petroff twice. In the third round of the Silver Lake Open I chose 3.d4 and was happy to see 3...exd4 rather than the more highly regarded 3...Nxe4, and I won the game pretty quickly. But I will instead focus on my sixth-round encounter against IM Rudolf Sertic. I was surprised he played the Petroff (although I should have probably guessed it), and this time I used a different line: 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 d5 6.Bd3 Nc6 7.0-0 Be7 8.Nc3!?
8.Nc3, instead of the main line 8.c4, is quite a practical way to play. White stabilizes the center and relies on his greater piece activity, particularly on the kingside. The c-pawns will be doubled, but this only becomes an important factor in rare cases. Additionally, White gets the chance to make the pawn break c3-c4 later on.
Here, I only knew of 8...Bf5 9.Re1 Nxc3 10.bxc3 Bxd3 11.Qxd3. It seemed logical to exchange the light-squared bishops, but White can fight for an edge there, too. Instead, Sertic chose 8...Nxc3 9.bxc3 Bg4.
Here I was on my own, not knowing anything about this line or the fairly large number of high-level games played from this position. Let's first see how my game went.
Perhaps not the most exciting game. It was quite frustrating too, since it seemed I was very near victory but unable to quite break down my opponent's position. In fact, right after the game it seemed like I had missed quite a few promising chances, although later I found that none of them could be shown to win.
Later, my opponent pointed out that he could have played the immediate 13...Qxd6 rather than inserting 13...Bxf3, since after 13...Qxd6, 14.Bxh7+ does not work -0 14...Kxh7 15.Ng5+ Kg6 16.Qxg4 f5! -- as you can see in the above annotation.
I began to look up games in this line. It was interesting to see the evolution of the players of the white side trying to find an advantage, and the Black attempting to counter this.
First came a really fascinating observation. Originally most players of the white side played as I did, with 10.Rb1 followed by Bf4 and Re1. Normally in chess, including moves like Rb1 and...Rb8 benefits the first player -- the white rook occupies an open file, while the black rook is tied down to defense, at least temporarily. Of course, these are general considerations. There might be a deeper justification either for or against playing Rb1.
Thus between the first instance of this line I have seen (Brown-P.Murray, Vancouver 1971) and 2008, almost everyone played as I did, with the inclusion of Rb1. For instance, the game Anand-Gelfand, Monaco 2006 went as my game did, with the exception that Gelfand realized he could recapture immediately on d6 after 13.Bxd6.
But then in 2008, Veselin Topalov played the same way, with Bf4 and Re1, but without the inclusion of Rb1. Immediately afterwards, all the other top players followed suit. Quickly I discovered the rather deep reason for this.
Let's look at the two positions, with Rb1 and without:
In the former, Black can play 12...Bd6 and meet 13.Bxd6 with 13...Qxd6. Now, as I said above, 14.Bxh7+? does not work: 14...Kxh7 15.Ng5+ Kg6 16.Qg4 f5. But it goes on: 17.Qh4 Rh8 18.Re6+ (not 18.Qxh8 Rxh8 19.Re6+ Kxg5) 18...Qxe6 19.Qxh8.
Now Black wins by 19...Qe7!. The rook on b8 is protected, while the white queen and knight on g5 are attacked.
Compare to the same line without the inclusion of Rb1. Here 11...Bd6 is met by 12.Bxd6. Now 12...Qxd6? simply loses: 13.Bxh7+! Kxh7 14.Ng5+ Kg6 15.Qxg4 f5 16.Qh4 Rh8 17.Re6+ Qxe6 18.Qxh8
Now we see the fantastic justification for the exclusion of the (otherwise useful) 10.Rb1. The rook on a8, unlike in the previous diagram, is not on a protected square. Thus if the black queen moves, White just captures the rook. Black has to instead play 18...Rxh8 19.Nxe6, with a winning ending for White, a pawn up.
This should give you an idea of the depth of the opening investigations by some of the top players. It would be impossible for me to discover the importance of excluding Rb1 over the board, especially in today's time control, where delving into such subtleties in the opening would be extremely impractical. Of course, nobody else had realized this until Topalov did (probably in his home laboratory).
Thus, if Black wanted to play the immediate 11...Bd6, he had to make one of two concessions: either recapture with the ugly 12...cxd6, or throw in 12...Bxf3, as Sertic did, unnecessarily, against me.
The first possibility was used in Topalov's game against Boris Gelfand. It seems to me that 12...cxd6 is the biggest of the possible concessions. It may even be the case that Gelfand did not realize that 12...Qxd6 was not playable until the last moment. Topalov used his opening discovery to fulfill the above three conditions for a successful outcome of the opening and won a smooth game:
While a few other strong players used 12...cxd6, soon it became clear that the loss of a tempo by 12...Bxf3 was a "better" concession. The endgame with 13.Bxc7 as in my game with Sertic was never considered a serious try. Although I managed to push my opponent to the wall and nearly won the game, in general this ending should not be any more than a minute advantage for White. It fails on the first point: a lack in richness.
Meanwhile, the simple 13.Qxf3 has been seen more optimistically than I saw it over the board in my game. Black is very near to equality, but doesn't quite get there. His level of precision must be much higher than White's. For example, Peter Leko won a game against Rustam Kasimdzhanov.
This has remained the "main line" of sorts in this variation. But probably Black can neutralize White's slight pressure. Not surprisingly, players of the white pieces have looked for some alternatives, since this line seems to lack somewhat on the first point: richness of the position. Additionally, Black has been able to delay ...Bd6 with, for instance, 11...Bh5, when White struggled to get any kind of game going.
One interesting attempt was made by Gata Kamsky, who played (after 9...Bg4) 10.h3 Bh5 11.Rb1 Rb8 12.g4!? Bg6 13.Ne5. The point was that after 13...Bxd3 14.Nxd3, White wants to pressure the d5 pawn and the kingside by moves such as Qf3 and Nf4. Bf4 also would come into the equation. It looks good at first, but it turns out that Black is able to neutralize White's pressure. Nevertheless, Kamsky outplayed his opponent and won:
In the last few years, White has tried some other methods. One involved a fairly early c3-c4 -- 9...Bg4 10.Rb1 (aside from the above considerations, 10.Rb1 is still a good move if White does not plan to play Bf4!) 10...Rb8 11.Re1 0-0 12.h3 Bh5 13.c4!? and White has had good results here, but probably Black can handle the pressure. Also, nobody has yet played 13...dxc4 14.Bxc4 b5!?, which seems to practically equalize.
Another quite promising-looking idea is the move Bf5, most likely after 10.Rb1 Rb8 11.h3 Bh5 12.Bf5!?
The idea is to play Qd3, unpinning the knight. By keeping the bishop back on c1, White avoids exchanges. The position stays richer than after some other ways of playing. We will see how the young star Wei Yi used this line to win a convincing game:
So here we have seen some part of the evolution of this line. It was interesting for me to discover (having played this variation completely without any prior knowledge) the various attempts by different players to show that White can fulfill the three criteria of success against a solid opening, and Black's counters.
While one might say that these openings -- such as the one we have witnessed -- where Black merely attempts to suck the life out of the game (or use the threat of doing so to force White to take undue risks) are rather dreary; they nevertheless have a right to exist.
It is a valid and indeed interesting part of chess to find new ways to counter this play, get a full position with prospects of an advantage, win the game, and then maybe our next opponent will play the Sicilian Najdorf.
On the other hand, then we too have our chance to lose.
