Friday, October 29, 2010

In defence of the Petroff Defence

If the position above - White to play - occurred in one of your games, what would you think about it? Would you think "how dull", would you feel contempt for your opponent for playing into such a dull position or shame at yourself for allowing it? Would you offer or accept a draw?

Or would you think that this was a challenging, a position, open, asymmetrical, plenty of pieces still on the board, very easy to lose this one quickly if I don't keep my wits about me?

Of course you wouldn't. Because this is the Petroff. The position above is probably the most crucial position in current Petroff theory. And everybody knows the Petroff is a drawing opening. Unless they actually look at it.

It cropped up in Nanjing, on Wednesday, in the game Gashimov-Wang Yue. Naturally it resulted in mass exchanges and a draw. In fact everything was exchanged until they got right down to the bare kings, and a draw on move 66.

It's not a battle royal or masterpiece, it's just a hard-fought grandmaster game like any other, in an opening like any other. Except most openings don't get talked about as if their employment was a sign of moral turpitude, or as if they represented a threat to the existence of the game rather than just another way of starting it.

Me, I like the Petroff. I've always liked it, ever since I saw David Hooper's Complete Defence To 1.P-K4 on my father's shelves when I was young, the first opening book I can remember seeing. In my beginners' books - and of course in my beginner's games - you only ever saw 2...Nc6, and so for me 2...Nf6 was from the start something new rather than something dull. It was something different, clever, off-the-wall.

What's wrong with it? It copies. By copying, it annoys. When you were a kid, did you ever annoy anybody by copying exactly what they did? I still do this, as it happens, and I'm 45. Don't! "Don't!" Stop it! "Stop it!" It's enough to get anybody's nerves, and for a chess opening, that's a start. It gets on people's nerves, the Petroff. I like that about it too.

But they say it's boring. It's not like the sort of game you get from the Semi-Slav, for instance, where lines like 1. d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nc3 e6 5.Bg5 lead to the most fantastic complications. Lines favoured by players like Topalov, who like to like to sacrifice their knight on f7 as early as possible, as in Topalov-Kramnik, Wijk aan Zee 2008.

Or do I mean Topalov-Kramnik, Linares 1999? It's so hard to tell.

I mean, let's go back to that thrilling, mind-boggling Semi-Slav line again. Suppose White had deviated a little earlier, specifically on move three? What was Black going to do for winning chances after 3.cxd5? (As it happens, even 5.cxd5 can be hard to play against, but let's keep this simple for the moment.) Horror! All of a sudden, instead being in the thrill-a-minute Semi-Slav, we're in the handshake-in-a-moment Slav Exchange.

So whose fault is that, assuming the language of blame has any relevance at all here? White's, presumably, for taking on d5, since we've already established that Black's up for a fight if White wants one. But in that case, isn't the same true if White plays for a draw in the Petroff?

If this doesn't seem quite right to the club player, it may be because the relevant drawing variation in the Petroff is much more drawish, at club level, than its Slav equivalent. I'd have a little bet, for instance, that the most common result at club level, in the Slav Exchange, is a win for Black, since it's most commonly employed against the higher-graded by people hoping for a draw. I've not seen the statistics on which I'd base that bet: I have, however, seen it happen any number of times in club and tournament games. It's not a strategy that tends to work.

The Petroff equivalent, however, goes like this: 1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.Qe2

and after 5...Qe7 6.d3 off the queens go early doors. Not quite as drawish as Gashimov and Wang Yue's bare kings, of course, there's still a game to be played, but if you were trying to win a game, you'd rather take the Slav Exchange, wouldn't you?

I don't think I've ever - unlike the Slav Exchange - seen this on a real-life chessboard in a serious game. Maybe at club level nobody ever has this position, because they avoid the Petroff in order to avoid it. Fair enough: I'd do the same. But it doesn't follow that when somebody plays the Petroff as grandmaster level, they're doing so because they're inviting the opponent to play right down it. On the contrary, it's because they're assuming the opponent won't.

If Kramnik plays 2...Nf6 against Topalov, it's because he's got every reason to think Topalov is going to play something more testing. And if Topalov isn't in a mood to play, he's going to play down a drawish line no matter what Kramnik does. Because, at the highest level, what makes the game hard to win is not that you're playing the Petroff. It's that you're playing Black.

But at that level, you don't normally waste a White. Or, for that matter, waste a draw offer if you've got Black and you get one nice and early.

That's the draw-hating Topalov who favours the Berlin Wall, by the way, which as much as anything points up the absurdity of attaching moral criteria to openings when played by super-grandmasters, or indeed of attaching them to preferences of style and approach.

If it's true that the Petroff is an unwise choice against the lower-rated player, because it makes it too hard to take advantage of your strength, it doesn't mean it's a cowardly choice when employed by a top ten player against another top ten player. It just means you - you, not them - shouldn't play it when it isn't suitable. Play it against the better players, pick something else against the weaker ones. This doesn't fit with the practice of most ordinary players, since we like to concentrate on playing just the one defence. But it's viable, if the openings are sound and you play often enough.

I don't play it myself, not at the moment. I gave it up as a kid - stupidly, for Sicilians about which I understood nothing - and then played it a little again in my twenties, until a game against Adam Hunt when, a few minutes before we played, he looked up what was then troubling the Petroff most, rolled it out and rolled me over. That was about fifteen years ago. I've not played it since.

But I would. If everybody played 3.Nxe5, which right now nearly everybody does, I'd play it now. But it's 3.d4, the move that Adam chose, that I've never been happy against. Only that. If not for that, I'd play it. And if other people don't like it? So much the better for that.


Martin S. said...

Didn't Frank Marshall play the Petroff as an aggressive defence, before it got its boring reputation? I think he went...Bg4....Ne4...f5...f4 & king's side attack.

Times have changed.

Alberto said...

Well, as a matter of fact I don't think things are so bad with the Petroff. I agree it tends to be somewhat more drawish due to the fact of the simple central pawn structure (like the Scandinavian, there is an early central pawn exchange) and also there are a few drawish lines White can play (1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d3 or 5.Qe2) but in the whole it's far from being drawish. I always play it, and though it's true a few weaker players who knew they were so tried to draw by early queen trade, it's true also that Black has a lot of potential attack in all the classical variations with takes takes and d4, or in that line where White tries to play with reversed castles. I've actually beaten some strong players in that variation (Nimzo attack, I believe). Players in the local circuit here were so very much fed up with my Petroff that they started trying 3.Nc3, but then I got the better of it with 3...Bb4 playing a reversed Ruy Lopez where black does fine. I think the Petroff can be played for players not looking for a draw and even quite aggressive players, like myself. The problem is that people use it for drawing - but the problem is with the people, not the opening