Yeah I admit the Scandi is a little dodgy. I play the Qd6 lines and tend to prefer a6 with Nc6 rather than c6. For a long while I looked very closely at Kramnik's g6 variation of the Qd6 Scandi, but ultimately I felt white just gets a dominant center for black's rather passive developement.

Really I'm considering going back to the french full-time. I remember abandoning it after feeling frustrated with Tarrasch and exchange line players. The Tarrasch always seemed to give white all the fun with black having to walk a tight-rope, while the exchange line seemed to just take black out of the whole point of playing the french in the first place.

Also I disagree with what was stated earlier about the exchange line being better for black. It isn't, and there are specialists in the exchange with white that can blow you off the board quite quickly if you aren't fully booked out on it. My initial attraction to the french was to avoid those quick-blast-super-theory lines in other openings like the Sicilian, and my results against the classical Nc3 players was quite positive. So when i faced the exchange lines, I naturally felt totally out of my element and never could fully feel comfortable in that opening line. Now when I play online and I face the exchange, I'll just play Qxd5 and deal with the somewhat worse position.

I am in the "Black is better in the Exchange variation" camp. The thing is - the position after 3. exd5 exd5 is a zugzwang Or at least the side to move has to make an immediate concession on move 4 (as if defining the pawn structure on move 3 wasn't enough of a concession; IMO, ability to keep things ambiguous in the center is probably an important factor in many lines, including on move 3 in the French). In the Exchange variation, both sides would -ideally- rather place their knights on c3 and f6 (c6 and f6). But the problem is that if white plays 4. Nc3 or 4. Nf3, black can get his bishops to their own ideal squares - by pinning those knights with Bg4/b4. But white can't do the same with his own bishops: obviously Bg5 on move 4 (or later - after Nf3-...Bg4 and/or Nc3-...Bb4) is impossible, and Bb5 just loses a tempo to c7-c6. White can try 4. Bd3, as that is a rather active square for a bishop, too, but black can play 4... Nc6 - and white will have to play c2-c3, taking away a square from his own knight- which again gives black a better knight.

All of that probably means that white's best try in the exchange variation is 4. c4 - which is a line that feels as different from the other lines in the Exchange Variation as the Exchange variation is from the rest of the French. Even that line gives black a concession - as black can play ...Bb4 which is rather uncomfortable, unlike 4. Bb5+ c7-c6 that I mentioned earlier. I would say black is at least even in the 4.c4 line, too - in fact, one of my last game in the 15 game pool on ICC was in this variation; here it is it is not atypical game for this line, IMO:

[Event "ICC 15 0"]
[Site "Internet Chess Club"]
[Date "2010.03.04"]
[Round "-"]
[White "fearlessone"]
[Black "Russianbear"]
[Result "0-1"]
[ICCResult "White resigns"]
[WhiteElo "1904"]
[BlackElo "2117"]
[Opening "French: exchange variation"]
[ECO "C01"]
[NIC "FR.01"]
[Time "22:21:42"]
[TimeControl "900+0"]

1. e4 e6 2. d4 d5 3. exd5 exd5 4. c4 Bb4+ 5. Nc3 Ne7 6. Nf3 Bg4 7. Be2 dxc4
8. Bxc4 O-O 9. Be3 Nd5 10. Bxd5 Qxd5 11. O-O Qa5 12. Qb3 Bxf3 13. gxf3 b6
14. Kh1 Bd6 15. Ne4 Qh5 16. Nxd6 Qxf3+ 17. Kg1 cxd6 18. Rac1 Nd7 19. Rc3
Rfc8 20. Rfc1 Qg4+ 21. Kh1 Nf6 22. Rc7 Rxc7 23. Rxc7 Qf3+ 24. Kg1 Nd5 25.
Rc1 Re8 26. Re1 Re6 27. h4 Nf4 {White resigns} 0-1

Note how the pressure of the black bishops on g4 and b4 was so big that white first gave up the bishop pair - and then lost a pawn.

As for the 3. Nc3 versus 3.Nd2 - it is a matter of taste - for both white and black. I think the moves are roughly equally strong, but while 3.Nc3 leads to more direct play, the Tarrasch variation leads to somewhat more positional play.

The above thoughts on the Exchange kinda make me think white's best bet is to make some sort of inconsequential non-move on move 4. For example, 4.a3 is kinda sneaky, because it passes the move to black, and prevents a pin on a5-e1 diag in some lines, but of course, black wouldn't have to develop a peace first and could just play 4.... a6! So maybe white could somehow lose a move by triangulation Say, 4. Bd2, and, for example, 4... Bf5 5. Bf4 and white is again a little better

I've lost games to the Naidni Sgnik French, but not because of the opening. I've lost because I am a patzer. I can say this: I don't like Uhlmann's method of meeting it, namely castling kingside and making a huge pawn storm up the queenside. I like to castle long and pawn storm the kingside.

In the Center Counter/ Scandinavian, I wouuld say the ...Q_QR4 lines are best for Black. As Bent Larsen wrote after beating Karpov, "It is a good idea to avoid weak pawns against Karpov." The main virtue of the Center Counter is just that: no weak pawns. Putting the queen on ...Q3 seems wrong to me.

FBX, if you can get your hands on [i]Chess Praxis[/i[ by Nimzowitsch, there is a great chapter in there called "The Asymmetrical treatment of Symmetrical Positions." All you need to know about facing the French Exchange is in there. If you can find it, PM me and I'll try to get you a photocopy of the relevant section.

Yeah, it does seem wrong. Why block the KB, unless you're planning to fianchetto it? (On the other hand, in the Q-QR4 line, you often end up playing P-QB3 and Q-B2, and you can do that just as easily in this line, and you get the QN1-KR7 diagonal earlier).

Just had a quick look at several Tiviakov games. He seems to play the 2...QxP line exclusively. That's good, cause I never liked the Marshall line, for some reason. It feels like you're not really playing the Center Counter Defense unless you're playing 2...QxP. It's hard to explain. Like 2...QxP is the point of the whole opening, and 2...N-KB3 is just a transposition to something else.

But in the 51 Tiviakov games I just saw, he scored +13-7=31 with the Center Counter, so his results are pretty darn good. And he's scored draws with it against the likes of Judit, Chucky, and Nakamura, so it doesn't look like he saves it only for weaker opposition. (He also plays the 3...Q-Q3 line exclusively, BTW.)

The CC Defense as a whole still "feels" wrong, though. Even if you can get away with bringing your Queen out that early, what's the point of doing it? Or is the point simply in proving that you can do it (sort of like Nakamaura and 1. P-K4 P-K4 2. Q-R5 a couple of years ago)? If the point is to exchange QP for KP, can't you do that equally well or better in the French or Caro-Kann? The only advantage of doing it this way seems to be to avoid the P-K5 lines, but who's afraid of those?

To try to answer this, I thought I'd have a look at what the World Champions played. This is a rough guesstimate, but should provide a nudge. Basically, I went to chessgames, and looked at the lists of most commonly played openings for all world champions. They generally register any opening played more than a dozen or so times. The numbers are thrown off a bit by the inclusion of blitz, simul, and offhand games sometimes. They're also thrown off by the exclusion of infrequently played openings (I know Petrosian played the Petroff a half dozen times, but it wasn't listed as one of his commonly played openings; ditto for Fischer's occasional 1...P-K4's), but this should at least provide a rough feel for what moves have been most popular.

It looks like the general rule is that before the war, 1...P-K4 was tops, after the war it's been 1...P-QB4. Of all the World Champions, the ONLY losing score that I saw notched by a World Champion with one of his most frequently played KP defenses was.... Kramnik and his Petroff!

Interestingly, of the Caro-Kann-using World Champions, Botvinnik and Petrosian both played the French too. But Karpov and Anand didn't. That's interesting, because I think of the French and Caro-Kann as being almost "Sister Openings". Both are geared towards getting an immediate, pawn-supported P-Q4 in. But the French secures the KN1-QR7 diagonal at the cost of locking in the QB, while the Caro-Kann keeps the Bishop free.





These break down thusly:




Of course, we're considering only world champions here. What's most popular at the top level is not necessarily what should be most popular at the Club Level, or even the GM level in general. (How many beginners came to grief trying to play the Najdorf just because Fischer did?)

But if the question is what is the "best" move, then we have a fairly clear answer. Over time, the best players have considered the best moves to be 1...P-K4 and 1...P-QB4. Also competing are P-K3 and P-QB3. Nothing else is in the running. (Even Alekhine made no appreciable use of the Alekhine's Defense).

It's interesting that even when the King's Indian was fairly popular, that the Pirc/Robatsch was never anywhere as near popular. They're almost sister openings also. Black relies on similar setups in both, with the main difference being that in the Pirc, White usually hasn't played P-QB4. Little things like that can make a difference. Fischer lost a very interesting game to Benko at Curacao against 1. P-KN3, because he tried to play it like the Long Varation of the King's Indian, but with White's QBP uncomitted, he just didn't have the same kind of play you get in the Long Variation.

I wonder how many games Tiviakov won with theoretical novelties. I think it is worth knowing; is he outplaying people who surprise him, or is he winning by surprising them?

I'm always leery of statistics, as they neglect the tournament or match standing at the time the game was played. If it is the last round and a draw is as good as a loss, you may as well play the Latvian Gambit as the Petroff.

I think the Pirc never caught on because White can launch an attack on the kingside that is a tempo ahead of a King's Indian. In the KID White spends a tempo playing P-QB4. In the Pirc White gets to play P-K4, P-Q4, and P-KB4, but doesn't need to play P-QB4. That extra tempo adds to Black's dangers.

I don't know. Theoretical novelties might explain why he's won almost twice as many games as he's lost. But all I'm interested in now is whether the opening is legitimately playable. If he's played it 51 times, how much surprise value can it have?




Well, in the end, all we're gauging is opinion (albeit the opinion of the game's greatest players). Statistics can be off, but not that much off. We may not be able to judge the Ruy Lopez against the Sicilian this way, but it seems absolutely clear that of the 20 possible moves, there are four of them that the World Champions have favored, and that is all. So collectively, it seems that the World Champions are of the opinion that the best reply is one of those four.

Of course in this case "best" might just mean "most to their personal taste". It doesn't necessarily mean that other moves are inferior or unplayable.

That could be, and the times I've tried it, I've really preferred the Robatsch Defense, with 1...P-KN3, to avoid comitting the Knight, keep open the option of a Suttles-like Hedgehog, and so forth. I do think it's a good idea to play as many lines as possible in casual play to get experience with as many types of positions as possible. I'm amazed at people who don't dare vary from their tournament repertoire even in blitz.