Scott wrote on 05/24/09 at 02:44:01:


3. Nc3 allows the Nimzo, but preserves favorable versions of the Benoni, and QGD. If you can find a variation of the NID that you like as white (4. e3, 4. Qc2 are probably the best tries for advantage).

3. Nf3 means you'll need to prepare the Benoni/QID/QGD precisely to hope for an edge.

Kasparov in the early 80's only played 3.Nc3 when was sure not to meet ...Bb4.

Scott wrote on 05/24/09 at 14:44:33:


If you can stomach the NID, then that would be the way to go. It'll make your life easier preparation-wise.

When deciding between 3.Nf3 and 3.Nc3, it is worthwhile considering what you play against other openings. If you play an early Nf3 against the QGD and Benoni, then 3.Nf3 is the logical choice. But if you refrain from an early Nf3 against the QGD/Benoni, then 3.Nc3 may be more suitable.

In my view, 3.Nc3 certainly carries a smaller workload than 3.Nf3, since you only have to prepare for the Nimzo (compared to the QID, Benoni and Bogo). If you aren't comfortable facing the Nimzo, consider 3.Nc3 Bb4 4.Nf3 which at least retains some similarity to 3.Nf3 and may even transpose after some of Black's most popular moves, namely 4...d5 (Ragozin) and 4...b6 (which often arises via. a QID). It also provides you with some flexibility against the Bogo with 4.Nc3 as well as 4.Bd2/Nbd2.

If you decide to stick to Nf3, then I would be tempted play some of the fashionable gambit lines against the QID as they are not very difficult to learn (you have to know the theory fairly well, but that is part of chess), but present your opponents with a lot of problems over the board, even if they recall the theoretically strongest variations. This especially applies to the d5 pawn sacrifices (5.Qc2 and the Polugaevsky Gambit).

TN wrote on 05/24/09 at 23:15:44:

After 3.Nc3 White has to be prepared for the Benoni as well.

TN wrote on 05/24/09 at 23:15:44:

Perhaps the most important move after 3.Nc3 Bb4 4.Nf3 is c5. If Scott likes 5.g3 it is a good choice. Maybe Scott should just bite the bullit and chose 4.Qc2, which has been most popular among GM for the last 20 years or so.

I am also preparing my transition from Nf3 to Nc3. Just to have something new OTB.
My experience is that Nf3 must/should be combined with g3 to get an advantage out of the opening. That means that you must be ready to play the Catalan and have something for the QID and the Bogo. With Nf3/g3 you also have an "easier" setup against the Dutch and the fianchetto variation of the Benoni is quite good for white.
Nc3 is for in entire new world but I find it quite interesting and, after all, you have to try everything before you really can choose...
Against the Nimzo, I like Qc2 and against the QGD, I will go for the exchange variation with Bg5. The last seems to give white a long lasting small edge and so small errors are not dramatic for white.

I play both.  For me, its purely a matter of mood or opponent. 

If you're going to learn the Nimzo as White, I would recommend starting with 4.e3 or 4.a3.  These will more often lead you to the basic structures that you need to learn in order to understand the opening. 

Scott wrote on 05/25/09 at 02:02:51:



While I understand the sentiment, and I might have written those words once, I now realise that in order to gain an advantage in the opening I have to be willing to sacrifice something.  My gambits usually occur later in the game than the Romantic players, but I still offer material or space in return for an interesting attack or at least counterplay.

The modern lines are replete with gambits, and I don't think it's possible to play theoretically topical chess without playing some sort of gambits. 

I choose 3.Nc3 about as often as I choose lines with 3.Nf3 (usually by transposition).  I usually make my choice based on what I think my opponent is most prepared for and go the other way. The problems, as several have mentioned, go beyond merely whether to play the QID/Benoni/Bogo systems or the NID.  White is even stating what sorts of Queen's Gambit he is likely to play.  I wonder if 3.Nf3 is actually as popular as Scott suggests, or if 1.Nf3 is popular, transposing to the the QGD with Nf3.

Antillian wrote on 05/25/09 at 21:19:17:


I am proposing that Scott take up 4.e3 or 4.a3 instead of 4.Qc2.  If you want to learn the Nimzo from scratch, I opine that you will find 4.e3 and 4.a3 easier to start with and, further, that they will do more to teach you the fundamentals of the Nimzo than most of the other lines.  As for 4.Qc2, while it is very fashionable at the pro level, I don't think its all that easy for a beginning Nimzo player to handle.  Just my opinion, your mileage may vary   

I think 4.a3 vs the QID and 4.a3 vs the NID are completely different species. The Sämisch is very principled for both sides exactly because it rises the question: pair of bishops or weak pawns?

Scott wrote on 05/25/09 at 02:02:51:


A "gambiteer" to me is someone who plays 19th-Century-style gambits.  I don't see how someone can play chess and not be willing to sac any given amount of material for corresponding play.  Pawn-sacs and exchange-sacs are a dime a dozen in modern chess theory, let alone practice.  The ability to spot them and the willingness to play them are just plain essential for good players -- but that doesn't make anybody a gambiteer, necessarily. 

I certainly don't think of people who play the 4.a3 versus the Nimzo as gambiteers, notwithstanding that their c4-pawn very frequently dies an early death.

@ everone who reads this: Is the theory of the Petrosian QID such that White can claim +=, or at least reasonable play for a win?

Markovich wrote on 05/27/09 at 14:30:44:


Last I checked the ...Bb7 variation left white with very good chances of playing for a win (tends to be +=), whereas the ...Ba6 variation leads to unclear positions where black isn't worse. As far as I know it's 4...Ba6 that has really put 4. a3 on the backburner theoretically.

Much like the French the QID has been in my repetoire since I started, so I've usually kept tabs on it theoretically (both in Bb7 and Ba6 variations since I've played both at varying points).

It might be mentioned that there is considerable material pertaining to 3. Nc3 and 3. Nf3, including the Petrosian QID and several NID systems, in Ivan Sokolov's recent pawn-structure book.

Aren't all QID lines either = or += leaning towards = ?  I believe the Petrosian QID is a good way to play for a win, and until the recent popularity of the Qc2/d5 gambit in the 4.g3 system, I kept saying I thought 4.a3 was a better way to fight for an advantage than 4.g3. Now my opinion is that both are good ways.

After 4.a3, I see 4...Ba6 about as often as 4...Bb7, and I'm comfortable facing them both. I can agree about the += leaning towards = evaluation, but I'm satisfied with that against such a solid opening as the QID. I wouldn't say the evaluation is different in the Nimzo.

As for the Nimzo, I disagree that 4.Qc2 is a harder start than 4.e3 or 4.a3. True, I can't really say I have much experience from the White side of the Nimzo, but I find the positions arising after 4.Qc2 easier to understand than the others. It's hard to say what will suit you, though, as the different 4th moves lead to quite different kinds of positions, and none of them really resemble the 4.a3 QID.

"Thus 4. Qc2 is an ideal piratical weapon at the non master level"

Does "piratical" mean attacking with cutlass, eye-patch and a parrot in tow?  

I agree that the 4.Qc2 line is a more natural choice for starting in the NID as it is more akin to QGD that 1.d4 players will be used to. Not the same of course, but a few similarities - and a few pieces ending up on similar squares. This should help the non-master player as they'll have some prior idea of what to do.

He has explained that: the pieces end up on similar squares. He was obviously thinking of the Queen's Bishop going to f4 or g5, which of course is not possible after 4.e3. Another question is if I agree with this argument.

I would say if nothing else, that  4. e3  is certainly the most instructive to learn.

Antillian wrote on 06/08/09 at 13:03:15:

Really? After 3 em. games with 4.e3 I only saw confirmed what I already suspected: Black better postpones ...d5 a bit and plays ...0-0 and ...c5 first; 4.e3 0-0 5.Bd3 c5 6.Nge2 is completely innocent; if 6.Nf3 is best White cannot avoid the Normal System; so now I try 6.a3 Bxc3+ 7.bxc3 transposing to the Sämisch.
Can't say this shattered my views on the question 3.Nc3 or 3.Nf3.

Smyslov_Fan wrote on 06/10/09 at 08:26:20:


I can see the logic of this approach. Teaching 4. a3 is the most effective way of illustrating the core  bishops pair versus pawn structure imbalance. But overall 4. e3 is the most instructive IMHPO because of the richness and diversity of ideas. You learn about every possible type of QP pawn structure, pawn transformations and the associated plans.

Antillian wrote on 06/10/09 at 13:31:46:


If I had to give only one recommendation, it would be for 4.e3, mainly because of its flexibility, as I mentioned in my previous post in this thread.

MNb has stated before that he likes the Hubner Variation as Black against the Rubinstein, but White can avoid this with 5.Nge2 or 5.Bd3 Nc6 6.Nge2. And against the move-order 4...0-0 5.Bd3 c5, White should consider 6.d5!?, as played by Korchnoi against Karpov in Baguio City. However, there is nothing wrong with 6.Nge2, which is certainly not 'completely innocent', as the 6...d5 7.cd5 cd4 8.ed4 Nd5 9.Bc2!? system (delaying castling in favour of Qd3, h4 and a kingside attack) shows.

Generally speaking, the Bd3/Nge2 system is the most tactically aggressive, the Nge2 system is the most positional, and the Bd3/Nf3 is both tactical and positional depending on how both sides play.

Finally, if White wishes to play the Samisch against the Nimzo, I would recommend that he does so via. an 4.e3 move order, e.g. only playing the Samisch against 4...0-0 and meeting 4...c5 and 4...b6 with something different.

You have not missed anything. But Carsten Hansen's book on the Rubinstein Variation from 2002 is a big help if you want to combat the NID with 4.e3. Of course there still be a lot of work to be done by yourself, but since when is that a bad thing? Especially if the prospect is finding ideas for your own?