Nernstian59 wrote on 06/24/24 at 22:30:04:

Thanks for the summation and link. I always thought of the Tarrasch as emanating from 1.d4 d5 2.c4 e6 3.Nc3 c5. However, I understand that committed Tarrasch players would want consider the Tarrasch as a universal reply and play it against both 3Nc3 and 3Nf3. As you explained, 3Nf3 appears to rule out the Dubov variation. Good to know!

After 1.d4 d5 2.c4 e6 3.Nf3 c5, Sam Collins makes some general comments in his Move by Move book that I would like to quote:

"As I hope has become clear by now, the Tarrasch is a principled opening and White should not be able to hope for too much by avoiding the main lines. That said, main lines change over time and this chapter contains one very serious attempt.

This is probably a good time to mention that, when confronted by a sideline (in the Tarrasch or elsewhere), please do NOT rush your response. Sidelines by their nature have slightly offbeat ideas which, if well neutralized, can leave Black with a pleasant game. However, often on seeing an early Bg5 or e2-e3 in the Tarrasch (I speak from bitter experience here, in particular from my game against Bischoff) the natural tendency is to relax and play a quick move. Black still needs to be precise in the sidelines as in every other Tarrasch variation - his structure is fundamentally more vulnerable than 
White's and we don't have the luxury of second-rate responses. Apologies if this seems blatantly obvious, but I have personally scored worse against sidelines than against main lines in the Tarrasch, and want to save you from a similar experience."

Tarrasch, or similar, lines can arise from the Reti, where White is already committed to Nf3. Here too, Black will have to choose his moves with care, but I suspect he is ok.

I'm rather tardy in pointing out John Watson's book review column in the March 2023 issue of Chess Life, but I feel it still may be of some relevance to those posting on this thread.

https://pubs.royle.com/publication/?m=40761&i=783906&view=articleBrowser&article...

Watson examines the works by Jorden van Foreest, Vassilios Kotronias, and Renato Quintillano discussed in this thread along with Cyrus Lakdawala's Opening Repertoire: Queen Gambit's Declined - Tarrasch, which was published after most of the posts in this thread were made. Along with perusing each work, Watson also provides some nuggets of Tarrasch theory. The one of most interest to me was how early development of the g1-knight and delayed deployment of the b1-knight allow White to avoid the Dubov Variation. For example, after 1.d4 d5 2.c4 e6 3.Nf3 c5 4.cxd5 exd5 5.g3 Nc6 6.Bg2 Nf6 7.0-0, the attempt to play Dubov-style with 7...cxd4 8.Nxd4 Bc5 is met by the strong 9.Nxc6 bxc6 10.Qc2!. A search of the Mega Database finds that White scores 80% here, and Stockfish evaluates the position as ±. Watson gives the possible continuation 10...Qb6 11.Nc3 Be7 12.e4! I wonder if this particular move order is the "problem with the Dubov" that RoleyPoley was asking about in Reply #51.

The delayed development of the b1-knight isn't a problem for the repertoires offered by Kotronias and van Foreest because they don't recommend the Dubov variation. Watson points out that Quintillano apparently "doesn't deal with early Ng1-f3 move orders". On the other hand, Lakdawala is aware of the issue and suggests deviating from the above sequence with 6...c4!? The main line then continues 7.0-0 Bb4! 8.Nc3 Nge7 9.e4! 0-0 10.exd5 Nxd5 11.Nxd5 Qxd5. Here Watson gives 12.Be3! as better than Lakdawala's 12.a3. Watson provides the following line: 12.Be3 Qh5 13.Ne5 Qxd1 14.Rfxd1 Nxe5 15.dxe5 c3 16.bxc3 Bxc3 17.Rab1 Bxe5 18.Bc5 Bf5 19.Rxb7 Rfd8 20.Rxd8+ Rxd8 21.Bxf3 a5 22.Ra7 Bc3 23.Bb6 winning Black's a-pawn. Watson's overall conclusion: "It's conceivable that Black can eke out a draw .... but the overall impression is not positive. For now, it seems to me that the Dubov is an excellent choice against 3.Nc3, but that you may want to play a normal Queen's Gambit or a traditional Tarrasch vs. 3.Nf3."

Using 3.Nf3 to avoid the Dubov variation involves trade-offs since White players choosing this move order will likely have to take on the Semi-Tarrasch, the Ragozin, and the Vienna (among others) while surrendering the possibility of playing Botvinnik's f3 +e4 plan in the QGD Exchange. This approach is used in the Grivas Opening Laboratory book series, where the author advocates 3.Nf3. Thus his repertoire covers the Semi-Tarrasch and the other lines mentioned above, but oddly he doesn't give the anti-Dubov variation pointed out by Watson. Perhaps Dubov's revival and popularization of the variation came while Grivas' books (which were published in 2019-2021) were in preparation.

For those who wish to stay with 3.Nc3, there could be a new resource. In a recent review of Sipke Ernst's new positional 1.d4 repertoire (one of ChessBase's FritzTrainer DVD series) , the reviewer, Christian Hoethe, notes that Ernst allows the Dubov and meets it with a pawn sacrifice, leading to "sharp compensation". Unfortunately no specifics are given for this line.

Another option against the Tarrasch is given in IM Kushager Krishnater's Modern Chess database Top-Level Repertoire Against the QGD Part 2. His preferred move order is 1d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 c5 5.cxd5 exd5 and now 6.Bg5 instead of 6.g3. (Obviously this move order allows Black to enter the Semi-Tarrasch with 5...Nxd5. This variation is also covered in the database). ChessBase denotes 6.Bg5 as a "hot" move with numerous recent games being played with the response 6...Be6. Krishnater's main line leads to a TN for White that apparently hasn't been played yet, resulting some practical attacking chances for White, though the author admits Black should be fine with accurate play. 

kylemeister wrote on 10/14/22 at 18:40:05:


That does sound interesting as 5...Nc6 6.dxc!? is a serious option. 5...Nf6 allows 6.Bg5, which is all I know. Thanks for letting us know.

FreeRepublic wrote on 10/08/22 at 18:05:50:


Kotronias appears to have put a lot of independent analytical effort into this book. I found one line where his work is complemented by prior work from Aagaard and Ntirlis.

After 1. d4 d5 2. c4 e6 3. Nc3 c5 4. cxd5 exd5 5. Nf3 Nc6 6. g3 Nf6 7. Bg2 Be7 8. O-O O-O 9. Bg5 c4 10. Ne5 Be6 11. Nc6 bc6 12. b3, Kotronias continues with 12...Qa5, whereas Aagaard and Ntirlis continue with 12...h6.

Part of the appeal of the QGD Tarrasch variation is the possibility that it can be played as a universal system against many lines and move orders. This has been claimed, but some are skeptical and I have not developed my own opinion.

The repertoire that I have been pursuing is:
QGD Tartakover, QGD Exchange variation, Catalan, Ragozin/Vienna.

I think it provides good chances for Black. However, I can see the appeal of just learning the Tarrasch QGD instead, even if Black may end up with an endgame that is objectively worse but adequate. Examples are a 4 vs 3 (pawns on kingside) rook endings or two bishops vs bishop and knight.

FreeRepublic wrote on 10/10/22 at 13:22:19:


In addition to 9.Ne5, White has 9.b3. It has been played a few times but as far as I know it goes unmentioned by theory. After 9. b3 cxb3 10. Qxb3 0-0, White has several reasonable moves:  11.Bg5, 11.Bf4, 11.Ba3 and 11.Rd1. Black will often play ...Na5 to drive the queen away and possibly go to c4.

I went ahead and bought this book. With the exception of the Dubov variation, this seems like the definitive work on the Tarrasch. Kotronias is exceptionally thorough. He was bothered by one variation and worked on it for ten days until he found a satisfactory answer. Sometimes he evaluates a position and chooses a logical move despite initial engine evaluations. The engine eventually comes around and validates his work.

Kotronias compares the Tarrasch to the King's Indian Defense, on which he is a well-known author. He thinks the KID requires more only-moves.

I find the comparison interesting because the KID is known as a hyper-modern opening. The Tarrasch, with it's strong emphasis on space and development (at the expense of pawn structure), is so classical that it could be called hyper-classical. Kasparov played both lines.

I think the unifying link is aggression. Initially the KID looks calm, but Black often aims to attack the White king. The Tarrasch  takes an aggressive stance in the opening. However, the character of the middle-game is then determined by both players.

FreeRepublic wrote on 10/07/22 at 15:45:35:


I thought there was a problem with the Dubov variation now?

Is he dealing with the 4.dxc5 Line?