Saturday, September 16, 2017

Opening Preparation: The French Defense - Tarrasch Variation

The Tarrasch Varation of the French Defense is said to be the second most popular response by White, behind only 3.Nc3. Well, let me summarize what a specialist of the Advance French, Evgeny Sveshnikov, had to say about it. Sveshnikov pointed out that 3.Nd2 does not fight for the center as d4 has been weakened. He brings up the fact that it blocks the development of the other White Queenside pieces, particularly the Bishop on c1, which in turn hems in the Rook on a1. Lastly, he makes the point that the only true plus to this line is defensive in nature. With no pin on the Knight (Any ...Bb4 idea will be answered with c3), Black is unable to weaken White's pawn structure. That said, why is White defending? White goes first. White should be the one doing the attacking. For these reasons, Sveshnikov is of the belief that 3.Nd2 is an error from the perspective that Black can gain immediate equality with 3...c5!, and argues that the games of Bareev prove his point. Under normal circumstances, this early advancement of the c-pawn would be a mistake until White has locked the center. This is why a line like 1.e4 e6 2.d4 d5 3.Nc3 c5 is a mistake. White can now trade on d5, and with the Knight on c3 pressuring d5, the attack on what will shortly become an isolated pawn would be too much for Black. However, with the Knight on d2 instead of c3, there is no pressure on d5 at all. In fact, as we will see in the first part of the article, White's Knights typically end up on b3 and f3. From both of those squares, it would take four moves to get the Knight to d5. So in this case, with the pressure on d5 being minimal, Black can play like this and use his piece activity to offset the structural weakness. For this reason, I will be covering 3...c5 as the response to 3.Nd2.

A brief word on one of the other main lines against the Tarrasch, 3...Nf6. I used to play this back in the late 90s and early 2000s when I first played the French Defense because the moves and ideas appeared simple. That said, after 4.e5 Nfd7 5.Bd3 c5 6.c3 Nc6 7.Ne2 with the idea of putting the d2-Knight on f3, White's development is harmonious.



His pieces are all able to find good squares as his Knight will go to f3, opening up the Bishop on c1, and he can usually stir up an attack on the e6-pawn, after Black plays ...f6 to chip away at White's center, or go for a direct attack on the Black King. All of this while avoiding weaknesses in the position, unlike variations like the Winawer where in return for the attack, White is saddled with immobile doubled c-pawns So to me, 3...Nf6 is simply playing into White's plan, and is why I will cover 3...c5 instead.

This article will be split into two parts:
  • Part One will cover 4.exd5.
  • Part Two will cover 4.Ngf3 along with move order tricks and transpositions that Black must watch out for.

Part One: White plays 4.exd5

After 4.exd5, we see White trying to take advantage of the fact that Black did not wait for White to resolve the tension with his e-pawn prior to playing 3...c5, and force Black to either deal with an Isolated Queen Pawn, or else recapture on d5 with his Queen, which enables White to gain time via hitting the Black Queen while developing his pieces. I am going to suggest going with the Isolated Queen Pawn position via 4...exd5.

After the moves 1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Ngf3, Black has two ways to handle the position. Back in the days of Viktor Korchnoi and Wolfgang Uhlmann, the main line of the Tarrasch ran 5...Nc6 6.Bb5 Bd6 7.dxc5 Bxc5 8.O-O Nge7 9.Nb3 Bd6 was often played, leading to the position below.



And now let's look at the more modern approach. After the moves 1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Ngf3 Nf6 6.Bb5+ Bd7 7.Bxd7+ Nxd7 8.O-O Be7 9.dxc5 Nxc5 10.Nb3 Nce4, we get the following position.



So let's compare the two positions and assess what we have. In both cases, Black is looking at an Isolated Queeen Pawn (IQP), and in today's generation, this is enough to lower the popularity of this variation in favor of the Closed Tarrasch (3...Nf6) or the Queen recapture in the Open Tarrasch (4...Qxd5). That said, this IQP in some ways is a strength for Black. Neither White Knight is anywhere near attacking it. White can block it all he wants by moving one of the Knights to d4. However, this doesn't prevent Black from having outposts on e4 and c4. This is one case for the argument in favor of the more modern approach (which happens to be the line I play when White does play the main line). In the lines with 5...Nc6, both Knights are very far away from these outposts on c4 and e4. There is no quick way for the c6-Knight to get to c4 as their paths are guarded, and the Knight on e7 is busy covering d5 since the Bishop on d6, while active, is blocking the Queen from guarding the isolated pawn. So Black is succeeding in not trading off pieces as in the case of IQP positions, the last thing Black wants to do is trade down to an endgame unless something has changed in the position, like a trade occurs that makes the d-pawn no longer isolated, or Black wins material. In the lines with 5...Nf6, one of the Knights is already on the outpost on e4 and can easily be transferred to c4 via d6, and the other Knight is more active on f6 than in the old main line on e7. The Bishop is slightly more passive on e7 than d6 in the older line, but the Bishop is easier to make active. Just move the Bishop again. Lastly, notice how the only difference in White's position is in the former line, he has his Bishop on b5. In the latter line, this Bishop and the one on c8 are gone. This makes getting the Rook onto the open c-file easier for Black than in the old main line. Now there is one downside to the more modern approach compared to the old approach. With the Bishop on c8, which may get moved to a square like e6, along with the Knight on e7, the square f5 is well covered. This square can be another soft spot in Black's position to go along with the IQP and the weak d4 square in front of it. With the Knight on f6 instead of e7 and the Light-Squared Bishops traded off, f5 can be weak, and the Knight on d4 blocking the pawn can easily see itself on f5, a very annoying square with which to deal with a White Knight if you are Black. The automatic thought is that Black can easily play ...g6 and cover the weak f5-square. This may be possible in some cases, but Black must also be on the lookout for his dark squares around his King, especially if the Dark-Squared Bishops and Queens are still on the board.

Given that our goal is to come up with a Repertoire where Black has the best chance at achieving equality, but also have the ability to play for a win against anybody below the GM level, the line we will be following here is the 5...Nf6 variation rather than the older 5...Nc6 variation. For those of you that prefer the older 5...Nc6 variation, I would suggest going through the games of Viktor Korchnoi and Wolfgang Uhlmann, both of whom were major advocates of the 5...Nc6 variation. As for 5...Nf6, two major advocates are Mikhail Gurevich and Rafael Vaganian, and I would suggest studying their games heavily. The feature game here is one that was won by Gurevich, and features White's best line against this system by Black.


W: Pawel Blehm (2380)
B: Mikhail Gurevich (2635)
Cappelle la Grande open 1998

(A word of note about this game. The majority of the games I am covering in this 7-article series on the French are games from databases where all annotations are completely mine. However, this particular game was first covered in Lev Psakhis's book "French Defence 3 Nd2", and the annotations of this game are a combination of Lev's and mine. So why am I using this game? This was the first game I saw that exposed me to the 5...Nf6 line, and is still to this day the best example I have seen, and so I figured it was only right that this be the game that exposes the readers to the 5... Nf6 variation as well. This game is why I agree with Sveshnikov's assessment of the Tarrasch Variation. We will see in this game that White had a couple of chances to equalize, but there was never any opportunity for White to gain an advantage barring a serious error by Black.)

1.e4 e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5 5.Nfg3 Nf6 6.Bb5+ Bd7 7.Bxd7+ Nbxd7 8.O-O Be7 9.dxc5 Nxc5 10.Nb3 Nce4

We now have the position shown in the last diagram above.

11.Nfd4

The correct Knight to use to block the d-pawn, keeping the b3-Knight there to maintain the ability to play Nc5. Black has no problems after 11.Nbd4 Qc8 (Because 12.Ne5 is possible in this case, 11...Qd7 is inferior after 12.Ne5 Qc8 13.f3 Nc5 14.Re1 O-O 15.Ng6 hxg6 16.Rxe7 Qd8 17.Re2) 12.c3 O-O 13.Be3 Re8 14.Nc2 Bd6 15.Bd4 Nh5 16.Ng5 Nf4 17.Nxe4 dxe4 18.Bxg7 Re6 and it's Black with the attack!

11...Qd7

Now you might be wondering why Black should play 11...Qc8 against 11.Nbd4, but instead play 11...Qd7 against 11.Nfd4. The reason is fairly simple. The first thing that Black is trying to achieve is keeping an eye on f5, which both Queen moves do. What's different is the position of the White Knights. With a Knight on f3, he can move it to e5 and attack the Queen. Because the Queen needs to continue to cover f5, he would need to move the Queen to c8 and White will have gained a free tempo. Trying to attack the Knight by going to a square like c7 completely abandons the fight for f5. With a Knight on b3 instead, White might appear to have the potential threat of gaining a tempo with an eventual Nc5, but it turns out, as we will see, that White doesn't gain much as the Knight will be under attack when the Queen goes to c8. So it becomes very important to understand the slight difference in the position, and when a move that attacks your Queen is a complete gain of tempo versus when it is more a case of enticement, where you actually want to encourage an opponent's piece to enter a less desirable square.

12.Qf3

White continues his strategy of trying to get the Knight to f5. The only other move that makes sense is 12.f3, but after 12...Nd6 13.Nc5 Qc8 14.Nd3 O-O 15.b3 Nf5 16.Nxf5 Qxf5 17.Be3 Rfe8 18.Qd2 a6 19.Rae1 Rac8 20.Rf2 Bd6 21.Rfe2 Nd7, the position is equal.

12...O-O 13.Nf5 Bd8

Black cannot allow White to trade his Knight for the Bishop.

14.Be3 Rc8

The alternative for Black is 14...g6, White has two options that both lead to unclear play. He can give the check with 15.Nh6+ Kg7 16.Rad1 Rc8 17.Nd4 Re8 18.h3 b5 19.c3 a5 20.Ng4 h5 21.Nxf6 Bxf6 22.Ne2 b4 (Berelovich - Gurevich, Hoogeveen Open 1999) or he can retreat the Knight via 15.Ng3 Re8 16.Rfd1 Qc8 17.c3 a5 18.a4 Ra6 19.Nxe4 dxe4 20.Qe2 Bc7 21.h3 Bb8 22.Rd4 Qe6 (Godena - Bareev, Aosta Open 1989)

15.c3 g6 16.Ng3 Re8 17.Rfe1 a5 18.a4

It probably would have been a wiser move to play 18.Nd4 b5 and accept an equal position.

18...b5



This cannot be what White was hoping for. Black has the initiative, particularly on the Queenside. As we will see, White's position will continue to get worse.

19.Nxe4 dxe4

Black gets a slight advantage after 19...Nxe4 20.Nd2 bxa4 21.Nxe4 Rxe4.

20.Qd1 Qb7 21.Nc5 Qc6 22.Nb3

And here White can equalize with 22.b4! The game basically goes downhill from here for White.

22...bxa4 23.Nd4 Qb7 24.Qxa4 Bb6 25.Nb5 Bxe3 26.fxe3 Qb8 27.h3 Re5 28.Nd4 Qxb2 29.Re2 Qxc3 30.Rf1 Nh5 31.Qd7 Rf8 32.Qd6 Qc5

Black also wins after 32...Ng3 33.Qxe5 Nxe2+ 34.Kh2 Qxe3.

33.Qxc5 Rxc5 34.Ra1 Ra8 35.Nb3 Rb5 36.Nd4 Rd5 37.Ra4 Nf6 38.Rea2 Rc5 39.Rc2 Rac8 40.Re2 Nd5 0-1

This game made me into a believer of the 5...Nf6 variation. Since then, that is all I have played here, and no amateur player (Under 2200) that has played 4.exd5 has even survived against me, and I have yet to be beaten by anybody since taking up this line. This comes after many hardships with 3...Nf6 or 3...c5 4.exd5 exd5 5.Ngf3 Nc6. Through careful study of this game and many of Gurevich's other games with this 5...Nf6 line, the move 4.exd5 should cause you no problems either!


Part Two: White plays 4.Ngf3

This move is becoming increasingly popular. The main motivation behind it is that after 4.exd5, Black not only has the option to take with the e-pawn, but he can also take with the Queen via 4...Qxd5. This move is by no means better for Black, and many of the highly theoretical lines lead to a draw. However, given its high amount of theory, often times White is not prepared for such a theoretical battle, and plays this move to sidestep the problem. Now Black has a number of options that lead to independent lines, like 4...a6, 4...c4, or 4...cxd4. The two main responses are 4...Nc6 and 4...Nf6. The one thing that Black must watch out for is direct transpositions. Those of you that decided to take the 5...Nc6 approach in Part 1, studying the games of Korchnoi and Uhlmann, will also need to play 4...Nc6 or one of the independent lines. White has side lines that he can play, like 5.Bb5. That said, 5.exd5 exd5 transposes directly to the 5...Nc6 line. If you are one that plays 5...Nf6, as recommended above in section 1, then you would need to avoid 4...Nc6 and instead play 4...Nf6 so that any trade on d5 will lead to a direct transposition to the 4.exd5 variation with 5...Nf6. So what you play against 4.exd5 matters when deciding what to play against 4.Ngf3. Since we looked at 4.exd5 exd5 5.Ngf3 Nf6, we are going to look at 4...Nf6 here.

So as mentioned prior, trading pawns now on d5 leads to a direct transposition. The alternative for White is 5.e5, and after 5...Nfd7 6.c3 c5 7.Bd3, we have a direct transposition to a line called the Korchnoi Gambit.



This position looks similar to the first diagram that comes from the 3...Nf6 line, and actually, this position can arrive from that variation as well via 3...Nf6 4.e5 Nfd7 5.Bd3 c5 6.c3 Nc6 7.Ngf3. If White didn't have 7.Ne2 available, I would have suggested 3...Nf6. However, via the 3...c5 move order, the only way to get a Closed Tarrasch is via the Korchnoi Gambit, and so this will be the only line of the Closed Tarrasch that we have to worry about. The difference between the Knight being on f3 and e2 is significant. With the Knight on e2, the d2-Knight has the f3-square available and White's development is smooth. Here, the Knight on f3 is occupying the square that the Knight on d2 wants to go to. The d2-Knight can't go to b3 due to a pawn fork. And so therefore, after a move like 7...Qb6, White has no real way to hold on to the d-pawn, and after 8.O-O, Black can capture 3 times on d4, netting a pawn. That said, I am not going to suggest this line as it gives White all the play. He will be chasing the exposed Black Queen around the board and establishing a substantial lead in development. For those of you that enjoy defending such positions just for a pawn, I suggest you study the theory of that line with a fine tooth comb before even attempting to play the position over the board.

Instead, I am going to suggest that Black instead play 7...Be7, which leads to a very aggressive line based on the White Knights tripping over themselves. In addition to the 3...Nf6 or 3...c5 line, this position can also come from the 3...Be7 variation of the Tarrasch. We'll use a game from 2004 to look at Black's ideas in this line.


W: Pavel Kolar (2280)
B: Petr Boukal (2415)
Czechia, 2004

1.e4 e6 2.d4 d5 3.Nd2 c5 4.Ngf3 Nf6 5.e5

5.exd5 exd5 transposes back to Part One.

5...Nfd7 6.c3 Nc6 7.Bd3 Be7 8.O-O g5



What? This has got to be a typo, right? No, what you see in the diagram is correct. Black's idea with 8...g5 is to pose threats on the d4-pawn. Unlike the line of the Closed Tarrasch where the White Knight goes to e2 so that the other can go to f3, in this case, the Knights get in each other's way, and there is no good place for the d2-Knight to go. He can't go to b3 to guard d4 because of a pawn fork. So Black's idea is to follow up with 9...g4 and then capture the d-pawn once the f3-Knight moves away.

9.dxc5

This is White's best move, giving the Knight on f3 the d4-square when Black advances the g-pawn. The move 9.h3?! is ineffective, and in the game Tong - McCartney, NC Open 2017, the game followed 9...h5 10.Re1 g4 11.hxg3 hxg3 12.Nh2 g3 13.fxg3 cxd4 14.cxd4 and now 14...Nxd4 would have given Black a significant advantage.

9...g4 10.Nd4 Ndxe5 11.Bb5 Bd7 12.N2b3 h5 13.Re1 Nxd4 14.Nxd4 Bxb5 15.Rxe5



15...Qc7

In Emms-Lalic, Southend 2001, Black instead played 15...Bd7 16.b4 Bf6 17.Re1 Qc7 with an unclear position. With the move played in the game, Black will win a pawn.

16.Bf4 Qxc5 17.Be3 Qc4 18.b3 Qd3 19.Nxb5

The only move! Trading Queens is not a good idea being down a pawn, and 19.Qe1 just doesn't work. After 19...Bd7, White has a couple of tries, but with correct defense, they simply don't work:

  A) 20.Bg5 should be answered by 20...Bxg5 21.Rxg5 O-O-O and Black is simply a pawn up.

  B) 20.Nf5 is trickier because of discoveries on the e-file. After 20...f6 21.Rd1 Qc2 22.Rd2 Qxc3 23.Rexd5 exd5 24.Nxe7 Kf7! (24...Kxe7?? 25.Bd3+ wins for White) 25.Nxd5 Qe5 26.Nf4 Bc6 and the smoke has cleared. Black is an exchange up with a significant advantage. White has nothing better than 27.Qb1 Qe4 28.Rd7+ Ke8 29.Qxe4+ Bxe4 30.Rg7 b6 31.f3 (31.Ne6 Bf5 32.Nc7+ Kf8 and Black holds the material) 31...gxf3 32.gxf3 Rh7 33.Rxh7 (33.Rg8+ Kf7 -+) 33...Bxh7 34.Nxh5 Kf7 and Black has a winning endgame.

19...Qxb5 20.Bd4 Qd7 21.Qd3

Another option for White is to regain the pawn with 21.Rxe6 Qxe6! (21...fxe6?! 22.Bxh8 O-O-O 23.Bd4 gives White an advantage) 22.Bxh8 f6 23.Bg7 Kf7 24.Bh6 and just about any reasonable move by Black, such as 24...Rc8, 24...Qf5, or 24...h4, leads to an equal position with chances for both sides.

21...h4 22.f4

Now, because White has less control over e1, 22.Rxe6 isn't as good. After 22...Qxe6 23.Bxh8 f6, 24.Qd2 can be answered by 24...Qf5 while 24.Rd1 or 24.Bg7 can both be answered by 24...O-O-O, all of which lead to a slight edge for Black.

22...gxf3 23.Qxf3 f6 24.Rh5 O-O-O 25.Rxh8 Rxh8 26.Bxh7 h3 27.g3 Qc6 28.Bf2 Kb8



29.Re1

And after a long sequence of moves where the position hovered around equality, White makes an error. 29.g4 should be played here, opening up access to the diagonal of the Black King, and the Rook belongs on a different square so that White can position himself to cover all weaknesses. After 29...Rg8 30.Rd1 Bd6 31.Bg3 Ka8 32.Bxd6 Qxd6 33.Kh1 Qe5 34.Qxh3 f5 35.Qg3 Qe4+ 36.Qg2, any advantage that Black may have is extremely minimal. More than likely, either the sides will repeat with White continuing to offer the Queen trade, or else Black will trade Queens and the Rook ending should be drawn.

29...e5 30.Kf1

And once again, 30.g4 =/+ is relatively best.

30...Rc8 31.g4

Too Late!

31...Qxc3 32.Qxd5 Qc2 33.Qf3 Bb4 0-1

The resignation may appear to be pre-mature, but it turns out that Black does win a pawn, and White must have felt that was enough to throw in the towel. 34.Re2 Qd1+ and 34.Qd1 Qc6 both lose instantly, and so that leaves 34.Rd1. After 34...e4! 35.Bg3+ Ka8 36.Qe2 (All other legal moves lead to mate in 6 or quicker for Black) 36...Qc6 and now the only move that doesn't lose on the spot is 37.Qc4 when 37...Qxc4 38.bxc4 Rxc4 wins Black a pawn. The position is clearly better for Black, but it's hard to believe that Black is just completely winning despite the extra pawn as the Black pawns are scattered. That said, Black is the only one pushing, and White apparently decided that it wasn't worth the energy to defend this position.

This variation of declining the Korchnoi Gambit has only recently (within the last 25 or 30 years) become popular, and while the Black King might look a little airy, nobody has been able to bust it from the White side. Therefore, this aggressive system, questioning White's piece development and trying to prove that the White pieces are tripping over themselves, is alive and well, and especially at the amateur level, it might surprise some White players that were expecting a calm, positional game with no real weaknesses. This will make them think again!


Well, that concludes this article on the Tarrasch Variation. Whether you study the games of Bareev or those of Gurevich, along with this fairly new aggressive system against 4.Ngf3 with the 8...g5 idea, you'll eventually become a believer, just like Sveshnikov and myself, that the Tarrasch Variation gives Black equality just as easily as the Exchange Variation does.

The next article will cover what to do when White plays the tricky King's Indian Attack.


Links to the rest of the articles.
Introduction and facing the Advance Variation
Part One: The Exchange Variation
Part Three: The King's Indian Attack
Part Four: The MacCutcheon Variation
Part Five: The Steinitz Variation
Part Six: Beating the French with the Advance Variation

No comments:

Post a Comment