Monday, 16 August 2010

The Reti Gambit vs the Caro-Kann

Over the years since I wrote The Fascinating Reti Gambit in 2006, people have occasionally asked me if the same set-up can be used against other openings than the French Defence. Especially the Scandinavian and Caro-Kann have been suggested. I don't know if this is because they've fallen in love with having a bishop on b2 or if they're not interested in booking up against openings they seldom encounter.

However, I have hardly taken this seriously, as it must be a completely different scenario when Black's Bc8 is allowed to come out without any loss of time - and in the Scandinavian Black also has c6 available for the knight, which should put a stop to this idea once and for all.

Though, in blitz I sometimes use a home brewed system after 2.exd5 Qxd5 3.Nc3 Qa5/Qd6/Qd8 etc and now just Nf3, Be2, and then either 0-0 and b3 or first b3, Bb2 and see what happens. Why not call it The Jokhansson-system, if it needs a name. ;)

It's not very ambitious of course, but in 3 minute blitz online, the Scandinavian with Qxd5 is both very popular and annoying as the Black player usually has an enormous experience of the mainline positions and responds instantly up to around move thirty, and by then you are already pressured on the clock...

And now, back to 2.b3 versus the Caro-Kann. It was only when I recently was looking at the Caro-Kann from Black's point of view that I noticed in two recent repertoire books that 2.b3 was actually dealt with - though with little enthusiasm from the authors! :)

Their lack of enthusiasm may have resulted in some slight wishful thinking though, as my impression when seeing their lines was that White was basically equal and not worse, but with a somewhat different position than CK-players are used to play. By the way, as White over the years I've mainly relied on the Fantasy variation (2.d4 d5 3.f3!?) and more recently on the Apocalypse variation (2.Nf3 d5 3.exd5 cxd5 4.Ne5!?), but after finally seeing the modern theory on the mainlines, I can see why they are usually preferred.

So, a reasonably ambitious White player should no doubt play 3.Nc3 or 3.e5 in my opinion - but there is something with the CK-solidity that often makes White want to play "something surprising" instead (probably because he never bothered learning the theory, as 1...c6 isn't all that popular, and weaker players punting it often crumble easily when surprised.) and there are many, many ideas to choose from and unlike in the French and Sicilian, it's hard for Black to really punish these off-beat ideas, as he has already committed to building a fort - so frequently black just builds the fort according to plan and waits for White to come to him.

And now finally a look at some concrete lines in the RetiCK - or whatever we should call 2.b3 here?! First according to Houska and then Schandorff.

1.e4 c6 2.b3 d5 3.Bb2 dxe4 4.Ne2!?

Houska mentions two ways to play for black.

A) 4...Nf6 5.Ng3 Nbd7 6.Nc3 Nc5 7.Qe2 (7.h3) 7...Bg4 Diagram

Now 8.f3 is forced according to Houska.

(But to me it looks like 8.Qe3! is a better move, e.g: 8...Qb6 (or 8...e5 9.Ngxe4 with a roughly equal position with fighting-potential, White can hardly expect more than this with 2.b3 vs the Caro-Kann.) 9.Bc4 (9.a4!?) 9...e6 10.h3 Bf5 11.Nxf5 (11.d4!?) 11...exf5 12.Qf4 (12.g4!?) 12...Be7 13.Qxf5 0-0 14.0-0 Rad8 15.d3! exd3 16.Rfe1 Ne6 (16...Bd6? 17.Na4!+-; 16...Rde8 17.cxd3 Qc7 18.d4) 17.Bxd3 +=)

8...exf3 9.gxf3 Bd7 10.0-0-0 (10.Nce4 Ne6 11.Bh3 Qc7) 10...Ne6 Bokuchava-Gurgenidze, USSR 1974. And now I think 11.Qe3!? is best again, covering most of Ne6's options, opening the way for Bf1 and perhaps supporting f4-f5 or d4-d5. Black has a pawn and no weaknesses, but his position is a little congested. In blunderchess white probably has decent compensation, but in corr perhaps not quite full comp..

B) 4...Bf5 5.Ng3 Diagram

The knight hits the bishop and pawn-e4, but also blocks the g-pawn - all necessary to recover the pawn, but also another thing that makes it less similar to the real Reti gambit. Throwing the g-pawn up the board, though weakening, has the usual effect of creating volatile positions with "cut and thrust" play. Now, with the pawn firmly on g2, Black has little to fear except the loss of his extra pawn.

5...Bg6 6.Nc3!? Diagram  Here I prefer to simply take back the pawn ASAP with the text move, without further ado.

Houska gives 6.h4 and even awards it an "!", h4-h5 may be a standard move in the mainline CKs, but here, to me at least, it looks more like a long term weakness. Worse still, is that it doesn't really lead to the kind of play you want as white - as can be seen from the humdrum examples below.

6...h6 (6...h5!? ) 7.Nc3 Nf6 8.h5 Bh7 9.Qe2 e6 10.0-0-0 Nbd7 11.Ngxe4
(11.Ncxe4 Qa5 12.Kb1 Nd5 13.Nc3 N7f6 14.Qf3 Nb4 15.d3 Nfd5 16.Nge2 Be7 17.a3 Nxc3+ 18.Nxc3 Nd5 19.Nxd5 cxd5 20.Qg4 Bxa3 21.Qxg7 Bxb2 22.Kxb2 Ke7 23.Rh4 Rac8 24.c4 b5 25.Rf4 Bf5 26.Qe5 f6 27.Qe3 dxc4 28.dxc4 Rhd8 29.Rfd4 bxc4 30.Qc3 Qxc3+ 31.Kxc3 cxb3+ 32.Bc4 Rxc4+ 0-1 Klee Helmut (GER) (2311)-Van Elsen Danny (BEL) (2248), corr.)
11...Qa5 12.Kb1 Be7 13.d4 (13.g3 0-0-0 14.Bg2 Nd5= Mazalon-Jaroch, Solec Kuj 2003.) 13...0-0-0 14.Qf3 Kb8 15.Bd3 Ka8 16.Nxf6 Nxf6 17.Bxh7 Rxh7 18.Ne4 Rhh8 19.Nxf6 Bxf6 20.d5 Bxb2 21.Kxb2 cxd5 22.Qxf7 Rhg8 23.Rh4 Rdf8 24.Qxe6 Qd8 25.Rg4 Rxf2 26.Re1 a6 27.Qe7 Qxe7 28.Rxe7 Rf5 29.Rb4 b5 30.Rg4 Rxh5 31.Rf4 Rg5 1/2-1/2 Cortifias,V-Zamora Aleaga,J, Cuba 1996, corr.

6...Nf6 7.Qe2 Nbd7 8.Ngxe4 Diagram

Why not unblock the g-pawn? The most natural way to develop now is g3, Bg2 and 0-0, (though 0-0-0 might be possible too). Playing for g2-g4 is hardly recommendable, as it weakens f4 and besides, since pawn e4 has already been recovered there's not much point in chasing Nf6 to an even more active post on d5 or encouraging a knight exchange on e4 for that matter.

By the way, 8.Ncxe4 Qc7 9.h4 h5 10.0-0-0 0-0-0 is preferred by my silicon friend, but it hurts my eyes to see Bf1 standing there boxed in, looking like a big pawn. What the engine likes is probably the chance to gain space with d4 and c4 - but it can hardly be wise to "storm the fort" with less than full force...

8...e6 Developing as usual - and incidentally stopping White's mate threat.  

9.g3 (9.g4? h5!? (9...Nd5!?; 9...Qa5!?) 10.g5 Nxe4 11.Nxe4 Bxe4 12.Qxe4 Qxg5) 9...Qa5 10.a3!? (10.a4?! Bb4 11.Bg2 looks more comfortable for black, as pawn d2 may be a tactical weakness.) 10...Be7 11.Bg2 0-0 (11...0-0-0 12.0-0 (12.b4!?) 12...h5 13.b4 would allow both players to "slug it out".)

12.0-0 Diagram

with a roughly level situation (though my computer prefers Black slightly), where the better player may still prevail. White's game plan will most likely involve advancing his q-side pawns, playing for an eventual b3-b4-b5 or perhaps d2-d4-d5. Black's play will probably involve piece manoeuvres and exchanges, saving c6-c5 or e6-e5 until white has declared his intentions.

Thus far we has some slight remedies against what Houska recommended in Play the Caro-Kann in 2007.

In my next post I'll take a look at what Schandorff recommends for Black in his very recently published repertoire book The Caro-Kann.

1 comment:

Anonymous said...

In line A) , I think your suggestion of 8.Qe3 is best met by the natural and plausible 8...e6. This seems harder to crack and is also the very CK move :-)