All,
After all this time I finally had to face the line, in fact virtually playing against myself. I did okay but had to face a form of the Catalan I wouldn't normally having played e6, d5, Nf6, Be7, 0-0, c5 and b6.
To rephrase, I play the Tartakower and would respond to the Catalan 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.g3 dc 5.Bg2 Bb4+. When facing the English with g3, as I play as White, 1.c4 e6 2.g3 d5 3.Bg2 dc is often recommended as a good line for Black. However, Marin makes a good case for 4.Na3 Ba3 5.ba, answering 5..., Ne7 with 6.Bb2 0-0 7.Qc2.
I'm happy with the line 1.Nf3 d5 2.c4 e6 3.b3 c5 that Davies recommends in the Dynamic Reti intending d4, Nh6, f6 and e5. But if white follows the Marin recommendation where should Black be going ? Does he do best to avoid Nf6 ? If not, why is 1.c4 e6 2.g3 d5 3.Bg2 dc a clearer path than 3..., Nf6 4.Nf3 dc ?

The great thing about chess is that most assessments are largely subjective, so it would be very strange for people not to disagree about the exact assessment of positions, GM's do it all the time. On this one though I think we're both quite close. What is more of a mystery to me is the goings on in the ...c6 line, which I intend to look at next!

I appreciate what you're saying and it is a fairly solid position for Black, to say its completely equal I think is a bit of a bold statement though, but I accept I could well be wrong, I just think Whites more flexible structure should count for something. Isn't the line with ...Ra7 the (very!) old Miles-Chernin game. Remember looking at that years ago I think theres a thread here somewhere, I thought White was slightly better there also but of course it ended in a draw!

In work now so not got a board in front of me but doesn't playing ...a5 for Black leave his Q-side a bit static, and eventually, after an exchange of light square bishops say, some nice juicy squares for White on b5 and c4. Not saying Black can't play it of course, but at least there is some potential there for the future from Whites point of view. For Black I think I'd prefer a plan involving ...Nd5. Like I said before I still dont think this line is much at all for White, just that in abstract terms this line gives him more chance to create something than the ...c6 line does. I'd say its a nibble for White, nothing more.  Mind you I just looked at the update on the main site and the pawn sac for White in the ...c6 line looks very interesting...at least in a practical game it seems nice to play at first glance, but a pawn is a pawn also.

On 7.d3 it surely must be better for White to have the Black rook on a7 than b8 I would have thought, so perhaps this is the more accurate move order.

BPaulsen wrote on 05/05/12 at 19:45:40:


Thats kind of a typical position you've quoted there that can come about through various move orders. I'm not claiming its much for White but he definitely is for preference after say 14.Rfc1 with ideas of Qa3 and playing for the b4 break at the appropriate moment. Thats kind of what I meant by saying White is the one trying to create. Its not that I'm saying Black should be desperate to avoid this mind, just that the other ...c6 line seems to be holding up better where White doesn't get this nagging pressure. Ultimately its personal preference and the whims of fashion I suppose, but its not an entirely innocuous line at all for White and deserves serious respect.

Regarding 7.d3!? - if this prevents Black establishing the ..a6, ..b6 structure on the Queenside by allowing White a4-a5 then I agree you are completely right, 7.d3 would be a more accurate move order. And it does make sense not to be dancing around with the Queen until entirely necessary. I've not checked it but very interesting if true. Something like 7...a6 8.Qb3 Bd6 9.a4 Rb8 10.a5 is kind of what I'd think White was ideally looking for. I suppose the reason 7.d3 is not more popular is that White players think Black can just reply 7...b6 but it seems like 8.Nd4 should be good for White, although Ponomariov played 7...b6 in a 2011 game so may be more here than meets the eye. Even more mystifying is that his strong GM opponent Eljanov didn't play 8.Nd4 so we dont know what Ponomariov intended...he can give an exchange with ...Ne5 or a pawn with ...Nd5. In both cases its not clear Black has sufficent compensation, but its also clear many White players may wish to avoid this kind of thing also.

John Bartholomew wrote on 04/04/12 at 06:17:21:


Very interesting discussion here. I think 5...c6 for Black  is actually the more solid continuation at the moment, although you could argue its personal preference. I don't see any clear way for Black to an easy life in the system 5...Nbd7 6.Qxc4 c5 7.Qb3!? followed by a4,d3,Na3 in all the games it seems to be White looking to create something. 5...c6 on the other hand is more clear-cut and looks like getting more and more worked out.

There is also this possibility, for which I have a question:

1 c4 e6 2 g3 d5 3 Bg2

and, like with the Catalan, having the bishop on g2 and not the knight on f3 gives white more pressure along with diagonal which may give him some extra resources if black takes on c4 immediately.

My first guess is that 3...dxc4 4 Qa4+ c6 5 Qxc4 b5 should transpose to the 3...Nf6 4 Nf3 dxc4 lines under discussion after white retreats, and looking at the games it seems that this is true.

3...dxc4 4 Qa4+ Nd7 5 Qxc4 is another case.  Here 5...a6 allows 6 d4 with a transposition to a slightly more favourable than normal version of the 5 Qa4+ Catalan proper (but still not terribly much).

but 3...dxc4 4 Qa4+ Nd7 5 Qxc4 c5! seems to force the transposition to the main "Neo-Catalan" lines after e.g. 6 Nf3 a6 7 Qc2 Nf6.

So the question is: am I missing anything here? Does white have any profitable ways to deviate?

The other question would be whether Marin's Na3 ideas have been completely defused or not...

BPaulsen wrote on 03/31/12 at 21:29:23:


Agreed, 5.Qa4+ Nbd7 is solid for Black. In the March update I examine the game Naiditsch-Papin (from the recently completed European Championship) which featured 6.Qxc4 a6 7.Qb3 c5 8.a4 Rb8 8.a5, discouraging the b-pawn advance. White has some nice positional ideas here, and I feel like his position is definitely the more pleasant to play. I'm probably biased though, cause I like White in a lot of these lines . Of course there's other lines - the immediate 6...c5 being a major alternative. The tremendous flexibility of both setups makes this a pretty tough one speak about in terms of absolute theoretical advantages, imo.

I assume in the line 1.Nf3 e6 2.c4 d5 3.g3 dxc4 4.Qa4+ the intended equalizer is 4...Nd7 5.Bg2 a6, using the tempo to achieve a quick ...b7-b5. This looks like a critical test, but I'm not sure it's so clear-cut. Delchev, for instance, claims a small advange for White in this line citing Svidler-Kramnik from last year. Definitely something to investigate.

I would argue that 5.Qa4+ c6 6.Qxa4 b5 is unusually popular at the moment. I was kinda shocked to see how many GM games have been played in this line already in 2012 and in the past year overall. Yes 5.Qa4+ is the main line of the Neo-Catalan, but you have to remember that historically Black has played 5...Nbd7 more than twice as often as 5...c6. Thus, to see so many sharp lines at GM level with 5...c6 6.Qxa4 b5 is noteworthy. Compared to 5...Nbd7 these lines are way more easily subjected to exhaustive computer analysis, so I suspect we'll have a more clear evaluation of 5...c6 by year's end.