I noticed in another thread that Palliser recommends 4.e3 against the QID.  Having played the Black side of this a couple of times in the past (and currently playing against it in a correspondence game), I have found it rather interesting and possible to be creative sooner rather than later, which is good.  But from the Black side of the board, this opening has never really bothered me after 5...d5 and ...Bd6. 

1.d4 Nf6 2.c4 e6 3.Nf3 b6 4.e3 Bb7 5.Bd3 d5 6.0-0 Nbd7 7.b3 Bd6 8.Bb2 0-0 9.Nc3 a6 (there are a million roads to Rome, here, and it seems as though most of these moves can be played in endless orders)

If White resolves the tension in the center with cxd5 (anywhere between move 5 and 12), s/he avoids worrying about ...dxc4, but without the tension in the center, Black can  turn to the kingside.

Having said that, I am very interested in the QID from both sides of the board.  Are there approaches to 4.e3 that allow White to be aggressive, play for a win, and avoid the more drawinsh lines associated with the mainline QID?

4 e3 is certainly a hard move to categorise! From my own experience, White can often get to attack. Black often either plays ...c5 and an early ...g6 (allowing one of the d5 gambits) or plays a ...d5 and ...Be7 too passively, allowing Ne5 and some kingside pressure. Attacking for W against ...d5 and ...Bd6 is more tricky, at least early on, since Black's often the side there looking for a kingside strike. It shouldn't be forgotten, though, that quite often White can offer to play with an IQP which might not be theoretically dangerous, but could suit the right sort of player over the board. Indeed, quite sharp struggles can spring up as shown by some classic Keres games.

I was thinking maybe 8.b4 is a bit more active. And yes I dont like 11.Nh4 either.

I found an oldie for you:
[Event "Havana m"]
[Site "Havana"]
[Date "1890.??.??"]
[Round "18"]
[White "Gunsberg,Isidor"]
[Black "Chigorin,Mikhail"]
[Result "0-1"]
[Eco "E14"]
1.e3 d5 2.d4 Nf6 3.c4 e6 4.Nc3 b6 5.Nf3 Bb7 6.b3 Nbd7 7.Bb2 Bd6 8.Bd3 a6
9.0-0 0-0 10.cxd5 exd5 11.Nh4 g6 12.g3 Re8 13.Re1 Ne4 14.Nxe4 dxe4 15.Be2 Qg5 16.Ng2 Nf6 17.Bc1 Rad8 18.Bd2 Bc8 19.b4 h5 20.Rc1 Ng4 21.Rf1 h4 22.gxh4 Bxh2+ 23.Kh1 Qh5 24.f4 exf3 25.Bxf3 Bd6 26.Qe1 Qf5 27.e4 Qe6 28.Bf4 Bxf4 29.Nxf4 Qd6 30.Bxg4 Bxg4 31.Nd5 Bf5 32.Rxf5 gxf5 33.Qg1+ Kf8 34.Qg5 Rxe4 35.Nf6 Re6 36.Nh7+ Ke8 37.Qg8+ Kd7 38.Qxf7+ Qe7 39.Qxe7+ Rxe7 40.Ng5 Rh8 41.Nf3 Re2 42.Rg1 Ke6 43.Rg5 Rf2 44.Rg3 f4 45.Rh3 Kf5 46.Kg1 Rxa2 47.Ne5 Rg8+ 48.Kf1 Rgg2 49.Rd3 Rgb2 50.Rd1 f3 51.Nxf3 Rf2+ 52.Kg1 Rxf3 53.d5 Kg4 54.d6 cxd6 55.Rxd6 Kh3 0-1

HgMan,

FWIW here are the games. Attached as a pgn file.

castlerock wrote on 09/17/07 at 11:55:31:



One of the games that Castlerock posted was:

[Event "Regensburg op"]
[Site "Regensburg"]
[Date "1996.01.??"]
[Round "1"]
[White "Piskov,Yury"]
[Black "Hohner,Karl Heinz"]
[Result "1-0"]
[Eco "E14"]
1.d4 e6 2.Nf3 Nf6 3.c4 b6 4.e3 Bb7 5.Bd3 d5 6.cxd5 exd5 7.Nc3 a6 8.0-0 Bd6 9.b3 0-0 10.Bb2 Nbd7 11.Rc1 Qe7 12.Ne2 Ne4 13.Bb1 Rac8 14.Nf4 Rfd8 15.Nd3 c5 16.Nfe5 Nxe5 17.Nxe5 f6 18.Bxe4 dxe4 19.Nc4 b5 20.Nxd6 Rxd6 21.Qg4 Qd7 22.Qf4 c4 23.Ba3 Re6 24.bxc4 bxc4 25.Bc5 Bd5 26.h4 Ree8 27.Rb1 Qc7 28.Qxc7 Rxc7 29.Rfc1 Rec8 30.a3 Rc6 31.Kf1 Kf7 32.Ke2 Ke6 33.Kd2 h5 34.Kc3 g5 35.Rh1 g4 36.g3 Kd7 37.Rb4 Ra8 38.Rhb1 Rcc8 39.Rb6  1-0

A couple of notes:  Instead of 17...f6, Black might have tried 17...Bxe5, removing both of White's knights from central squares.  It seems a shame to give up the bishop to do so, but I would imagine that Black is at least equal here.  Secondly, in the game, 19...b5 looks like a mistake: White gets the bishop as in my previous suggestion, but without weakening his central pawns.  19...Bc7 looks very good for Black: White's knight is now on the wrong side of the board to attack or defend and White's pieces seem to lack harmony.  Black enjoys the bishop pair and both bishops are pointing happily at White's king position.  Surely ...f6-f5-f4 is in the offing with an attack.

I don't think this refutes White's plan, which does look very good, and I would need to compare the game above with the other efforts the Castlerock posted, but I think Black has resources here...

One thing I have noticed is that a couple of the most regular 4 e3 users over the past few years (Chatalbashev and Malaniuk) don't actually play it as a QID.  i.e. after:

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 e3 Bb7 they tend to play 5 Nc3 more or less forcing 5...Bb4 with a transposition to the 4 e3 b6 5 Nf3 Nimzo.  I assume the benefit is this hugely cuts down on what white needs to know,  and if Black doesn't play 4...b6 against the Rubinstein he can be rather on his own!

"Forces" was perhaps too strong a word!

My point was that against the normal 5 Bd3 the following are considered to be black's best defences:

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 e3 Bb7 5 Bd3 d5 followed by ...Bd6 e.g. 6 cxd5 exd5 7 O-O Bd6 as discussed by HgMan

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 e3 Bb7 5 Bd3 c5 6 O-O and now both 6...g6!? and the risky 6...Be7 7 Nc3 cxd4 (7...O-O?! 8 d5!) 8 exd4 d5 are theoretically doing well

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 e3 Bb7 5 Bd3 Bb4+ 6 Nbd2 (again 6 Nc3 transposes to the Rubinstein Nimzo).  This is still a viable defence for black against 5 Nc3 of course.

The point is now that:

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 e3 Bb7 5 Nc3 d5 6 cxd5 exd5 7 Bb5+! c6 8 Bd3 is an improved version since the blunting of the Bb7 makes ...Ne4 harder to handle.  Still not a complete disaster, but enough to favour white. 8...Bd6 9 O-O O-O 10 e4 or 9...Nbd7 10 e4 also score well.

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 e3 Bb7 5 Nc3 c5 is now just a good Benoni for white after 6 d5!, most likely transposing to the 7 O-O 8 d5! line in the first note above.

So white avoids black's best lines in the 4 e3 QID, although for me the 4 e3 b6 5 Nf3 Nimzo lines are not that great for white IMHO.

On a related note, in a NiC YB some years ago Igor Stohl sent a letter to the forum talking about how he used to favour the 5 Bd3 c5 6 0-0 g6 line but had been put off by 6 Nc3!? since 6...g6 7 d5! is good for white again.  Yrjöla + Tella (and I think Wells also) give 6...cxd4 7 exd4 Bb4 as equalising, but I don't think either mentioned that 8 O-O O-O 9 Na4!? is another direct transposition to the Rubinstein Nimzo. (1 d4 Nf6 2 c4 e6 3 Nc3 Bb4 4 e3 b6 5 Bd3 Bb7 6 Nf3 O-O 7 O-O c5 8 Na4 cxd4 9 exd4).

Hope this helps and is of interest  

Well, it is at least better/more interesting for White than the normal ...d5/...Bd6 line.  Whether it is really that good is another matter!  5 ...Be7 should transpose and is ok, but IMO it is not something white should be worried about facing.