Confused_by_Theory wrote on 08/30/22 at 22:29:32:

It requires quite some theory, but I'd still prefer Nf6 looking for favourable transpositions to the Modern Benoni. Recently a book by GM Ivanisevic was published about this (move order 1 d4 N6 2.c4 c5)
So fans of the Pirc/Modern do have yet another option and may even consider a repertoire consisting partly of the KID, partly of the Benkö and partly of the Modern Benoni.
The big problem, as already mentioned, is the optimal timing of ...c5.

Hi.

I was thinking yesterday about how 1...g6 could actually be quite a clever move order if you want to avoid the fianchetto KID. If you go with 1.d4 d6 and white goes for a fianchetto setup (2.Nf3 and 3.g3 or 2.Nf3, 3.c4 and 4.g3) I find there are two basic problems.
A) It is quite hard to get in e7-e5 and B) there seems to be no really attractive point where you can just go c7-c5 instead. At least if white goes for 
1.d4 d6 2.Nf3 g6 3.g3 Bg7 4.Bg2
and not
1.d4 d6 2.Nf3 g6 3.g3 Bg7 4.c4 
where you can argue...

4...c5 5.d5 b5!? Is not that bad of a Benko
4...c5 5.Nc3 cxd4 6.Nxd4 Nc6 Is not that bad of a Symmetrical English
4...c5 5.dxc5 Qa5+ 6.Nc3 Bxc3+!? 7.bxc3 Qxc5 8.Qd4 Nf6 9.Qxc5 dxc5 Is not that bad in general.

In contrast for example:
1.d4 d6 2.Nf3 g6 3.g3 Bg7 4.Bg2 c5 5.0-0 cxd4 6.Nxd4 Nf6 7.c4
Is some sort of symmetrical English line where black has played an early d6. It should be playable for black of course. For someone who doesn't play the symmetrical English normally though, it is hard to get enthusiastic about. 

This difficulty in getting effective e5 or c5 moves in means there is some underlying pressure for positions to move towards ones where black just does not care about going e5 or c5 and develops instead. This usually involves the move ...Nf6 because it's just a qualitatively good developing move that it is hard to do without. If it is played though the position basically just transposes to the King's Indian. This is not bad in itself. Both Vassilios Kotronias and Gawain Jones for example have presented excellent and wide-encompassing repertoires where they show different ways of treating the positions in a competent manner as black. Yet when playing 1...d6 or 1...g6 I've found there is big practical value in avoiding KID transpositions in general. I mean for one Gawain Jones covers the fianchetto KID in ca. 150 pages and Kotronias in a whole book of around 700 pages. Even 150 pages is quite a lot to understand.

So why would 1.d4 g6 be better?

Well... Apart from allowing a whole new set of theoretically well regarded lines where black faces the fianchetto with d5 instead of d6 e.g.
1.d4 g6 2.Nf3 Bg7 3.g3 Nf6 4.c4 c6 5.Bg2 d5
Which in some circumstances lowers the attractiveness of playing the fianchetto to begin with for white. 

It is also quite easy to end up in the position after:
1.d4 g6 2.Nf3 Bg7 3.g3 c5 4.d5
There are other playable moves like 4.c3 and 4.dxc5 but none look terribly threatening to me. Black can play...
4...Nf6
And irregardless of if white goes 5.Bg2 or 5.c4 black can go for 5...b5!?. If white instead goes 5.a4!? I guess black could try and steer towards a Modern Benoni where white has played a very early a4. This seems like it wouldn't necessarily be optimal for white.

White can try:
1.d4 g6 2.Nf3 Bg7 3.c4
Hoping to wait with committing to the g3 move until he knows if black is actually going to play c7-c5. It seems to me like black has a decent waiting move in:
3...d6
When...
4.g3 c5!?
Is possible and positions could easily become a bit irregular e.g. after
5.Bg2 Nc6!? 6.d5 Na5
5.d5 b5!?
5.Nc3 cxd4 6.Nxd4 Nc6 7.Be3 Nh6!?

Or white can try:
1.d4 g6 2.c4 Bg7 3.g3
but then again black can aim for b5 moves if he wants
3...c5 4.d5 b5
(4...d6!?, 4...Qa5+!?)

Or white can go:
1.d4 g6 2.c4 Bg7 3.Nc3
When black's options for original setups seems the most limited but still he has some tries e.g.
3...d6
3...c5 4.d5 Bxc3+ 5.bxc3 e5 or 5...d6 or 5...Nf6 before a 6...e5.
4.g3 e5!? 5.Nf3 exd4 6.Nxd4 Ne7!?

Edit:
With 1.d4 g6 there is actually also
1.d4 g6 2.g3
Which I forgot. After 1.d4 d6 2.g3 there is 2...e5!? so there I wouldn't try g3 this early. 
2...Bg7 3.Bg2!?
The idea. Now
3...Nf6 4.Nf3 c5 5.c4 is a mainline symmetrical English.
3...c5 4.d5 d6!?
4...b5 5.d6 seems annoying; which would be the other part of the justification for the early bishop development.
5.a4
And we get a position where white has a little bit of a bind and the computer likes white very slightly, yet black is quite flexible and will probably not have to large of difficulties with developing.

Have a nice day.

Hi.

I tend to play people who don't usually nurse any minimal advantage they may get in various queen exchange lines very well, so I'd also consider:
1.d4 g6 2.c4 Bg7 3.e4 d6 4.Nc3 c5 5.Nge2 Nf6!? 6.dxc5
6.d5 may become a Modern Benoni line, though both sides have a little bit of wiggle room.
6...dxc5 7.Qxd8+ Kxd8
But it's probably not one of the better versions practically.

Have a nice day.

Thanks for the welcome (and edit button tip)

Confused_by_Theory wrote on 08/30/22 at 22:29:32:


In this line here, what about 5.Nge2, where white is simply trying to play d5 without Bxc3 or Bg4? 
5...cxd4 seems to transpose to a Maroczy Bind again
5...Qa5 6.d5 b5 7.cxb5 a6 8.Ng3
   8...Bxc3
   8...h5 9.h4 9.Bxc3 
   Black fails to transpose to a Benko and it seems that Bxc3 only yields axb5, where the position is all weaknesses for black.
5...Nc6 6.d5 Na5 ...I have no idea, I guess black is trying to put together ideas like a6->Rb8->b5 or if white moves the knight Bxc3.

Just to put it out there 1.d4... g6 has a pretty cool line trying to transpose to a slightly improved Benko in case white plays Nf3.
1.d4 g6 2. c4 Bg7 3. Nf3 c5 4. d5 b5 
6.cxb5 a6 7.bxa6 Bxa6 7. Nc3 Nf6 8. e4
5. e4 Nf6
and due to having provoked Nf3 black avoids the annoying e4 lines (the King Walk as its known) and the a7 line, which is pretty common nowadays if you try to delay Bxa6 in the normal Benko.
There is still this line though...
7.g3 Nf6 8. Nc3 d6 9. Bg2 Nbd7 10. Rb1, which I think is supposed to be a good try for an advantage by white.

If white plays 4. e4 trying to transpose into a Maroczy Bind there is 4...Qa5 5.Nc3 d6, which is pretty good for black.
or 5.Bd2 Qb6 6.Bc3 which is quite dry, but IMO still preferable to the Bind... I'm sure people would disagree on that though.

Not that this line is all that worthwhile... I mean you can see it's quite a Frankenstein of of different types of positions.
Just an observation on the difference between d6 and g6.

Hi.

Somewhat interestingly, to me at least, Leela indicates 1.c4 d6 2.Nc3 e5 3.Nf3 c5!? as quite playable. If white goes 4.g3 black can play a Reversed Botvinnik with 4...g6. I would venture out on a limb and say having played d6 early here is far from catastrophic for black. That being said I kinda recall being told about the Wedberg system at some point (c5, e5, Nc6, Nge7, g6, Bg7) and how not doing d6 early was mildly clever for some reason.

If instead white tries 4.e3 black can apparently aim for a slightly offbeat 1.d4 d6 2.c4 e5 3.Nf3 position with 1.c4 d6 2.Nc3 e5 3.Nf3 c5 4.e3 f5!? 5.d4 (5.h4!? seems higly interesting as well) 5...e4 6.Ng1. White has played e3. This seems not necessarily like a move that would have been done as early in 1.d4 d6 2.c4 e5 3.Nf3 e4 4.Ng1 f5. At the same time black has played the mildly offbeat early c5. Usually black tries to go for c6 at some point to hint at d6-d5.

Have a nice day.

It seems to do quite well on my database of TWIC's - 4 Nf3 g6 has a slight positive score for black in over 1100 games (loads via the 3.. g6 4 Nf3 Bf5 move order) and 4 f3 e5 is about 50%.

That looks like mild over performance vs objective expectation of course, but you are dragging white into a mess.

4 g3 has scored quite a bit better for white in practice (68%). Leela and SF14 seem to think that black's best might be 4.. e6.

Then 5 d5 is some sort of weird benoni and 5 Nf3 Ne4 works reasonably now Qd3 isn't there. (or 5.. d5!?)

5 Bg2 critical, then their idea seems to be 5.. d5. A slightly absurd tempo waste, but if white plays slowly it must be OK. The direct tactics after (eg) 6 cd ed 7 Qb3 Nc6 8 Bg5 Na5 9 Qa4+ c6 10 e4!? are dangerous but don't seem to win either.

Strange game!

an ordinary chessplayer wrote on 02/20/22 at 02:59:39:

I actually tried the Janowski Indian briefly as a junior. I can't remember now where I picked it up; probably just browsing a database. Over a handful of games I scored OK with it (and a lot better than with my other defence to 1.d4 at the time, the Chigorin Queen's Gambit). But I couldn't quite believe it was sound, so I soon moved on.

I quoted your 4.f3 line just to mention it's the suggestion for White in FM Daniel Barrish' Chessable course The Principled Queen's Gambit - Part 2 (2020). He claims an edge but doesn't cover it in much depth.