Indian Game: main lines, transitions and statistical evaluations

Sort:
Yigor

The Indian Game 1. d4 Nf6 with its 416000+ master games in the Explorer, is the most popular opening in 2 plies blitz.png:

https://www.chess.com/forum/view/general/list-of-all-400-legal-openings-in-2-plies-master-games

In this thread I'll make subdivisions with statistical results. Engine evaluations are taken from the competitor's database.

 


2 plies:

  • 2D 1. d4 Nf6 Indian, 416K+, ev=+0.0, sev=+0.39

 3 plies, 10K+ (K=1000) games:

  • 2CD 1. d4 Nf6 2. c4 Indian: Main, 296K+, ev=+0.1, sev=+0.46
  • 2D 1. d4 Nf6 2. Nf3 Indian: KK, 103K+, ev=+0.1, sev=+0.27
  • 2D 1. d4 Nf6 2. Bg5 Trompowsky, 15K+, ev=-0.1, sev=+0.23, sh1=1.01 W-edged

 4 plies, 10K+ games:

  1. 2CD1e 1. d4 Nf6 2. c4 e6 Indian: Main, 144K+, ev=+0.1, sev=+0.40
  2. 2CD1g 1. d4 Nf6 2. c4 g6 KID, 107K+, ev=+0.1, sev=+0.54
  3. 2D1g 1. d4 Nf6 2. Nf3 g6 East Indian, 43K+, ev=+0.2, sev=+0.32
  4. 2Dd 1. d4 Nf6 2. Nf3 d5 QPG: Symmetrical, 42K+, ev=+0.2, sev=+0.35
  5. 2D1e 1. d4 Nf6 2. Nf3 e6 Horwitz: Symmetrical KK, 38K+, ev=+0.1, sev=+0.22
  6. 2CDc 1. d4 Nf6 2. c4 c5 Benoni, 33K+, ev=+0.4, sev=+0.48, sh1=1.24 W-edged

5 plies, 10K+ games:

  1. 2CD1g 1. d4 Nf6 2. c4 g6 3. Nc3 KID: Main , 89K+, sev=+0.52, sh1=1.01 W-edged
  2. 2CD1e 1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 Anti-Nimzo-Indian, 83K+, sev=+0.46, sh2=0.48
  3. 2CD1e 1. d4 Nf6 2. c4 e6 3. Nc3 Nimzo-Indian Invitation, 68K+, sev=+0.35
  4. 2CD1g 1. d4 Nf6 2. c4 g6 3. Nf3 KID: Symmetrical KK , 28K+, sev=+0.57
  5. 2CD1Ge 1. d4 Nf6 2. c4 e6 3. g3 Catalan, 14K+, sev=+0.37

6 plies , 10K+ games:

  1. 2CD1g 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 KID: Normal, 61K+, sev=+0.57, sh1=1.16 W-edged
  2. 2CD1e 1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 Nimzo-Indian, 56K+, sev=+0.25

  3. 2CDd1e 1. d4 Nf6 2. c4 e6 3. Nf3 d5 QGD: Symmetrical KK, 46K+, sev=+0.53, sh2=0.42

  4. 2CD1be 1. d4 Nf6 2. c4 e6 3. Nf3 b6 QID, 36K+, sev=+0.42, sh2=0.48
  5. 2CDd1g 1. d4 Nf6 2. c4 g6 3. Nc3 d5 Grünfeld, 28K+, sev=+0.39
  6. 2CD1g 1. d4 Nf6 2. c4 g6 3. Nf3 Bg7 KID: Normal 2, 61K+, sev=+0.55
  7. 2CDd1e 1. d4 Nf6 2. c4 e6 3. Nc3 d5 QGD: QK vs KK, 22K+, sev=+1.10, sh2=0.41
  8. 2CD1e 1. d4 Nf6 2. c4 e6 3. Nf3 Bb4+ Bogo-Indian, 16K+, sev=+0.72, sh2=0.45

(to be continued)

 

P.S. My special thanks to Bernard (Differentiation2) who contributes a lot to these statistical calculations. blitz.pnghappy.png

Yigor
  1. East Indian and QPG: Symmetrical have low black's sharpness sh2<0.5. It means that the number of draws is 2+ times higher than the number of black's wins.
  2. KID and Benoni have the statistical evaluation sev>0.50. It means that white has, at least, 50% more wins than black.

P.S. Sorry, I've corrected these results cuz, mistakingly, I considered the database (2300+ vs 2300+) before. tongue.png

Yigor
Differentiation2 wrote:

While Yigor does just four plies, I will do deeper variations like the lines of Nimzo-Indian, Queen's Indian, King's Indian, etc.

 

That's nice! happy.png I'll go deeper too but more slowly and systematically. tongue.png

Yigor

Differentiation2: Hey, Bernard, could U also indicate when the sharpness sh1 (or sh2) is >1 or <0.5 ?

Yigor
Differentiation2 wrote:
Yigor wrote:

Differentiation2: Hey, Bernard, could U also indicate when the sharpness sh1 (or sh2) is >1 or <0.5 ?

I will get to it, but I am in a rush now as I have to leave in a few minutes. Expect results by 7:00 PM EST because today I am very busy.

 

All right, take your time and see ya later. wink.png

Yigor

We see that statistically Nimzo-Indian is better for black than KID and QID.

Yigor
Differentiation2 wrote:

Nothing as far as I can see with sh(x)<0.5 for Benko!

 

And with sh1>1 ? I thought it should be sharp in white's favour.

Yigor

cjxchess16: Sure, Gruenfeld is a legal successor of the Indian game as well as of QPG and QGD ( Marshall defense) too. wink.png

Yigor

Gruenfeld is even a legal successor of English: Great Snake grin.png:

 

 

Yigor

I've added the Bogo-Indian. Curiously, contrary to the Nimzo-Indian with the same pawn structure 2CD1e, it's statistically bad for Black.

Yigor
cjxchess16 wrote:

There is another defense called East Indian, and that is 1 d4 Nf6 2 Nf3 g6.

But this is King's Indian Defense = KID (4 moves, #2). wink.png

tongue.png

P.S. Oops, sorry, I've confounded 1. d4 Nf6 2. c4 g6 with 1. d4 Nf6 2. Nf3 g6. I called the later Knights Variation I in OP (4 moves, #3).

Yigor
cjxchess16 wrote:
Yigor wrote:
cjxchess16 wrote:

There is another defense called East Indian, and that is 1 d4 Nf6 2 Nf3 g6.

But this is King's Indian Defense = KID (4 moves, #2).

But 1 d4 Nf6 2 Nf3 g6 is A48, not E60!?!!?!??!?!

So, we have discovered that A48 is also an ECO code for KID blitz.pnggrin.png:

https://www.365chess.com/eco/A48_King's_Indian_East_Indian_defence

Yigor

 Well, there is definitely a lot of ambiguities in ECO codes. The use of PSCC codes is better imho.

Yigor

cjxchess16: Finally, according to the Wikipedia, U was right about the name East Indian:

https://en.wikipedia.org/wiki/East_Indian_Defence

I've corrected it. happy.png

Yigor

The line chosen by Explorer (Stockfish 8) after 1. d4 Nf6 at d=20 (completed):

 

 

So, it opts for QGD: Symmetrical KK (KK = King's Knight).

Yigor

I'm now adding variations with 10K+ master games more systematically. In particular, Trompowsky attack was added at 3 plies.

Yigor

I've added Catalan and Anti-Nimzo-Indian at 5 plies. settings.png

Yigor

The investigation continues, I'm adding openings to 5 plies.

Yigor
cjxchess16 wrote:

The Pirc defense is also reachable from 1 d4 Nf6:

 

Yes, U absolutely right but, at the moment, I fixed the low limit of 10000 master games in Explorer.

Yigor

I indicated W-edged openings. settings.png