Eval of mainline King's Indian

[EroSennin]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Nd3 f5 11. f3

So computers love white's position here. I have analysed this opening a bit but couldn't really come to a full conclusion. Black's best try might be some blockade with like c5 which computers don't understand. Anyway the question is black bust or not? I would really like to play the king's indian but computer eval scares me.

[velmarin]

With this position and as a reference a database "PlayBase Chess Engines (3 million games)":
Percentage from the White side:




This opening book says my consult, "18 Gbs and 15 million games"
Percentage from the White side:




After c5:
Percentage from the White side:

[lucasart]

Easier to see like that:
[pgn]
1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Nd3 f5 11. f3
[/pgn]

[Sven]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Nd3 f5 11. f3

So computers love white's position here. I have analysed this opening a bit but couldn't really come to a full conclusion. Black's best try might be some blockade with like c5 which computers don't understand. Anyway the question is black bust or not? I would really like to play the king's indian but computer eval scares me.

Isn't the following line even slighty better for white than the one above?
[pgn]1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Be3 f5 11. f3 f4 12. Bf2[/pgn]
Sven

[Don]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Nd3 f5 11. f3

So computers love white's position here. I have analysed this opening a bit but couldn't really come to a full conclusion. Black's best try might be some blockade with like c5 which computers don't understand. Anyway the question is black bust or not? I would really like to play the king's indian but computer eval scares me.

An observation that Larry made was that almost all computers misevaluate this type of position - the type where white generally has a big space advantage but black is quite solid.

I would say if you are comfortable with the black side of this, play it! It's probably ok theoretically.

[jdart]

This opening is notoriously hard to evaluate by computers. The problem is, Black's plan is basically a very gradual but often deadly attack on the kingside. If White parries this he is ok. If White doesn't take effective measures in time, he will lose. This kind of super long-range planning is not something current chess programs do well.

--Jon

[Don]

This opening is notoriously hard to evaluate by computers. The problem is, Black's plan is basically a very gradual but often deadly attack on the kingside. If White parries this he is ok. If White doesn't take effective measures in time, he will lose. This kind of super long-range planning is not something current chess programs do well.

--Jon

You are describing a position that is hard to play but the question is whether it can be evaluated with a more or less correct score - one that sums up the chances well. Surprisingly, the two things are not always the same. A program can often win positions (i.e. play the right moves) without even knowing a position is a win.

In this case, you make the argument that the computer doesn't know that white has a good position but I don't think that is the problem because programs overestimate whites potential. The problem is that they do not understand that black has resources.

This is something that Larry and I are definitely interested in because we are keenly interested in improving Komodo's understanding of all kinds of difficult chess positions.

[Eelco de Groot]

This opening is notoriously hard to evaluate by computers. The problem is, Black's plan is basically a very gradual but often deadly attack on the kingside. If White parries this he is ok. If White doesn't take effective measures in time, he will lose. This kind of super long-range planning is not something current chess programs do well.

--Jon

You are describing a position that is hard to play but the question is whether it can be evaluated with a more or less correct score - one that sums up the chances well. Surprisingly, the two things are not always the same. A program can often win positions (i.e. play the right moves) without even knowing a position is a win.

In this case, you make the argument that the computer doesn't know that white has a good position but I don't think that is the problem because programs overestimate whites potential. The problem is that they do not understand that black has resources.

I think that you maybe did not interpret that correctly, because you seem to be arguing the same thing here Don, as Jon does. The problem is not with the evaluation of White, because the program can, by measuring mobility, see that White has a lot of extra space. But it is much harder for a program to see if Black could still mobilize his forces quickly once it is clear where the pieces have to go. That is much harder to see in the short term. And what from Jon's description would follow that there is at least part of a King safety problem, which is encouraging if you look at the position at the end of Sune's line

[D]r1bq1rk1/pppnn1bp/3p2p1/3Pp3/2P1Pp2/2N2P2/PP2BBPP/R2QNRK1 b - -
This should look familiar for the readers of Lyudmil Tsvetkov's posts, the outline of at least a pawn formation that could lead to King Safety asymmetry: the "sniper wall" rears its ugly head again. Any attacks on Black's shortcastled position would be blocked effectively, but if Black could get some pieces looking in the direction of the long diagonal c8 - h3 the white King is not yet under direct attack, but there is indirect danger I think. I am sure the structure is not hard to program for you Don. And I am even more sure Larry would love to play with it

I can't guarantee the program would actually play better though, if it recognized the structure. In this case it is also about long rang planning if I read Jon's description. And the mobilization of Black's forces.

Eelco

This is something that Larry and I are definitely interested in because we are keenly interested in improving Komodo's understanding of all kinds of difficult chess positions.

[CRoberson]

1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4 d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Nd3 f5 11. f3

So computers love white's position here. I have analysed this opening a bit but couldn't really come to a full conclusion. Black's best try might be some blockade with like c5 which computers don't understand. Anyway the question is black bust or not? I would really like to play the king's indian but computer eval scares me.

An observation that Larry made was that almost all computers misevaluate this type of position - the type where white generally has a big space advantage but black is quite solid.

I would say if you are comfortable with the black side of this, play it! It's probably ok theoretically.

I coded Telepath to be more human like and it says that black is as happy as a rooster in front of the hen house which is not as happy as a rooster in the hen house.

[carldaman]

Funny thing, if I had not been busy, I would gladly have posted on this topic of "Engines & The Classical KID" myself, and a lot sooner. It would appear that the engines are perhaps getting too much of a 'bad rap' about their handling of this opening.

The question is actually two-fold - are the engines underestimating Black's attack, or are they even capable of carrying it out to begin with?

Just a few weeks ago, I ran a brief test at an intermediate time control (40 moves/20 min), pitting Zappa Mexico II vs Deep Hiarcs 14, running on 4 threads, with 1GB of RAM each. The exact opening line is from Noomen's Kings Indian Suite.

Let me first post this diagram to whet one's appetite

[D]5rk1/p5b1/1p1p4/3P4/2N1pppp/8/P7/1R3KB1 w - - 0 40

This surreal position occurred after Black's 39th move, e5-e4. Hiarcs conducted the attack with great mastery, and I must say it wasn't the only time it did so, to its credit. Out of two games with Black, it won both in dazzling fashion.

The trick is to get Hiarcs (and other top engines) to play 11...f4 on their own. Some engines do this, but they're not among the top ten, as I found out. The strongest engines seem more concerned with playing b6/a5 first to slow down White's Queenside advance, and only then pushing f5-f4, or just opting for the less convincing f5xe4, which is not in the spirit of the position.

Till later,
CL

[pgn]

[Event "tourn87-4020-E99-1"]
[Site "?"]
[Date "2013.01.17"]
[Round "1"]
[White "Zappa Mexico II x64"]
[Black "Deep HIARCS 14 WCSC"]
[Result "0-1"]
[ECO "E99"]
[Annotator "0.61;0.64"]
[PlyCount "88"]
[EventDate "2013.01.17"]
[EventType "tourn"]
[Source "Doe"]

{Intel(R) Core(TM) i3-2120 CPU @ 3.30GHz 3292 MHz W=14.3 plies; 1,974kN/s; 55
TBAs B=20.6 plies; 1,569kN/s; 2,064 TBAs} 1. d4 Nf6 2. c4 g6 3. Nc3 Bg7 4. e4
d6 5. Nf3 O-O 6. Be2 e5 7. O-O Nc6 8. d5 Ne7 9. Ne1 Nd7 10. Be3 f5 11. f3 f4
12. Bf2 g5 13. Rc1 Rf6 14. Nd3 Rh6 {[%eval 64,20] [%emt 0:00:39]} 15. h3 {
[%eval 61,17] [%emt 0:03:37] (c5)} Nf6 {[%eval 48,20] [%emt 0:01:01] (c5)} 16.
Re1 {[%eval 75,15] [%emt 0:02:33]} b6 {[%eval 57,20] [%emt 0:00:52]} 17. b4 {
[%eval 64,18] [%emt 0:01:59]} Bd7 {[%eval 54,19] [%emt 0:00:29] (a5)} 18. c5 {
[%eval 72,17] [%emt 0:02:25]} Ng6 {[%eval 53,19] [%emt 0:00:40] (Rg6)} 19. Qd2
{[%eval 79,16] [%emt 0:01:07]} Nh4 {[%eval 38,18] [%emt 0:00:40]} 20. Bf1 {
[%eval 74,15] [%emt 0:01:17] (a3)} Qe8 {[%eval 14,19] [%emt 0:00:30] (Nh5)} 21.
Red1 {[%eval 79,13] [%emt 0:00:47] (Bxh4)} Nxf3+ {[%eval -125,19] [%emt 0:02:
30] (Nh5)} 22. gxf3 {[%eval 0,15] [%emt 0:00:43]} Bxh3 {[%eval -125,18] [%emt
0:00:37]} 23. cxd6 {[%eval 0,14] [%emt 0:00:58] (Bxh3)} cxd6 {[%eval -222,19]
[%emt 0:00:28]} 24. Bxh3 {[%eval -96,15] [%emt 0:01:32]} Rxh3 {[%eval -227,18]
[%emt 0:00:05]} 25. Kf1 {[%eval -105,15] [%emt 0:00:23]} g4 {[%eval -272,20]
[%emt 0:00:26]} 26. Qb2 {[%eval -151,15] [%emt 0:00:54]} Qg6 {[%eval -275,19]
[%emt 0:00:10] (Qh5)} 27. Rc2 {[%eval -145,13] [%emt 0:00:07]} Rxf3 {[%eval
-309,19] [%emt 0:00:22] (Qh5)} 28. Ne1 {[%eval -166,13] [%emt 0:00:11]} Rh3 {
[%eval -310,18] [%emt 0:00:11]} 29. Bg1 {[%eval -177,13] [%emt 0:00:09]} Rf8 {
[%eval -392,19] [%emt 0:00:43] (f3)} 30. Rh2 {[%eval -189,12] [%emt 0:00:07]
(Rcc1)} Rxh2 {[%eval -419,20] [%emt 0:00:47]} 31. Qxh2 {[%eval -221,14] [%emt
0:00:11]} Nxe4 {[%eval -426,21] [%emt 0:00:40] (Nh5)} 32. Nxe4 {[%eval -235,14]
[%emt 0:00:07]} Qxe4 {[%eval -426,18] [%emt 0:00:06]} 33. Qc2 {[%eval -247,15]
[%emt 0:00:12] (Rc1)} Qxb4 {[%eval -443,19] [%emt 0:00:14]} 34. Rb1 {[%eval
-251,14] [%emt 0:00:04] (Rc1)} Qa3 {[%eval -467,20] [%emt 0:00:26]} 35. Rb3 {
[%eval -251,14] [%emt 0:00:06]} Qa4 {[%eval -475,21] [%emt 0:01:08]} 36. Rb1 {
[%eval -285,14] [%emt 0:00:10] (Qe2)} Qxc2 {[%eval -591,23] [%emt 0:00:54]
(Qa6+)} 37. Nxc2 {[%eval -248,9] [%emt 0:00:00]} h5 {[%eval -581,25] [%emt 0:
01:25] (g3)} 38. Na3 {[%eval -290,12] [%emt 0:00:02]} h4 {[%eval -594,24]
[%emt 0:00:55] (g3)} 39. Nc4 {[%eval -350,13] [%emt 0:00:07]} e4 {[%eval -617,
23] [%emt 0:01:23] (g3)} 40. Nxd6 {[%eval -323,12] [%emt 0:00:05] (Re1)} e3 {
[%eval -655,23] [%emt 0:00:58] (Rd8)} 41. Rd1 {[%eval -462,15] [%emt 0:00:51]
(Re1)} Bh6 {[%eval -807,22] [%emt 0:00:55] (g3)} 42. Nc4 {[%eval -736,14]
[%emt 0:00:55] (Re1)} h3 {[%eval -928,22] [%emt 0:00:25]} 43. Ne5 {[%eval -923,
15] [%emt 0:00:50] (d6)} f3 {[%eval -1598,21] [%emt 0:00:33]} 44. Nxg4 {[%eval
-923,15] [%emt 0:00:00] (Re1)} e2+ {[%eval -2042,25] [%emt 0:00:17] (h2)} 0-1

[/pgn]