Analysis of my recent King's Indian Defense game

Question

I'm looking for any feedback on positional ideas I've missed on a recent KID game I played (as black). We were both about 1900 Elo, although I was slightly higher rated. I've obviously already analyzed it with an engine, pointing out a few of my missed tactical shots in the game in these annotations. But this was a very complicated game, so I would like any and all feedback on better lines and/or positional ideas for either side.

[FEN " "]
1. c4 Nf6 2. Nc3 g6 3. e4 d6 4. Be2 Bg7 5. d4 O-O 6. Rb1 a5 { Trying to shut down the obvious b4 plan by white. The more natural c5 is slightly better, but I had already decided on going for a classical KID with an e5 break. } 7. Nf3 e5 8. Bg5 (8. dxe5 dxe5 9. Qxd8 Rxd8 10. Nxe5 Nxe4 11. Nxe4 Bxe5 $17) 8... Nc6? { Mistakenly assuming that white's only play was d5, and completely missing that this move loses an exchange. } (8... h6 {The move I'll play next time I see this position. } 9. Be3 (9. Bh4 g5) 9... Ng4 10. Bc1) 9. dxe5 dxe5 10. Qxd8 Rxd8 11. Rc1? { This mistake fails to capitalize on my poor play. } (11. Nd5 Nxd5 12. cxd5 f6 13. dxc6 fxg5 14. Nxg5 bxc6 { Engine-like best play temporarily holds on to all material but completely ruins all chances of black fighting for a win. }) 11... h6 12. Be3 Ng4 (12... Nd4!? 13. Nxe5 Nxe2 14. Kxe2 Nxe4 15. Nxe4 Bxe5 { Wins the two bishops and retains active pieces. }) 13. Bd2 Be6?! (13... Nb4!? 14. a3 Na6 { This maneuver keys in on the weak c5 square. But making repeated knight moves means the queenside development is lacking. }) 14. O-O Nd4!? { Acquiring a passed pawn almost by force. The knight lands on a beautiful square and cannot be kicked away. } 15. Nxd4 exd4 16. Nb5 Rd7?! {An awkward move - I thought I was better here, but my opponent comes at me with a lot of pressure.} (16... Rac8?! 17. Bxa5 d3! 18. Bxg4 Bxg4 19. f3 Be6 20. b3 Rd7 { I'm unconvinced by the computer line which makes no sense to me. }) 17. f4!! { An extremely strong reply and a move I had completely missed. With f5 coming, the knight can be left stranded on g4 in many lines. } 17... f5!? { Gives white a passed pawn with e5, but I figured black's would be stronger. } (17... c6 18. f5)  (17... Re8 18. f5 Bxf5 { The computer move to stop f5. }) 18. c5!? { A very interesting pawn sacrifice that turns the knight (and eventually c1 rook) into incredibly active pieces. } (18. exf5 Bxf5 19. c5) 18... fxe4 19. c6! bxc6 20. Rxc6 d3!! { There's no time to cower from each other's threats. Black is pushing through in the center and white is crashing through on the queenside. } 21. Bd1 Bf5? { Immediately after deciding that I had to play actively, I falter with a slow move. I needed to keep up the active play with the brilliant Bd5! } (21... Bd5!! 22. Bxg4 (22. Rxc7 Rxc7 23. Nxc7 Bd4+ 24. Kh1 Nf2+ 25. Rxf2 Bxa2!!)  (22. Rxg6 Nf6 23. Nc3 Kh7 24. f5 Bf7 25. Rg3) 22... Bxc6 23. Bxd7 Bxd7) 22. h3 Nf6 23. b4?? { I truly have no idea why this move was played. } 23... axb4 24. g4 Rb8!? { Attack - counterattack. } (24... Rxa2?? 25. Bb3+) 25. Nxc7 Rc8 26. gxf5 Rdxc7 (26... gxf5 27. Bb3+ Kh7 28. Rfc1 {The reason you have to immediately take the pinned knight.}) 27. Bb3+ (27. Ra6 gxf5 28. Bb3+ Kh7) 27... Kh7 28. Ra6? (28. fxg6+ Kxg6 29. f5+ Kh7 { The line I expected. Does the bishop pair and active rook make up for black's 2 passed pawns? }) 28... gxf5 29. Kf2?? { An interesting idea - allowing the king to become a blockader and the rook to use the open g file. But it's ultimately way too slow. White is down 2 pawns and must show compensation. } (29. Be6 Re8 30. Bxf5+ Kh8 31. Bxb4 { The line I expected. While material is balanced, I felt that black's two central pawns should rule the day. }) 29... Rc6 30. Rxc6 (30. Ra7 R8c7 31. Rxc7 Rxc7 32. Be6 Rc2 33. Bxf5+ Kh8 { I was sure I had enough to win in this line. }) 30... Rxc6 31. Be3 Bf8! { Trading bishops and ensuring that black's king can rely upon the dark squares. } 32. Rd1 (32. Bd4 Bc5 33. Ke3 Bxd4+ 34. Kxd4 Nh5 35. Ke3 Rg6 { An interesting sideline I was trying to evaluate. }) 32... Bc5 33. Rg1 Bxe3+ 34. Kxe3 Rd6!! { Giving up the open file, but preparing Nd5+ after which the game is immediately decided. } 35. Rg2 (35. Rc1 Nd5+ 36. Bxd5 Rxd5 37. Rc4 d2) 35... Nd5+ 36. Bxd5 Rxd5 37. Rg1 d2 { Simplification. From this move onwards, I saw to the end. } 38. Ke2 d1=Q+ 39. Rxd1 Rxd1 40. Kxd1 Kg6 41. Kc2 Kh5 42. Kb3 e3! { And the white king must return to stop the queening e pawn. } 43. Kc2 Kh4 44. Kd3 Kxh3 45. Kxe3 Kg3 {White surrenders.} *

In particular, I'm curious about what people think about:

• 14... Nd4 - is it too premature to start thinking about passed pawns as soon as the queens are off?
• 16... Rd7 - this awkward move was very hard to make. Is black just worse here, even with the passed pawn?
• 18. c5 - sacrificing a pawn, but drumming up a lot of counterplay on the queenside. I felt that this move was a blunder during the game, but the pressure I came under seems to prove that it was not.
• 23. b4 - What's the point of this move?
• 29. Be6 - Is my assessment of this sideline accurate? Does black have enough to win?

Looking forward to your analysis! Hopefully others find this game instructive; I certainly learned a lot playing (and subsequently analyzing) this middlegame.

Answer 1

14... Nd4 - is it too premature to start thinking about passed pawns ...?

...Nd4 is highly desirable for a number of reasons! Besides the fact that White is inclined to capture an give you a passed pawn, you get more space, extend your dark square bishop and free your c-pawn to cover the d5 hole.

16... Rd7 - this awkward move was very hard to make. Is black just worse here?

Nothing wrong with Rd7 that I can see.

c5 - sacrificing a pawn, but drumming up a lot of counterplay on the queenside. I felt that this move was a blunder during the game, but the pressure I came under seems to prove that it was not.

18.c5?? a terrible blunder! White gives you connected passers in the middle of the board for what -- a chance to get back the crappy pawn on c7? This is the losing move in my opinion.

Besides those questions let me comment: I don't agree with a lot of your annotations for example, why is 17.f4 a double exclam? At best it is !? of course after 17...f5 (which is standard) White should have played 18.e5

I do agree 13...Be6?! is poor, but favor 13...f5 over your suggested Nb4.

Answer 2

14...Nd4: it's never too early to think about passed pawns.

16...Rd7: I think black is fine here, but I would consider Nf6 with a counterattack on e4 or d3.

18. c5: I would consider e5, shutting down the B-g7, or Bxg4 to win a tempo and play c5 next move.

19. b4: This is a question for your opponent. I prefer a4, fixing the weakness on a5, or Bb3+, controlling d5 and allowing the other rook to enter the game.

20. Be6: The assessment is correct as 32... Nd5 wins outright.

In the first variation, 10. Nxe5 is incorrect. Better is 10. Bg5 with the threat of Nd5.

I've suggested this before: post the game on lichess.org. It's free to join, the analyses will be better, and this forum isn't really meant for this type of question.

Note: I'm only a 2000, so my superficial opinions may not be any better than yours.

Answer 3

It is a little awkward to comment while knowing that you have already checked the game with an engine, so you may be aware of tactical flaws in our suggestions, but does 23.g4!? wins material ?

Bf5 is trapped, and after 23...e3 24.Bb3 Kh8 25.Bxe3 Be4 26.Rxc7 seems to collect two pawns. 23...Rb8 is much less strong than in the game since 24.a4 (probably best), 24.Na3 and even 24.Nxc7 are available.