SmithyQ – Checkia, Feb 2017: A KID Endgame

Let me start by saying this: I don’t understand the KID and I’m terrible at endgames.

The King’s Indian Defence is one of those openings I’ve never understood.  I mean, I get the basics.  Black gives up the centre so he can counter-attack with pieces and pawn breaks later.  When I play it as White, no matter how well I play, Black always ends up with active counterplay.  When I play it as Black, I lose in 12 moves.

The KID appeals to aggressive players, but it’s a high-risk opening.  What’s the worst possible scenario?  You don’t get to attack and you get stuck in an endgame.  I’m not the best endgame player, but if I can end up in a no-risk position with decent winning chances, I take it every time.  That’s what happened here.

My opponent played much stronger than his rating, and this game was quite intense.  Let’s take a look.

[Event “2- Chess is for fun – Round 4”]
[Site “”]
[Date “2017.01.05”]
[Round “?”]
[White “SmithyQ”]
[Black “Checkia”]
[Result “1-0”]
[ECO “E70”]
[WhiteElo “2154”]
[BlackElo “1772”]
[PlyCount “117”]
[EventDate “2017.??.??”]
[TimeControl “1”]

{I win a pawn, but it takes 60 moves to turn that into a win.} 1. d4 Nf6 2. c4
g6 3. Nc3 Bg7 4. e4 d6 {The KID leads to positions I generally don’t
understand. My overall record, with both colours, is around 40%.} 5. Nge2 O-O
6. Ng3 {Because of that, I looked for various sidelines, and my favourite is
this, the Hungarian variation. The main idea is to keep the f-pawn mobile:
White can play f3 of f4-f5 as well as h4-h5. Even better, Black’s standard
attacking plans don’t work as well, as it’s harder to play f5. Most Black
players don’t know how to play against this, which can lead to quick wins.
I’ve included two in the annotations below.} e5 (6… c5 7. d5 e6 8. Be2 a6 9.
O-O exd5 10. exd5 Nbd7 11. Bg5 Qc7 12. Qd2 Ne5 13. f4 Ned7 14. f5 Ne5 15. Bxf6
Bxf6 16. Nge4 Bg7 17. f6 Bh8 18. Qh6 Re8 19. Ng5 {1-0, SmithyQ – Geralt, 2004.}
) (6… b6 7. Be2 c5 8. d5 e6 9. O-O exd5 10. cxd5 Na6 11. Bg5 Nc7 12. f4 Qe8
13. Re1 Bb7 14. Bxf6 $5 Bxf6 15. e5 dxe5 16. Nge4 Bg7 17. f5 gxf5 18. Nd6 Qb8
19. Nxf5 Rd8 20. Bc4 Ne8 21. Qg4 Bc8 22. Nh6+ Kf8 23. Qh5 Nd6 24. Rf1 Rd7 25.
Nxf7 Rxf7 26. Rxf7+ Nxf7 27. Rf1 {1-0, SmithyQ-Blanker, 2004.}) 7. d5 c6 8. Be2
cxd5 9. cxd5 a6 10. Bg5 h6 11. Be3 b5 {White has played standard KID moves so
far. Black has played a6 and b5. When does Black ever play that in the KID?
Not often. Black is already feeling uncomfortable in my pet variation.} 12. O-O
Nbd7 13. Qd2 Kh7 {This is the natural move, stopping the threat against h6,
but h5 appears even better.} (13… h5 {[%csl Rg3][%cal Rh5h4] The computer
really likes this move, and thinks the game is equal. It highlights the one
downside of the Ng3 variation, namely, the Knight has no good retreat squares.
White either needs to play h4 himself, which is slightly weakening, or Bg5,
which is only a temporary solution. For instance,} 14. Bg5 Qe8 {Breaking the
pin.} 15. Rac1 Nh7 16. Be3 h4 17. Nh1 {White’s Queenside advantage is negated
by his miserable Knight. Black will have time to start his counterplay on the
Kingside.}) 14. b4 {[%cal Ga2a4,Ga4b5,Gc3b5,Ge2b5] At this point, I need to
think of a plan. My main idea is to play on the Queenside. I want to play a4,
exchange on b5 and then group all my pieces against the resulting weak pawn.
If I play a4 now, though, Black can reply with b4. White’s probably still
doing okay, but that would make it much harder to break through. b4 stops that
idea, and so I played it first.} ({The computer moderately prefers} 14. Rfc1 {
which also looks fine.}) 14… Nb6 15. a4 bxa4 16. Nxa4 Nxa4 17. Rxa4 {[%csl
Ga6][%cal Ge2a6,Ga4a6,Gf1a1,Gd2a2] My plan has gone swimmingly. I have open
lines all across the Queenside, and I have two weaknesses to attack: d6 and a6.
d6 isn’t too easy at the moment, but I can bring everything against a6 very
quickly. That pawn probably isn’t long for this world.} h5 {Black tries to
stir up active counterplay, which is the right decision. The threat is h4,
where my Knight would be stuck on the awful h1 square.} ({Black also has
another option, one I looked at quite deeply.} 17… Nxe4 $5 18. Nxe4 f5 {
After the Knight moves Black will play f4 and win back his piece. Things now
get sharp, but I think White maintains the balance with} 19. Nxd6 $1 Qxd6 20.
Bc5 Qd7 21. Bxf8 Qxa4 22. Bxg7 Kxg7 23. Qc3 $1 {[%cal Gc3g7]} Kf6 24. f4 {
[%cal Gf1a1,Gf1c1,Gc3c6,Gc3c7,Gd5d6,Ge2f3,Ge2a6,Gf1e1,Ge1e5] Black now looks
forced to play Qe8, retreating. White now has lots of active plans, whereas
Black can only defend. An interesting variation, but better for White I think.}
) 18. Bg5 Qb6 19. Rfa1 {Very logical, bringing the last piece into the attack
and ratcheting up the pressure, but taking advantage of the pin also looked
promising.} (19. b5 $5 {[%cal Ga4a8] The pawn is immune because of the pin. If
I can capture on a6 with a pawn instead of a Rook, my position becomes much
better. Much harder to stop the a-pawn from Queening.} a5 ({If Black doesn’t
play a5, then} 19… Nd7 20. Be3 Nc5 21. Bxc5 dxc5 22. Rfa1 {and I’m about to
win the pawn with a winning position.}) 20. Rfa1 Ng4 {[%cal Rg4f2,Rb6f2]} 21.
Qe1 {Black is attacking, but he only has two pieces, and it’s hard to bring
the others into the game. White can easily push them back and take over the
game.} Bh6 22. Bxh6 Kxh6 23. Bd3 Kg7 24. h3 Nf6 25. Rxa5 $16) 19… Rb8 {
A surprisingly good move. Sure, it weaknes a6, but that pawn was going to fall
anyway. It gets out of the pin, stopping b5 ideas like the above variation,
and it pressures by b-pawn. If the a- and b-pawns gets traded, my advantage
becomes almost nothing.} 20. Be3 ({The computer prefers} 20. Bd3 {as it
defends e4 (taking the sting out of h4 from Black) and gives the e2-square for
the Knight.}) 20… Qc7 21. f3 {Instead of playing 20.Bd3, using the Bishop to
defend the pawn, I wanted to play f3, permanently stopping any Nxe4 ideas like
in the variation on move 17. If I keep the center solid, Black shouldn’t have
any counterplay. I was both right and wrong!} h4 {[%cal Gg3h1,Gh1f2,Gf2d3,
Yg3f1,Yf1e3,Ye3c4,Ye3c2] This lead to a painfully hard decision: where to put
my Knight. Neither square is great, but they both allow long routes to decent
squares. I picked Nh1 because it appeared faster, as Nf1 requires the Be3 to
move to get to a good square, and it really doesn’t have a better square.} 22.
Nh1 Nh5 23. Bxa6 Bxa6 24. Rxa6 Rfc8 25. Ra7 $6 {It’s amazing how this move,
which looks so tempting, hitting the Queen and targeting the weak f7-pawn,
getting my Rook on the seventh rank … is nonetheless inaccurate. This one
move gives Black an incredible amount of play.} ({I was so fixated on my
control of the a-file that I missed the power of occupying the c-file.} 25. Rc6
Qd7 26. Rac1 Rd8 27. Rc7 Qe8 $16 {This would be the logical way to play,
keeping complete control. White should be winning, as Black has nothing.})
25… Qc3 26. Qxc3 ({I originally wanted to play Nf2, but it lead to a
terrible looking position.} 26. Nf2 Qxb4 27. Rxf7 Qb3 28. Qe1 Rc2 29. Kh1 Qb2
30. Qg1 {The computer calls this best play for White, and even says White has
an advantage, but just look at it. White is passive, it’s very hard to
untangle and Black just keeps increasing the pressure. I’d much prefer Black,
pawn defecit or no.}) 26… Rxc3 27. Bd2 Rc2 28. R7a2 {I spent forever coming
to this decision. I wanted to keep a Rook on the seventh rank, but it seems
that it’s more important to curtail Black’s activity.} ({The other option} 28.
Be1 h3 ({I was more worried about this, as my back rank is potentially weak
and Black has dangerous activity.} 28… Rbc8 29. Rxf7 Kg8 30. Rd7 Bh6 31. Kf1
h3 32. Nf2 hxg2+ 33. Kxg2 Rc1 34. Rxc1 Rxc1 {I got to hear in my analysis and
assumed Black was winning a piece. I missed} 35. Ng4 $1 {with the
counterattack, securing a big White advantage.}) 29. gxh3 Nf4 30. Nf2 f5 {
Again, the computer really likes White, but I think Black has the far easier
play. White is one inaccuracy away from losing, and his minor pieces are
terrible.} (30… Rbc8 31. Rxf7 Rc1 32. Rxc1 Rxc1 33. Kf1 {is another awkward
looking line, though White’s three-pawn advantage makes it hard for Black to
win I’d imagine.})) 28… Rc7 {I was expecting R8c8, where I wanted to play b5.
} 29. g3 $6 {Because my opponent played (what I felt at the time) an inferior
move, I thought I had time to keep his Knight passive. In all the variations
above, Black kept getting good play by bringing his Knight to f4. I figured
I’d stop that, but it costs a tempo.} (29. Nf2 {This is the most obvious and
best move. Rather than worry about restricting Black’s Knight, I should work
on increasing my own Knight, which is sitting sadly in the corner, alone.})
29… hxg3 30. hxg3 f5 $1 {[%csl Gg3][%cal Gh1g3,Gh5g3] Black plays actively,
and we really see the problem with g3. I need to waste time to defend the
newly weak pawn, which Black can then use to create more counterplay. None of
my pawns are especially strong after an exchange on e5.} 31. Kg2 fxe4 32. fxe4
Bh6 {This is a well-timed exchange. Black’s Bishop is passive, and he trades
it for the main defender of b4. If that pawn were to fall, so would any
serious winning chances I had.} (32… Nf6 33. Nf2 Bh6 34. Rh1 {This shows why
Black cannot delay in offering the exchange.} Ng8 35. Ng4 {[%cal Gd2h6,Gg4h6,
Gh1h6]}) 33. Ra7 {On the downside, going for the trade of Bishops allows me to
exchange a pair of Rooks. As Black has more activity in most of the previous
lines, this should benefit me. Indeed, if I could somehow trade all Rooks, I’d
be winning rather trivially in almost all minor piece endgames.} Rbc8 34. Rxc7+
Rxc7 35. Bxh6 (35. Ra2 {This was the main alternative, keeping the tension and
not allowing the enemy Rook to infiltrate.} Rc4 36. Bxh6 Kxh6 37. Rb2 $1 {
is the critical idea that I underestimated. I was too concerned about losing a
pawn to see that my b-pawn becomes a monster.} Nf6 (37… Rxe4 $6 38. b5 Rc4
39. b6 Rc8 40. Nf2 Nf6 41. Kf3 Nxd5 42. b7 Rb8 43. Ne4 Nc7 44. Nxd6 {Material
is equal, but the e-pawn will fall and Black is struggling to keep my pawn
from promoting. In particular, my piece coordination is much better.}) 38. Nf2
Nxe4 39. Nxe4 Rxe4 40. Kf3 Rc4 41. b5 {This position is probably drawn with
perfect play, but it offers White lots of winning chances.} Rc7 42. b6 Rb7 {
White can now slowly push his King forward, invade on the squares, marking
time where needed with Rook moves up and down the file. Black must sit and
defend perfectly, and a single inaccuracy likely leads to a loss.}) 35… Kxh6
36. Rb1 Nf6 37. Nf2 Rb7 {Perhaps not a bad move, but unnecessarily passive.} (
37… Rc2 38. Kf1 Nd7 {[%csl Gb6][%cal Gd7b6] The Knight will blockade the
pawn and the Rook will keep my King under wraps. g3 is always weak. I don’t
know how I make progress without giving up a pawn.}) {I should point out here
that White, despite being a pawn up, has no clear winning strategy. Okay,
obviously I want to promote a pawn, but that’s not easy to do. Black’s Knight
is in a perfect spot to hit my vulernable pawns, and if my King tries to
support my b-pawn, Black’s King can invade on the Kingside. Don’t get me wrong,
White has an advantage and winning chances, but not enough to make it a
winning position.} 38. Kf3 {[%cal Gf6e4,Yf6h5,Yh5g3]} Nd7 39. Nd3 Kg5 40. Rc1
$6 {I spend the next five moves essentially wasting time, but it doesn’t
change the character of the position.} ({Passed pawns must be pushed!} 40. b5 {
This was the right plan, which I saw, but I hoped I could get Black to fall
for a trick.}) 40… Kf6 ({I had hoped Black would overreact to my Rook coming
to c6} 40… Nb8 41. Rc6 $1 {When I just play it anyway!} Nxc6 42. dxc6 Rc7 43.
b5 {The two connected pawns sweep the Rook away.} Kf6 44. Nb4 Ke6 45. b6 Rc8
46. b7 Rb8 47. Na6) 41. Rc6 Ke7 42. Rc4 Nb6 43. Rc1 Nd7 44. Rb1 {My Rook
finishes its journey and is back where it started. Not the greatest endgame
technique on my part, but I didn’t lose anything: I’m still up a pawn with
chances for a win.} Nb6 45. Kg4 ({The computer says this is best play} 45. Rh1
Kf6 46. Ra1 Nd7 {but it just looks like flailing around to me.}) 45… Nc4 46.
b5 $1 Na3 {Black sees a fork and immediately jumps for it…} 47. Rb3 Nxb5 {
[%cal Gb3b7] But now he falls into a pin, one he cannot escape.} (47… Nc4 {
was better, but it allows me to make progress.} 48. Rb4 Ne3+ (48… Nb6 49. Kg5
Kf7 50. Nb2 Kg7 51. Nc4 Nxc4 52. Rxc4 Rxb5 53. Rc7+ {is winning and similar to
the game.}) 49. Kg5 Nf1 50. g4 Kf7 51. b6 {White has made excellent progress,
and Black’s Knight has mostly moved itself out of play.}) 48. Kg5 Rb8 49. Nb2 {
[%cal Gb2c4,Gc4a3,Ga3b5] If Black does nothing, I will simply maneuver my
Knight to attack and win the pinned piece.} Kf7 50. Nc4 Rc8 51. Nxe5+ $1 {
Winning a pawn, but more importantly, securing a dominating position.} dxe5 52.
Rxb5 Rc4 53. Rb7+ Kf8 54. Kf6 $1 {[%csl Ge7,Gf7,Gg7][%cal Gb7b8] Even stronger
than taking on g6, as I now threaten mate.} Ke8 55. Kxe5 {And it allows me
take on e5, keeping my e-pawn. Two connected passed pawns give me an easily
winning position, which I convert in just a few more moves.} Rc3 56. Ke6 Kf8
57. d6 Rc8 58. d7 Rd8 59. Rc7 {[%cal Gc7c8] The threat of Rc8 cannot be
parried, so my opponent resigned. This was a ridiculously tough endgame to
convert, even being up a pawn, and it really shows the latent activity
inherent in the KID. Black’s pieces got active out of nowhere, and it took
great care to keep winning chances.} 1-0


First, I need to start playing the KID so I can better understand these positions.  I’m tired of having good positions, being in complete control, only for Black to get play seemingly out of nowhere.

Second, don’t get discouraged in the endgame.  I didn’t play this endgame very well, but it didn’t matter.  I had an extra pawn, I had all the winning chances, and I kept going until something popped up.  Again, it might not be the most exciting way to win, but the rating points all count the same.

Finally, I want to give huge props to my opponent.  Though there was a large rating gap between us, Checkia played incredibly well.  I had to fight tooth and nail for every inch, and it took 50 moves to finally prove an advantage.  Well played, sir, and if you keep playing like this then our ratings will get much closer.

2 thoughts on “SmithyQ – Checkia, Feb 2017: A KID Endgame

  1. Gringo

    This is a very detailed analysis on your part. It’s going to take a few days to go through and learn from it. Your games are a step above since your ascention into the 2100+ range. Your analysis is easier to follow as only one side needs more of the revising instead of both both sides requiring it.

    Perhaps keeping a running log of games going forwards from here isn’t a bad idea. I do learn from your older games as well. Perhaps, having the analysis in this new chess base format makes it easier to follow different paths instead of looking at a static screen, which for someone at a lower rating isn’t the easiest thing to do.

    Checkia must be congratulated here for playing through such a wide rating difference and still keeping it so close.

    You mentioned that your endgame isn’t the strongest. Is it because you haven’t gone through the Endgame Course yet or is it because you need more specific practice or training to get a read edge?


  2. JP Post author

    I had spent between 2-3 hours thinking and analyzing this game while it was going on, so writing the analysis was fairly easy. All the lines were already fresh in my head, and I just needed a quick double-check with the engine to confirm.

    The ChessBase format makes it so much easier to do things. Before, I had to manually create each diagram, manually type the .pgn and so many other headaches. Now I just analyze the game as always, click one button and have an interactive board embedded in my blog. So easy, and it’s going to let me do far more games and give more variations, knowing that the reader can follow. It also takes far less time.

    Eventually I plan to have every game I’ve played on uploaded and analyzed.

    As for the endgame, I’ve just never studied it. I know the bare basics (simple mates, square of the pawn, that stuff), but that’s it. My only real endgame knowledge comes from watching Endgame Expert videos, and I hope to find time to study it in depth this summer.

Leave a Reply

Your email address will not be published.