SmithyQ-Themis-Neo, April 2017: My First Loss of 2017

Well, here it is, my first loss of the new year, and my first loss in months.  Most of that is because I’ve only played a handful of games in those months, but still, that’s a good streak.  Time to start another one.

I believe in studying your own games, and this is especially true of your losses.  As such, this analysis is more directed at me personally than as for educational or entertainment value.  I need to learn from this game, and so the content is presented differently.  Mostly the information is concentrated in the critical positions, where I’ve made mistakes, and so there is less general analysis in other places.

If you dare to look anyway, though, you’ll see a smashing game from my opponent: he bravely sacrifices material for a devastating attack, and he then converts after a tough endgame.  A lovely game, one I wish I could have played as Black.  Alas, I was the victim, but I learned a great deal.  Let’s take a look.

[Event “2- Chess is for fun – Round 5”]
[Site “”]
[Date “2017.03.18”]
[Round “?”]
[White “SmithyQ”]
[Black “themis_neo”]
[Result “0-1”]
[ECO “D37”]
[WhiteElo “2135”]
[BlackElo “1855”]
[Annotator “Pettit”]
[PlyCount “100”]
[EventDate “2017.??.??”]
[TimeControl “1”]

{A fantastic game from my opponent, who sacrifices a pawn for a vicious attack.
I defend until the endgame, but his extra material wins out.} 1. d4 Nf6 2. c4
e6 3. Nf3 d5 4. Nc3 {So far, so standard.} Be7 5. Bf4 {This Bishop can go to
g5, f4 or b2 in most cases. Bg5 is the most classical, but if it has a
downside it increases the likelihood of piece exchanges, easing Black’s
position. b2 would be an excellent square, though it costs extra time to play
b3 first. Bf4 is then my preferred choice, as it avoids exchanges and doesn’t
cost time. It also pressures c7 is Black is careless. On the downside, it lets
Black play dxc4 and Nd5 in the middlegame, hitting the Bishop on f4. Well,
it’s chess. Every move has a potential drawback if you look hard enough …
except maybe mate.} c5 {My opponent strikes out at my centre. A good move, and
certainly more aggressive than c6.} 6. dxc5 {This breaks the tension and
allows the Bishop to come to a good square, but the Bishop has already moved
once to e7, so Black does not gain any time.} (6. e3 {This normal moves allows
Black to get a better structure.} cxd4 7. exd4 {[%cal Gd5c4] At some point,
maybe even now, Black can play dxc4 and leave White with an isolated pawn.
That’s not losing or anything, but you don’t always want to have an isolated
pawn.}) 6… Bxc5 7. e3 O-O 8. a3 {Strictly speaking, this move may not be
necessary right now, but it serves a useful role in this set-up. I plan on
playing Qc2 and possible Bd3, and I want to avoid Nc6-b4. I also may wish to
play b4. Finally, if I can delay developing by f1-Bishop, perhaps Black will
play dxc4 and I can play Bxc4 in one move. This is my thought-process at least.
} a6 9. Qc2 Nc6 {We’ve reached a normal 1.d4 position, more or less. We both
have three minor pieces developed, we’ve both advanced c-pawns. He has castled,
whereas I have moved my Queen. There are a few plans here. One is to play Bd3
and 0-0, normal development reaching a normal game. Another option is to play
cxd5, with Rd1 coming to eye Black’s Queen and possibly give Black some
structural weaknesses. The third option is:} 10. O-O-O {My play makes this
move look worse than I think it actually is. I still bring my Rook to d1, but
now I’ve castled and have time to launch a pawnstorm against Black. I was
feeling very aggressive and wantted to attack, so that helped push me in this
direction. I’d definitely play this again, even though I got rolled in this
game.} b5 $5 {My opponent reacts with a perfect blitz move: sac a pawn to open
lines. I refuse to believe this is correct, though, and I’m convinced that
perfect play refutes this. In a practical game, though, it certainly gives
good chances. Well, let’s see if you can play better than I. What would you do
here?} 11. cxb5 {I begin with the strategically dubious idea of capturing both
a- and b-pawns. This wins material but opens all sorts of dangerous lines
against my King. Honestly, I only did this because of a miscalculation on move
12.} (11. cxd5 {This capture makes far more sense, keeping lines somewhat
closed.} exd5 12. Kb1 {An important prophylaxis, which I missed during the
game.} (12. Nxd5 {I jumped straight into this line during my calculation,
which wins a pawn but leads to a pretty terrible position.} Nxd5 13. Qxc5 Be6 {
[%cal Ga8c8,Rc1b1,Re6f5,Rf5b1,Gc8c1,Gc6b4] Black will transfer the Rook to the
c-file and aim everything at my King. It’s hard to run with Kb1 because Bf5+
will always come with tempo. All of Black’s minor pieces are near my King, and
none of my own can easily help. White may well be losing here.}) (12. Bg5 {
This is a suggested Stockfish improvement, but it apparently leads to a draw
by repetition with best play.} Be6 13. Bxf6 gxf6 14. e4 d4 15. Nd5 Rc8 16. Kb1
Ne5 17. Qb3 Nc6 18. Qc2 {and repeats, so says the almighty Fish.}) 12… b4 {
Black aims to open lines against my King.} 13. Nxd5 Nxd5 14. Bd3 $1 {The only
move, says the computer. White finishes development before recapturing a piece.
There’s no way to hang onto material for Black, and he needs to becareful, for
moving the Nd5 allows Bxh7+ and wins the Queen.} bxa3 15. Bxh7+ Kh8 16. Qxc5 {
The position is crazy, but White is not worst. Both sides can throw the game
in a heartbeat, but this is the single best line for White you’ll see in this
entire game. Interestingly, I spent several minutes wondering what the best
reply to} axb2 {was, before I suddenly realized} 17. Rxd5 Qe8 18. Rh5 {wins
both material and a mating attack. Opps.}) 11… axb5 12. Nxb5 Qe7 {I won’t
lie, I overlooked this move.} ({It’s amazing that I can calculate all the
above variations accurately, even getting similar evaluations as the computer,
and yet here I thought the only move for Black was} 12… Qb6 $2 13. Bc7 {
and I win a piece. Chess blindness is funny.}) 13. Ne5 $2 {Annoyed that I
missed the last move, I now play too quickly here. I assumed that moving my
Knight to a good central square had to be good, and if Black exchanges then my
Bishop comes to a good diagonal that should protect my King. I spent maybe two
minutes on this move. I needed to spend far more.} (13. Kb1 {I need to think
prophylaxis above everything else. Even though the c-file is currently clogged
with pieces, that’s Black’s main trump. If he gets a Rook there, my King and
Queen are in trouble. So, move the King.} Bd7 14. Rc1 {Now if Black just
retreats then White can pressure for some advantage, maybe make that extra
pawn count. That’s why Black needs to move forward.} Bxa3 15. bxa3 e5 $1 {
Both opens lines for the light-square Bishop and, at least temporarily, stops
a Knight from coming to d4.} 16. Nxe5 Nxe5 17. Qc5 {White does best to
exchange Queens.} Qxc5 18. Rxc5 Ne4 $1 {Leads to a forced sequence.} 19. Rxd5
Bxb5 20. Bxb5 Nc3+ 21. Kb2 Nxd5 22. Bxe5 {Material is technically equal, as
White has two pawns for the Exchange. Two Bishops on an open board gives White
the sunny side of this endgame, though Black’s Rooks should hold without much
effort.}) (13. Nfd4 {This is the other move I briefly considered, and it’s
better than Ne5.} Bd7 {Simple development, preparing to play Rfc8 as soon as
possible.} ({Here, exchanging like in the game is a trap.} 13… Nxd4 $6 14.
exd4 {Now the Bishop must move.} Bb6 (14… Bxa3 {This sacrifice is the better
chance, but the pressence of the Knight on b5 helps White hold.} 15. bxa3 {
[%cal Gb5a3]} Ne4 (15… Ba6 {Trying to remove the defending Knight is too
slow.} 16. Bd6 Qb7 17. Bxf8 Bxb5 18. Bxb5 {[%cal Ga3a4,Gf8c5] and now,
depending which piece Black recaptures, White can play a4 or Bc5 and block
some of the open lines. With only one Rook, it’s harder for Black to take
advantage. White is winning here.}) 16. f3 Bd7 {Black prepares to sacrifice
his minor pieces to get White’s Queen.} 17. fxe4 Bxb5 18. Bxb5 Rfc8 19. Qxc8+
Rxc8+ 20. Kb2 $16 {White has two Rooks for the Queen, two Bishops for the Rook
and a passed a-pawn. Again, White has the better chances in the ensuing
endgame.}) 15. Bd6 Qd8 16. Bxf8 {I’ve both won more material and also pushed
Black’s Bishop off its great diagonal, making it far harder for Black to
sacrifice to open my King. White is much better.}) 14. Nb3 Bb4 $5 {Rather than
retreating off the diagonal, Black offers a dangerous piece sac.} 15. Kb1 (15.
axb4 {Accepting the sac is very risky, if not fatal.} Ne4 {This plays first,
stopping any Bg5 and then Bxf6 ideas.} 16. Kb1 Nxb4 {[%cal Gb4c2,Gb4d3,Re4d2,
Re4c3,Re4c5] The White Queen has no moves on her half of the board.} 17. Qc7 (
17. Qe2 e5 18. Bg3 Bf5 {and there’s no defence against the coming discovered
check.}) 17… Rfc8 18. Qb6 e5 19. Bg3 Bf5 {and the discovered check will soon
claim White’s King … or at least most of its army.}) 15… e5 {Black has the
initiative, and I have doubts White hangs on in a practical game.}) 13… Nxe5
14. Bxe5 {My Bishop can retreat to c3, covering the c-file while also aiming
at the Black King. I assumed that would be enough.} Ba6 $1 {I had assumed this
was impossible because of Nc7. Now I had to look deeply to try to save my
bacon.} 15. Kb1 {I finally clue in that the c-file is dangerous and move my
King.} ({First, the best defence is} 15. Bxf6 gxf6 16. Nc3 Bxf1 17. Rhxf1 Rfc8
18. Rd3 $15 {and though Black is definitely better, White isn’t losing, either.
The fewer pieces, the fewer chances for Black to attack.}) ({Second, let’s
look at the move I thought refuted Ba6.} 15. Nc7 Bxf1 16. Nxa8 Bc4 $1 {Even
better than taking on g2, which really just opens lines for White. Now White
needs to spend a tempo to save his Knight.} 17. Nc7 Ne4 {[%cal Gf8b8,Gc5a3,
Gc4b3] Every Black piece is aimed at White’s vulnerable King. Black’s attack
plays itself. He has full compensation is likely even winning. I wasn’t going
to calculate this any further, as I’d much rather be Black!}) 15… Rfc8 {
Notice how Black just makes natural moves? He puts his Bishops on open lines,
he connects his Rooks and now he places them on open lines. White has a much
harder time finding such moves. This tells us that Black has the better
position.} 16. Bc3 $6 {I try to block the c-file, but really, all this
succeeds in doing is pinning this Bishop to my Queen after Black moves his Bc5.
} (16. Nc7 $5 {My position isn’t great, but this is relatively best, blocking
the c-file for a moment and exchanging a pair of pieces.} Bd6 $1 {[%cal Gc8c2]
And really, this difficult to see move is key for Black, as it exploits the
pin on the c-file.} 17. Bxd6 Qxd6 18. Rc1 Bxf1 19. Rhxf1 Rab8 {This position
is not easy for White, but it’s better than the game.}) 16… Bxb5 17. Bxb5
Bxa3 $1 {Boom! A beautiful sacrifice to rip open my Kingside.} 18. bxa3 {
If I’m going to lose, I might as well lose in the most artistic way possible.}
(18. f3 {The computer suggests this as an option, but Black is still +2 at
least in all lines.}) 18… Rxa3 $6 ({Taking with the Queen would end the game
much faster.} 18… Qxa3 19. Rd3 Rxc3 20. Qxc3 Qa2+ 21. Kc1 Ne4 22. Qc6 (22.
Rd2 {This clever move may give Black pause, though he is still winning after}
Nxd2 23. Qxd2 Qa1+ 24. Kc2 Qxh1 $19 {Not only is Black up the exchange, but
he’ll gobble most of White’s pawns as well.}) (22. Qb2 Rc8+ 23. Kd1 Qxb2) 22…
Qa1+ 23. Kc2 Ra2+ 24. Kb3 Qb2#) 19. Rd3 {In my defence, I find the best
defence given the chance.} d4 20. Bb4 $1 {This move, at first glance odd, is
the only chance. Otherwise Black takes on c3 and summarily executes my monarch.
Here I at least get Queens off.} (20. exd4 Nd5 {is of course not what we want.}
) 20… Qxb4+ 21. Qb2 Qxb2+ 22. Kxb2 Rxd3 23. Bxd3 {After all the smoek clears,
I’m only down a pawn, and I may even be okay …} Ng4 $1 {… if not for this
move. Amazingly, I can’t defend the simple threat to win a pawn.} 24. Rf1 dxe3
25. fxe3 Nxe3 (25… Nxh2 {At first, I thought this move must be better, as it
cripples my structure. In fact, it just gives me play!} 26. Rh1 Ng4 27. Bxh7+
Kf8 28. Rh3 {I stopped calculating here, as I assumed Black could play the
standard Bishop trapping manuever and win easily … but things are trickier!}
g6 $2 29. Bxg6 fxg6 30. Rh8+ Ke7 31. Rxc8 {Suddenly Black is struggling to
draw.}) 26. Rg1 f5 27. g3 Kf8 $2 {Black, up two pawns, now makes a careless
move. Simply Kf7 and I’m close to resigning.} 28. Re1 Nc4+ 29. Bxc4 Rxc4 30.
Rxe6 {We’ve reached a Rook and pawn endgame … which I’m fairly certain is
drawn with best play. I don’t know endgames well enough to say for sure, but
I’ve played around for over an hour with my engine, and though it says Black
has a huge advantage, it can’t show a forced win. I learned a lot about Rook
endgames studying the following twenty moves.} Kf7 31. Re3 (31. Rb6 {[%csl Rc6,
Rd6,Re6,Rf6,Rg6,Rh6] To start, this move is a slight improvement, making it
harder for Black to advance his King. You might think, wait, can’t Black win a
pawn?} Re4 32. Kc3 Re2 33. h4 Rg2 34. Kd4 Rxg3 {The short answer is yes, Black
can.} 35. Ke5 g6 36. Rb7+ {And it then claims this position is a draw. It’s
amazing, but try to find a way to win for Black. Heck, just try to get off the
back rank. Play around with an engine and see for yourself.}) 31… Kf6 32. Kb3
Re4 33. Rf3 {I of course cannot exchange Rooks. The main idea is to trade
pawns and use my Rook to harass him with checks.} h5 34. Kc3 {Also, I need to
get my King over to the Kingside.} g5 35. Kd3 (35. Kd2 {[%cal Gf3a3,Ga3a6]
Here’s another tip: don’t let your King get in the way of your Rook. My pieces
kept tripping over themselves, and I only have two!}) 35… Ke5 {[%cal Yd3d2,
Gf3a3,Ga3a5] We see here that I have no useful move. Had my King been on d2, I
could have played Ra3 and then threatened to chase him away with checks.} 36.
Rf2 h4 37. gxh4 {It’s hard to tell if this move is a mistake or not. On the
one hand, the engine thinks Black is winning regardless of what I do … but I
can’t get a forced win. In fact, I get the opposite.} Rxh4 38. Ke2 (38. Rg2 {
This move forces a pawn move, which makes it much harder for Black to advance.}
g4 39. Ke2 Rh3 40. Ke1 Rf3 41. Ra2 f4 42. Rb2 {I’ve put this position into two
engines (Stockfish and Komodo) and had them play games against themselves.
Four games, all ending in draws after the 50 move rule. With correct play,
Black can’t seem to win!}) 38… Ke4 39. Rg2 g4 40. Kf1 {If my King occupies
the Queening squares, then Black needs to work much harder to win. This I know
from basic endgame theory.} Rh3 41. Ra2 Rf3+ 42. Kg2 Rd3 43. Kf2 {I just need
to wait.} f4 44. Ra4+ Rd4 45. Ra3 Rd2+ 46. Kg1 f3 {Black had made progress,
but interesting, even here the computer can’t force a win. All White needs to
do is keep his Rook on the back rank and/or check the Black King all over the
place.} 47. Rb3 $4 {I unfortunately allow Black to use his only winning idea.
I could have simply occupied the back rank with Ra1 instead.} (47. Ra1 {
and the question becomes how to make progress.} Ke3 48. Ra3+ Ke2 {This is just
one possible line.} 49. Ra1 Rd1+ (49… f2+ 50. Kg2 {And how does Black Queen
his pawn?} Ke3 51. Ra3+ Ke2 52. Ra1 Rd3 53. Ra2+ Ke3 54. Rxf2) 50. Rxd1 Kxd1
51. Kf2 Kd2 52. h4 $1 gxh3 53. Kxf3 Ke1 54. Kg3 h2 55. Kxh2 $11 {Once more,
the drawing idea is to OCCUPY THE BACK RANK and then CHECK THE KING whenever
necessary.}) 47… Rd1+ 48. Kf2 Rh1 {Losing the pawn means losing the game.}
49. h4 {I tried a desperate trick, but I’m of course lost.} (49. Kg3 {The
point is that the natural move fails tactically.} Rg1+ 50. Kf2 (50. Kh4 f2 {
and Queens.}) 50… Rg2+ 51. Kf1 Rxh2 $19) 49… Rh2+ {Black finds the
cleanest move. Kudos to him.} (49… Rxh4 50. Kg3 {and I have some small
swindle chances, such as} Rh3+ 51. Kxg4 f2 52. Rb4+ Ke3 53. Rb3+ Ke4 (53… Ke2
54. Rb2+ Ke1 55. Rb1+ {There is no escape from the checks.} Ke2 56. Rb2+ Kf1
57. Kxh3 Kg1 58. Rxf2 Kxf2) 54. Rb4+ Kd3 {and if Black tries to get closer to
the Rook, then} 55. Rf4 {and the pawn is stopped. I’ve pulled some great
swindles in my time, so this was worth a punt.}) 50. Kf1 g3 {And I can’t stop
the pawns. Incredibly impressive play by my opponent, pulling off a fantastic
attack … but I should have drawn this game. If I knew the drawing idea, I do
it easily. In fact, I had some inkling of it, but it wasn’t enough to guide me
in the right direction. On the plus side, by studying this game, I’ve gotten
to practice this ending, and I’ll be ready for my next game.} 0-1


Let’s list my main mistakes.

  • First, I grabbed the wrong pawn.  Why open lines against your own King?  Taking the d-pawn would have opened lines for my Rook against his Queen, a strategically much better idea.
  • Second, I made a strange miscalculating, completely missing Qe7, a natural move.
  • Third, I rushed with Ne5, not bothering to calculate, and this right after missing the Qe7 idea literally one move earlier.  IF YOU MISS A MOVE IN YOUR CALCULATION, YOU NEED TO REDO ALL YOUR CALCULATIONS AGAIN.  And by you, I of course mean me.
  • Forth, I thought of prophylaxis far too late.  Kb1 is a common move in Queenside castling, and I never considered it until it was too late.
  • Fifth, though I correctly identified the c-file as my main problem, I did not address it in the right away.
  • Finally, though I defended tenaciously until the end, I misplayed the endgame quite badly.

These are the things I need to work on.  Most of these are thought-process based.  That is, I need to make sure I see all candidate moves, that I restart my calculation when necessary, that I think of prophylaxis earlier (or at all…), those types of things.  Only the last two points relate to chess itself, namely defending properly and endgames, specifically theoretical endgames.  This makes sense, as these are the two areas I have the least experience in, and I’ll be sure to study this in the future.

Losing isn’t fun, but I certainly learned from it, and I can take some satisfaction in seeing that I (seemed) to reach a drawn endgame.  I mean, I didn’t draw, but I could have.  That’s certainly better than the alternative, and it shows how hard it is to win a won game.

And, of course, kudos to my opponent for a beautiful game.  Congratulations.  No one will ever accuse me of being a sore loser … even if you were several hundred rating points lower than me.  You have to win more games and raise your rating, Themis_Neo, so I don’t as many rating points next time!

Leave a Reply

Your email address will not be published.