Recent Endgames

Wed, 2014-08-27 10:59 -- IM Max Illingworth
[Event "TCh-TUR Super League 2014"]
[Site "Kocaeli TUR"]
[Date "2014.08.21"]
[Round "4.1"]
[White "Ding Liren"]
[Black "Mamedyarov, S."]
[Result "1/2-1/2"]
[ECO "D52"]
[WhiteElo "2742"]
[BlackElo "2743"]
[PlyCount "108"]
[EventDate "2014.08.18"]
[EventRounds "13"]
[EventCountry "TUR"]
[Source "Mark Crowther"]
[SourceDate "2014.08.25"]
[WhiteTeam "T.S. Alyans Satranc Spor Kulubu"]
[BlackTeam "Besiktas Jimnastik Kulubu"]

{Apologies for the long delay in blogging! I've been quite busy with my
overseas tournaments (namely the Olympiad and a GM round robin in Kecskemet,
Hungary), but rather than reporting on those (which I might save for when I
return home), I want to look at some interesting recent endgames. My reason
for this choice of topic is that this is an area where juniors (who comprise
the majority of readers of this blog) struggle most, due to a lack of study of
and experience in them. Anyway, let's take a look at some of these 'practical'
endings. My definition for this article in that I count any position without
the queens on the board.} 1. d4 d5 2. c4 c6 3. Nf3 Nf6 4. Nc3 e6 5. Bg5 Nbd7 6.
e3 Qa5 7. Nd2 Bb4 8. Qc2 c5 9. Bf4 cxd4 10. exd4 O-O 11. Be2 Re8 12. O-O Bxc3
13. bxc3 e5 14. Be3 exd4 15. cxd4 Nb8 16. Nb3 Qc7 17. Rac1 Bg4 18. Bd3 dxc4 19.
Qxc4 Qxc4 20. Bxc4 {Looking at this position, you probably think that White is
quite a bit better. After all, he has the pair of bishops in a very open
position, a modest lead in development, and the d-pawn is passed and far from
easy for Black to attack. However, Mamedyarov manages to neutralise Ding's
initiative and it's very instructive to see how he does it.} Be6 $1 {First of
all, Black controls the square in front of the IQP so it can't safely advance
further.} (20... Nbd7 {was the obvious move, but after} 21. Na5 $1 {either
Black will be tied to the defence of the b7-pawn, or if the pawn advances,
White's knight will get a sweet outpost on c6 where it ties up Black's entire
position.} Nb6 22. Bb5 Re7 23. Bd2 {with the idea of Bb4 to kick the defending
rook on e7 must be a little better for White.}) 21. Bb5 {White preserves his
bishop pair and continues to harass Black's pieces.} (21. Bxe6 {would be met
with} fxe6 $1 {- again, it is more important to keep tabs on the d4-pawn than
it is to keep Black's pawn structure together. The e6-pawn is not easy to
attack (being on the opposite colour complex to White's bishop) and helps
anchor a knight on d5.}) 21... Rd8 22. Nc5 Bd5 {In general the knight is a
better blockader of a passed pawn, but this is fine too - now White needs to
find a concrete continuation as if Black gets time to play ... Nbd7-b6 (or
Nxc5) and ...Rac8, he will have completed his development and avoid any
problems. Unfortunately for Ding, he doesn't have any way to breach Black's
solid structure.} 23. Bf4 (23. a3 Nbd7 24. Nxd7 Nxd7 25. Rfe1 {is the simple
choice of the computer, but then the light-squared blockade won't ever be
broken down after say} Nb6 {.} (25... Nf8 {and ...Ne6 was also fine, to put
some pressure on the d4-pawn.})) 23... a6 24. Ba4 b5 25. Bb3 Nc6 {Black is
doing very well in this position, even though White didn't make any obvious
mistakes. White's bishop pair is about to be traded off and White's weak pawn
on d4 isn't running away.} 26. Be5 $1 {Ding realises that the situation could
easily turn against him, and therefore uses a typical defensive technique -
simplification.} Na5 (26... Bxb3 {would be the most consistent continuation,
as after} 27. axb3 (27. Nxb3 Nxe5 28. dxe5 Nd5 {gives us an opportunity to
understand why the queenside majority is stronger than White's kingside
majority - because the Q-side majority is a lot further away from the
respective kings. While a White passed pawn on the kingside can be stopped by
the king, White's pieces will become tied up preventing the promotion of a
passed pawn on the queenside.}) 27... Ne8 $1 {Black's knights are better than
the bishop and knight, because the knights have stable posts (e.g. on b4 and
d5) and the e5-bishop is a ghost as far as Black's pieces are concerned. Yes,
the c5-knight looks nice, but can be played around easily enough, and while
the a-pawn can be guarded with ...a5, the d4-pawn is pretty fixed, making it
more vulnerable to attack.} (27... Nxd4 28. Bxd4 Rxd4 {doesn't win a pawn
because of Black's exposed back rank:} 29. Nxa6 $1)) 27. Bxf6 (27. Bc7 Nxb3 28.
axb3 Rdc8 29. Ba5 {is preferred by the engine, but the presence of
opposite-coloured bishops without any penetration points for either side make
this a draw.}) 27... Nxb3 28. Bxd8 (28. axb3 gxf6 29. Ra1 {was perhaps more
accurate.}) 28... Nxc1 29. Rxc1 Rxd8 30. Nxa6 Bxa2 {You would think that there
is some way to force the exchange of the b5 and d4 pawns to get a drawn ending,
and indeed White can achieve that with a couple of precise moves.} 31. Rc5 Bc4
32. f3 (32. Nc7 Rxd4 33. f3 {actually regains the pawn as Nxb5 is threatened
even when Black guards the pawn with} Rd1+ 34. Kf2 Rd2+ 35. Kg1 Rb2 36. Nxb5 $1
{and Black's weak back rank is the problem.}) 32... Kf8 33. Nc7 Rb8 {Now Black
has some small chances based on his passed pawn being better than White's 
(further away from the kings), though White can always play Na6-b4 to blockade
it.} 34. Kf2 (34. Na6 Rb6 35. Nb4 {would be quite sensible too.}) 34... Ke7 35.
Ke3 Kd7 36. Kd2 $2 {Accurate calculation is important in these endgames and
White misses Mamedyarov's next trick to penetrate to the 2nd rank.} (36. Nd5
Re8+ 37. Kf4 {is the correct way to defend, to threaten the liquidating Nb6.
Probably White didn't like} Bxd5 38. Rxd5+ Kc6 39. Rc5+ Kb6 {which does make
the passed b-pawn look scary, but by repositioning the rook with} 40. Rc2 $1 {
(stopping the nettlesome ...Re2 too)} b4 41. d5 f6 (41... Kb5 {allows the rook
behind the enemy passed pawn with} 42. d6 {followed by Rc7-b7, and if Black
stops that with} Rb8 {,} 43. d7 {threatens Rc8 and forces} Rd8 44. Rc7 {and
Black can't shepherd his b-pawn to promotion.}) 42. Rd2 {(rooks belong behind
passed pawns!) White is able to hold the draw.}) 36... Rc8 37. Na6 Re8 $1 {Now
White lacks the Nd5 resource.} 38. Kc3 {White activates his king in search of
counterplay. In such a sharp position you can't just defend passively.} (38.
Re5 Ra8 39. Nb4 Ra4 40. Kc3 Ra3+ 41. Kd2 Rb3 {for instance would break White's
blockade and give Black quite good winning chances after} 42. Nc2 Kd6 {as his
pieces are much more active.}) (38. Nc7 $142 Re2+ 39. Kc3 Rxg2 40. Nxb5 Bxb5
41. Rxb5 Rxh2 {seems winning for Black at first, because his passed h-pawn is
a long way from White's king and the d-pawn can be contained, but the engine
demonstrates that White should draw with} 42. Rb7+ Ke6 43. Rb6+ Kf5 44. Rb5+
Kg6 ({If} 44... Kf4 45. Rb7 f5 46. Rxg7 Kxf3 47. d5 {and again the passed
d-pawn provides enough counterplay to draw.}) 45. Rb6+ f6 46. d5 {as Black had
to seriously misplace his king to escape the checks.}) 38... Re3+ 39. Kb4 (39.
Kb2 Re2+ 40. Kc3 Rxg2 {would be winning for Black as after} 41. Nc7 Rxh2 42.
Nxb5 Bxb5 43. Rxb5 Rf2 $1 {White can't hang to the f3-pawn, and the three
connected passers will be decisive.}) 39... h6 $2 {The players were probably
very short of time, so it wasn't easy to play accurately, but the position is
that sharp that losing a tempo like this is enough to turn the position from a
win to a draw.} (39... Rb3+ 40. Ka5 Rb2 {is correct - we can understand from
Black's game choice that he didn't want to allow} 41. Rg5 {, but after} g6 {
the rook is just passive on g5, and even better is} (41... Ra2+ 42. Kb6 Rxa6+
$1 43. Kxa6 b4+ 44. Kb7 b3 {, when the rook doesn't get back in time to stop
the passed b-pawn:} 45. Rc5 Bd3 46. Rd5+ Ke6 47. Kc6 b2 {and wins.})) 40. Nc7
$1 {White is on red alert and saves the half-point by eliminating Black's
trump card on b5.} Rb3+ 41. Ka5 Rb2 42. g4 $1 {Good technique - White wants to
keep his pawn structure as united as possible.} (42. Nxb5 $2 Bxb5 43. Rxb5 Rxg2
{would not be so simple as White's king is a long way from the action on the
kingside, and will likely lose a pawn there given that all his pawns are
isolated.}) 42... Rd2 43. Nxb5 Bxb5 44. Kxb5 Rxd4 {You have to be careful when
your king is cut off from the main battlefield like it is here - but White has
everything worked out.} 45. Rc4 $1 {You should double-check your calculations
when you could potentially enter (or choose to enter) a pawn endgame, but it's
quite easy to calculate that the pawn ending will be a draw.} Rd2 (45... Rxc4
46. Kxc4 Ke6 47. Kd4 {is just a draw as neither king has any way to breach the
enemy fort.} Kf6 48. h4) 46. h4 Rd5+ 47. Kb4 f5 48. Kc3 {The White king has
come around and can continue to cross the board with Rd4, Kd3, etc. So the
draw is obvious.} fxg4 49. Rxg4 g5 50. hxg5 hxg5 51. f4 gxf4 52. Rxf4 Kd6 53.
Rd4 Rxd4 54. Kxd4 Kc6 {The main principles we learned from this game are: 1)
When fighting against an IQP or passed pawn (in our example, both), the first
priority should be to prevent its safe advance. 2) We don't have to worry
about having an isolated pawn if there is no feasible way for the opponent to
attack it. 3) Not all outposts are equal - Black was able to play around the
c5-knight which didn't do that much, while Black's outpost on d5 stopped White
from having any entry points into Black's position, including while he was
catching up in development. 4) Throughout the game we saw how Black's
queenside majority/passed pawn was more important than White's passed d-pawn,
because it was further away from the kings - as it ties up the opponent's
pieces more. Keep this idea in mind in your own endgames. 5) The bishop is
better than the knight when there are passed pawns for both sides, especially
with pawns on both wings. 6) White's bishop pair was of little use early on
because it couldn't find any targets, Black's minor pieces had stable posts 
(especially d5) and the bishop pair was soon exchanged in any case. 7) When
the endgame becomes sharp, as it did from around move 33, it's not enough to
play on general considerations - you have to calculate precisely and be on the
alert for tricks. When White overlooked the nice ...Rc8-e8! trick to make
White lose a tempo with his knight, things proved quite difficult. In turn,
Black played too slowly at the critical moment (move 39) when the direct
continuation (going after White's weak pawns) was the way to win. 8) Finally,
be careful when your king is cut off from the main battlefield (especially
with pawns on one side of the board). In this case White brought his king
across by offering rook exchanges into a drawn king and pawn endgame. That's a
lot of ideas for one side, but if you study similar examples with this
material balance (or imbalance if you prefer) you will see these themes come
up again. And I highly recommend you play through and analyse more games with
R+B vs. R+N or bishop pair vs. superior pawn structure situations, as they
occur quite frequently in practice.} 1/2-1/2 
[Event "89th ch-FRA 2014"]
[Site "Nimes FRA"]
[Date "2014.08.23"]
[Round "6.5"]
[White "Gharamian, T."]
[Black "Fressinet, L."]
[Result "0-1"]
[ECO "C65"]
[WhiteElo "2657"]
[BlackElo "2708"]
[PlyCount "106"]
[EventDate "2014.08.17"]
[EventRounds "11"]
[EventCountry "FRA"]
[EventCategory "14"]
[Source "Mark Crowther"]
[SourceDate "2014.08.25"]

{Meanwhile, Fressinet has been on fire in the recent French Championship, and
at the time of writing is leading by a significant margin with his 6.5/8.
Let's learn from his technique in a superior knight endgame against another
strong GM.} 1. e4 e5 2. Nf3 Nc6 3. Bb5 Nf6 4. d4 exd4 5. O-O a6 6. Ba4 Be7 7.
Qe2 O-O 8. e5 Nd5 9. Rd1 Nb6 10. Bxc6 dxc6 11. Rxd4 Qe8 12. Nc3 c5 13. Rd1 Qc6
14. Bg5 Bxg5 15. Nxg5 Bf5 16. g4 Qg6 17. f4 Bxc2 18. Rd2 h6 19. Rxc2 hxg5 20.
f5 Qh6 21. Ne4 Nd5 22. Rxc5 Nf4 23. Qe3 Rad8 24. Nf2 Rd7 25. Rac1 c6 26. Re1
Rfd8 27. e6 fxe6 28. fxe6 Re7 29. Re5 Rd6 30. Qc5 Rexe6 31. R1e3 Qg6 32. Rf5
Qe8 33. Rfe5 Rxe5 34. Rxe5 Re6 35. Qc4 b5 36. Rxe6 Qxe6 37. Qxe6+ Nxe6 {
Botvinnik once said that knight endgames were very similar to pawn endgames,
as you can't lose a move with the knight (if you aren't sure why this is
important, think of King and Pawn vs. King zugzwang positions). Well, the pawn
endgame would be an easy win for Black (the simplest winning plan would be,
after activating the King, to create a passed pawn on the queenside, and while
White's king is capturing it, take the White kingside pawns and queen the
g-pawn). So it seems fair to say that the extra pawn should also give Black
the win in the knight endgame.} 38. Ne4 c5 {OK, the e6-knight is tied up to
the weak g5-pawn for now, so things aren't that easy.} 39. Kf2 (39. a4 $5 {
might seem weird, but the idea is to exchange pawns to increase the drawing
chances. The problem is that after} bxa4 {White will waste time grabbing the
a4-pawn later, and Black also has ...a3 at some point to create a passed
c-pawn. After} 40. Nc3 Kf7 41. Nxa4 Ke7 42. Kf2 Kd6 43. Nc3 c4 {Black is ready
to penetrate with either his king (...Kc5-b4-b3) or knight (...Nc5-d3) and
that should give him the win. In fact,} 44. Ke3 Ke5 45. Ne4 a5 46. Nd2 Kd5 47.
Ne4 {gives Black the opportunity to free his knight with} Nc5 $1 {as following}
48. Nxg5 Na4 49. Ne4 Nxb2 50. Nc3+ Kc5 {the two queenside passers will decide
the game.}) 39... Kf7 40. Ke3 Ke7 41. a4 {So White tried my idea from earlier
after all. Fressinet doesn't even wince though.} (41. b3 Kd7 42. Kd3 Kc6 43.
Ke3 Kd5 44. Nc3+ Kd6 $1 45. Ne4+ (45. Ke4 {is better, but Black should still
win after} Nd4 {followed by ...g6, moving the d4-knight and pushing with ...c4.
}) 45... Ke5 {would be how Black would otherwise play. Notice how Black
triangulated with his king with ...Kd5-d6-e5 so that White has to worsen the
position of his king or knight here (unless he moves a pawn, but that weakens
his position too, and in any case he doesn't have an infinite number of pawn
moves). If he moves his king, the Black king can penetrate to the 4th rank,
while if he moves the knight, Black's knight can take up a more aggressive
stance.}) 41... Kd7 $1 42. axb5 axb5 {Black's winning plan is not very
different to what I explained in the note to 41.b3.} 43. Nd2 Kd6 44. Ke4 {
White tries cutting Black off with his king, but Black can still create a
passed pawn on the queenside (a key part of the winning plan).} c4 45. Nf3 {
White goes for counterplay as passively waiting with} (45. h3 Kc5 46. Ke3 Kb4 {
followed by ...Nf4 doesn't help. White's d2-knight is in a sort of zugzwang as
he can't allow ...Kb3xb2, but otherwise White is completely bound up.}) 45...
b4 46. h4 gxh4 47. Nxh4 b3 $1 {More precise than the immediate ...c3 as after}
(47... c3 48. bxc3 Nc5+ 49. Ke3 b3 50. Kd2 {White is in the square of the pawn.
}) 48. Nf3 c3 $1 {As I mentioned in the previous game, you need to see the
little tricks too to be able to cap off the win.} 49. Kd3 (49. bxc3 Ng5+ $1 50.
Nxg5 b2 {is the key point - there is no more Nd2 to stop the pawn. Every
Russian schoolboy knows this device for winning knight endgames and you should
rehearse it too.}) 49... c2 {With the passed pawn protected as well, the rest
is easy.} 50. Kd2 Kd5 51. Ne1 Nf4 52. Nxc2 bxc2 53. Kxc2 Ng6 0-1 
[Event "21st Abu Dhabi Masters 2014"]
[Site "Abu Dhabi UAE"]
[Date "2014.08.23"]
[Round "3.1"]
[White "Sumets, A."]
[Black "Kryvoruchko, Y."]
[Result "0-1"]
[ECO "D45"]
[WhiteElo "2622"]
[BlackElo "2708"]
[PlyCount "72"]
[EventDate "2014.08.20"]
[EventRounds "9"]
[EventCountry "UAE"]
[Source "Mark Crowther"]
[SourceDate "2014.08.25"]

{Finally, a fun example to show how a bishop and two connected passed pawns 
(not necessarily extra pawns) can beat a rook. Club players tend to be afraid
to sacrifice material because material is a lot easier to prove and maintain
than an initiative or positional compensation, but it will be a big step for
your chess development if you can muster the courage to sacrifice when your
analysis or intuition indicate it's the best continuation.} 1. d4 Nf6 2. c4 e6
3. Nf3 d5 4. Nc3 c6 5. e3 a6 6. a3 Nbd7 7. Qc2 g6 8. e4 dxe4 9. Nxe4 Nxe4 10.
Qxe4 Bg7 11. Be2 O-O 12. Bg5 Qe8 13. Bf4 c5 14. Bd6 f5 15. Qc2 cxd4 16. O-O e5
17. c5 Kh8 18. c6 bxc6 19. Qxc6 Rb8 20. Bxb8 Nxb8 {Black has very nice
compensation for the exchange in a pawn, a powerful centre (which deprives
White's pieces of good squares) and the bishop pair, but let's focus on the
arising sharp endgame.} 21. Qc7 Bd7 22. Bc4 e4 23. Ng5 h6 24. Nf7+ Kh7 25. f4
Qc8 26. Qxc8 Rxc8 {Black's advantage has sharply increased from the position
after the exchange sacrifice, as Black now has connected passers and he has
much better piece placement (the f7-knight is a bit offside).} 27. b3 {The
position is very complicated so we shouldn't chastise the players for their
mistakes.} d3 {Not bad, but I quite fancy the engine's} (27... Nc6 $1 28. Bxa6
Rb8 29. b4 d3 {, giving up a pawn but taking a very powerful initiative. True,
it's not so easy to be courageous like this when there are other good options
available.}) 28. Ne5 {Blocking Black's bishop pair is the only way to try and
survive.} Bxe5 $2 {Black wants to clarify the position, and this does grab a
second pawn for the exchange, but it also gives White time to blockade the
passed pawns and coordinate his pieces, which earlier didn't have any harmony.}
(28... Bb5 $1 29. Bxb5 axb5 {was very strong, as White's rooks are very
ineffective defenders against the connected passers (indeed, if Black could
play ...e3, all the rooks in the world would not save White). If White tries
to get active with} 30. Rac1 {,} Rxc1 31. Rxc1 d2 32. Rb1 g5 33. g3 Bxe5 34.
fxe5 e3 35. Kf1 Nc6 36. Ke2 f4 {gives Black a beautiful and crushing chain of
pawns - Black will simply walk his king up to take on e5 and then all will be
ready for the finale.}) 29. fxe5 Re8 30. Kf2 f4 $6 {Very tempting, for sure,
but this actually makes White's task easier by opening lines for his rooks.} (
30... Nc6 31. Bxa6 Nxe5 {was again better, activating the knight so that it
may support the passed pawns. Noteworthily, it's better to have two
well-supported passers than three blockaded ones.}) 31. g3 g5 32. gxf4 gxf4 33.
Ke1 f3 34. Kd2 {I think both players were very short of time here without an
increment to help them.} Bc6 $2 (34... Rxe5 {was more obvious and also defends
e4, so I don't understand why this wasn't preferred.}) 35. e6 $1 {Suddenly
White's one passer is better than Black's three!} Kg6 36. Ke3 Kf6 {The game
curiously ended here, just as White was getting a winning position - either
White lost on time or there are lots of moves missing. Anyway, I hope that
game has opened your eyes to a new dimension of chess (the dynamic element)
and shown that you don't have to fear positions where you have strong dynamic
and/or static compensation for your material deficit.} 0-1 
[Event "?"]
[Site "?"]
[Date "2014.08.26"]
[Round "?"]
[White "Capablanca Study"]
[Black "Without the Solution"]
[Result "*"]
[SetUp "1"]
[FEN "8/7r/2p2pp1/2p3p1/k5p1/3KN1p1/8/1R6 w - - 0 1"]
[PlyCount "0"]
[EventDate "2014.??.??"]

{Finally, something a little bit different - I saw this study by Capablanca
earlier today and can't work out the solution! One thing to keep in mind with
studies is that in a good study, every piece is placed where it is for a
reason, so we can reason that every pawn stops Black from playing some
defensive resource against our attempt to checkmate Black's king. I won't give
the solution to this as I want to work it out for myself, but instead will set
a different puzzle: what checkmate do you think White is trying to set up from
this position? You'll know it once you've figured it out! I will see you in
the next blog post!}  *