|
ALMOST EVERYONE SUCKS AT THE ENDGAME
Rugsy (2072) vs. neilbridge (2078)
ICC, 2007

WHITE TO MOVE
This position should be drawn. But keep in mind that just about
everyone sucks at the endgame, and the blunders that these two Experts
uncork should let lower rated players lose all respect for their higher
rated brethren.
In other words, when you reach an endgame, don't fear anyone -- they
are more than likely to hack the position to bits. This means that if
you learn to play the endgame in a competent fashion, you'll get an
enormous boost in both results and rating.
1.Ne2
White is about to play Nc3 and take aim at both b5 and d5. Black reacts in classic fashion: he panics!
1...b4??
Black feels the heat, but this move just makes it worse.
1...a4? 2.Nc3 b4 3.axb4 Nxb4+ 4.Kc4 is unpleasant for Black, but the
correct 1...Bd8! puts out the flames: 2.Bxd8 (2.Bc5 Be7) 2...Nxd8 3.Nc3
Nb7! (the saving idea)
NICE!
4.b4 (Bailing out. Better for Black is 4.Nxb5 Nc5+ followed by 5...Nxe4.) 4...axb4 5.axb4, =.
2.a4
Fixing the edible target on a5 and intending to approach it with the
King via Kd3-c4-b5. Also possible was 2.axb4 axb4 3.Kc4 Nb8 4.f3 when
further King penetrations and/or Nc1-d3 will put Black under tremendous
pressure.
2...Kd6?!
This makes things a bit too easy for White. Like it or not, he had to
give 2...Bd8 a try. Of course, this doesn't mean that Black will be happy
with the resulting positions: 3.Bxd8 Nxd8 4.Kc4 f3 (Both 4...Kd6 5.Kb5
and 4...Nb7 5.f3 Kd6 6.Kb5 promise Black nothing but suffering) 5.Ng1 Nb7
6.Nxf3 Nd6+ 7.Kd3 Nb7!
DEAD BUT STILL FIGHTING
Though things are indeed grim, Black is still fighting and might
confuse his opponent. An example (which I can confidently say isn't
anywhere near best play): 8.Nxg5+ Kf6 9.Nf3 (9.Nh3!?) 9...Nc5+ 10.Kc4
Nxe4 11.g5+? (11.Kd5 seems to be very strong) 11...Nxg5 12.Nxg5 Kxg5
13.Kb5 Kf4 14.Kxa5 Kf3 15.Kxb4 Kxf2 16.a5 e4 17.a6 e3 18.a7 e2 19.a8=Q
e1=Q+ with a draw due to the absolutely horrible placement of White's
Queen.
3.Kc4 Bd8 4.Bxd8
Quite good, but I would prefer keeping Black bound a bit longer with
4.Bc5+ Ke6 5.Kb5 Nb8 6.Nc1 heading for either b3 or d3, depending on
what Black does.
4...Nxd8
WHITE SHOULD WIN
5.Nc1??
5.Kb5! was a dead win.
5...Nc6??
Now Black misses an obvious draw: 5...Kc6! 6.Nd3 (6.Nb3 Nb7 idea
7...Nd6+) 6...Nf7 7.f3 (And not 7.b3 Nf7 mate!) 7...Nd6+ 8.Kb3 Nf7, =.
6.Nb3
Black's busted. Any Knight move hangs the a-pawn, and any King move fails to address the problem of White's threatened Kb5.
Another way for White to get the victory (without the tricks he runs into with 6.Nb3) is 6.Kb5 f3 7.Nd3.
6...f3
A good try! Now White can win with both 7.Nc5 and 7.Kd3 followed by
8.Ke3 and 9.Kxf3 (note that this involves zero risk). Instead, he turns
an easy win into an agonizing defeat by stepping into a King and pawn
endgame and the complications that such an ending often brings.
7.Kb5??
White saw the upcoming King and pawn endgame and thought it was just
winning. And White was right, the King and pawn endgame IS winning, but
for Black!
7...Nd4+ 8.Nxd4
There's no going back: 8.Kc4 Nxb3 9.Kxb3 Kc5 10.Kc2 Kd4, 0-1.
8...exd4 9.Kc4 Ke5 10.Kd3 b3
BLACK WINS VIA OPPOSITION
GAG! Now Black has the opposition and White has to jettison the e-pawn.
11.Kc4 Kxe4 12.Kxb3 Kd3!
Icing the (lucky) victory. Instead, the hasty 12...d3?? loses: 13.Kc3 Kf4 14.Kxd3 Kxg4 15.b4 axb4 16.a5.
13.Ka3 Kc2, 0-1.
In my new book, SILMAN'S COMPLETE ENDGAME COURSE, I have a chapter in
the Master section titled, KING AND PAWN ENDGAMES ARE CONFUSING! I
was trying to beat home the point that King and pawn endgames can be
incredibly complicated and the slightest miscalculation can turn a
winning game into a mind rending defeat.
The lesson? Only enter a King and pawn endgame if you are completely
sure you know what's going on, or if you have no other choice. If there
is a promising and safe alternative, go for that every time!
Our next example illustrates a proper transition into a King and pawn
endgame. Pure calculation was necessary to assure Black that he was
making the right decision (if he wasn't able to make that calculation,
then he should avoid the King and pawn endgame), but in this case he
proved up to the task.
BLACK TO MOVE
G Jones vs. M Carlsen, Gausdal 2007
1...Kd4!, 0-1.
It looks like this lets White trade Rooks and promote his pawn first.
The resulting Queen endgame doesn't look easy at all (even if it is an
easy win, one usually tries -- if at all possible -- to avoid Queen
endgames in favor of simpler solutions). Surely Black had better than
this?
However, if this is true, we're left wondering why White resigned after
1...Kd4. Clearly, there's more going on than meets the eye!
The obvious point of 1...Kd4 is that his c-pawn is free to advance. Also,
White can't push his own passed pawn due to Rxb6. So the only chance
White has is to go for that Queen endgame:
2.Rxe6 fxe6 3.a6 c3 4.a7 c2 5.a8=Q c1=Q+
TURNS OUT TO BE EASY!
So far, neither side has had any choices to make. However, now that
we've reached this position we can see that the game ends abruptly in
favor of Black.
6.Kf3
The only legal move.
6...Qf1 mate! Picking up the White Queen by 6...Qh1+ is also "promising," but why not end things immediately?
This was a very easy calculation (he saw this
position when he played 1...Kd4) because all the moves were forced.
|