- Endgame improvement

Reply

Anonymous says:

August 27, 2012 at 3:29 pm

b5+.. Kb7

seems a good start

Reply

Yancey Ward says:

August 27, 2012 at 4:19 pm

It is oh so tempting to take that pawn at g3, but one must see one overriding factor right from the start of this problem- white has the wrong bishop for the a8 square. In other words, if black can sacrifice the knight for the b-pawn at any point, black will gain a well known draw. Let’s just play through 1.Kg3 to make the point clear:

1. Kg3? Ne4

I am not completely sure this is the only draw for black, but I think Nd3 or Nd1 lose similarly to the main winning line for white I will discuss below. Continuing:

2. Kf3 Nc3

Another case where I am not completely sure if black draws with other moves. 2. …Kd5 looks ok to me, too, but I don’t want to spend the time and effort proving it so. Other knight moves just look losing to me, however. Continuing:

3. b5

If white plays a5, black just plays Nd5 followed by Nxb4 leading to a drawn B+P vs K ending (white’s bishop is the wrong color to keep black king out of a8). Also, if white plays 3.Bc5, black can draw with Nxa4, or he can just play Nd5 still followed by Nxb4. Either is a draw. Continuing:

3. …..Nb5 wins both pawns for the knight obtaining the draw by insufficient material.

So, where does white put the king? It should be clear that 1.Kf4 is an immediate draw after Nd3+ winning the b4 pawn outright. This pretty much leaves only Kf3. However, this ending is technically complex, and I need quite a bit more time to work this line out to determine whether or not white can win it.

Reply

Yancey Ward says:

August 27, 2012 at 6:41 pm

In my previous comment, I discussed how black draws easily if white takes the time to capture at g3 on the first move. Now for Kf3

1. Kf3 Nd3 (alts. later)
2. b5 Kb7 (alts. later)

Now white has the pawns protected, for the moment. White can capture at g3, play a5, or play Bd4 (preventing an immediate Nb2 or Ne5) as plausible plans. The instinct is to just clear the g-pawn so that the white king can join the fray without having to worry about it. Cont:

3. Kg3 Nb4

Black can try Nb2 here, but I will cover it later. Cont:

4. Kf3 Nd5

Black must avoid a move like Na2 with the intention of forking the pawns from c3 since white will just immobilize the knight on a2 by playing Bd2 and win. He can play Nc6 or Na6 since either capture with a white pawn leads to a drawn ending. Nd5 is a plan to bring the knight to either b6 or c7, or to do the fork from c3. Continuing:

5. Bd4

White can prevent Nc3 with Bd2 too, but I don’t like the idea of taking eyes off of b6: [5.Bd2? Kb6 6.Ke4 Nc7! with Nxb5 unstoppable.] Cont:

5. …..Nc7 (Nb4 6.Ke3)
6. Ke4 Na6
7. Kd5 Nb4 (harrass the king)
8. Kc4 Nc6 (planning harrass)
9. Bc3

White can now allow Kb6 since the white king now protects b5 and can respond with a5+: [9. …Kb6 10.a5 Kc7 (or 10. …Na5 11.Ba5 Ka5 12.Kc5+-) 11.Kc5 Na7 12.Be5 (12.b6?! Kb7 13.Bd4 Ka6 might not be winning for white) 12. …Kd7 13.a6! Nc8 (or 13. …Nb5 14.Kb5+-) 14.b6 Nb6 15.Kb6+-]. Cont:

9. …..Na7
10.Kc5 and this is surely won.

The major holes in this analysis are at moves 1, 2, and 3 but I now think they are clear losses. At move 3, black could have tried Nb2

3. …..Nb2
4. a5 Nc4
5. a6! Kc7
6. Bd4 Na3
7. b6 Kc8
8. b7 Kc7
9. Be5 wins. There are variations in this line, but they don’t look all that different from what I have discussed previously- in those, white’s king gets involved much as I showed previously.

At move 2, black could have played Kc7 instead of Kb7. This forces white to come up with a plan different from 3.Kg3 since that allows black to play Nb2 effectively. For examp:

2. …..Kc7
3. Kg3? Nb2!
4. a5 Nc4! and the a-pawn falls since white doesn’t have a6 with check as when the black king was on b7. At move 3 in this line, white can try 3.Bd4:

2. …..Kc7
3. Bd4 Nc1 (protect g3 with Ne2)
4. a5! Nb3
5. Bb6

Yes, I did look at Be5, too, but it looks iffy to me as a win. I like protecting a5 with the bishop rather than pushing the pawn which makes both b5 and a6 weak (light squares!). Cont:

5. …..Kb7
6. Kg3

Here, what can black do? He would like to play Nd4, but can’t. He can try to maneuver to attack b5 from a3 or c3, but white has time now to prevent this:

6. …..Nd2
7. Bd4! Nb3
8. a6!

This now works since white has already captured at g3 and he is playing a6 with check while controlling the g1/a7 diagonal. Cont:

8. …..Ka8
9. Be3

White now controls d4 and c5 preventing the knight from getting to c7 to fork the pawns. In addition, it is now harder to even reach a3. Continuing:

9. …..Na5
10.b6

The most direct win. Black cannot play 10. …Nc4 since white pushes b7 with check once again followed by Bf4+ and b8Q. Continuing:

10. ….Nc6
11.Kf3

Do not take f4 from the bishop: [11.Kf4? Nb4! 12.a7 Nd5=]. Cont:

11. ….Ne5 (Nb4 12.b7 Kb8 13.Bf4)
12.Ke4 Nd7 (Nc4 13.b7 Kb8 14.Bf4)
13.b7 Kb8
14.Kd5 Kc7 (Nf6 15.Kc6)
15.Bd4

A powerful waiting move. Cont:

15. ….Kd8 (Kb8 16.Kc6 Nb6 17.a7)
16.a7 wins.

Ok, so I think I have shown that the line 1.Kf3 Nd3 2.b5 Kc/b7 loses for black. I will consider 1. …Nd1 next.

Reply

Yancey Ward says:

August 27, 2012 at 6:51 pm

So, I earlier discussed the line 1.Kf3 Nd3, and I think I showed that black loses. The last thread I see is 1. …Nd1, but this move is obviously weak after just a moment’s consideration:

1. Kf3 Nd1
2. b5 Kd5

If Kb7/c7 white still plays Bd4. Continuing:

3. Bd4! Kd4 (what else now?)
4. b6 and the pawn can’t be stopped by either the king or the knight.

This was very nice endgame study. My favorite type of problem.

Reply

Yancey Ward says:

August 27, 2012 at 6:52 pm

Anonymous,

White must move the king on the first move since he is in check.

Reply

Anonymous says:

August 27, 2012 at 11:49 pm

I Think that 1.Kf3 Nd1 2.Bd4 (Wins, since the King controls the black “g” Pawn advance, and Threats to advance de a,b pawns to promote; and also can captute de g pawn. Note that the black Knight is full controlled). And, if 2. Bd4 Kd5, black can even play the bishop to the other diagonal cutting-off the knight, with a mov like 3. Bg7 +-).
Pn the other hand, if seems to be a mistake if white captures directly the pawn 1. Kxg3 ??, considering that the black Knigth can gain a tempo return to the game to control de advance of a,b pawns and even sacrifice on the b pawn to reach an essy draw, since “a8” is light squared!!

Reply

pht says:

August 28, 2012 at 7:46 am

White king can approach black king without bothering too much about black pawn.

1. Kf3!

Attacking the knight.
Black will not play Nh3? Kxg2 and knight is dead.
And he will not play Nd1? Bd4 and knight looks stupid. So:

1. … Nd3 (enforced)
2. b5+ Kb7
3. Ke4!

Attacking the knight again.

3. … Nb2 (Nb4? Bc5! Na2/Nc2 looks good for white)
4. a5 Nc4
5. Kd4
and now black would hardly play the loosing:
5. … Nxe3?
6. Kxe3 g2
7. Kf2

My basic idea is that white king, if taking care of his steps, has large freedom to participate in the attack on black king, while bishop controls the g-pawn.

But here are too many variations for me to analyze this further….

Reply

Anonymous says:

August 28, 2012 at 8:12 am

i think it should be done like this:
1. Kf3 Nd1
2. b5+ Kany
3. Bd4
and the knight is dominated. the king can`t attack the bishop because the b-pawn would walk through. a5-a6 and Kxg3 to follow. this should be decisive.

Reply

Jassi says:

August 28, 2012 at 12:20 pm

Susan, your test for robot is is a trouble. At least 5 times now I have failed to publish my comment. Why are you so bothered by that thing and why would anybody do that in the first place?

Reply

JGC says:

August 28, 2012 at 9:58 pm

Susan,

This position has been analysed in

http://escaque.blogspot.com/2012/03/algunas-reflexiones-sobre-el-campeonato_25.html

where the Uruguayan Championship 2012 has been where the Uruguayan championship has been carefully scrutinized.

The proper credits or source should go to the Jovenes Promesas blog.

All the best, Julio GC

Reply

JGC says:

August 28, 2012 at 9:59 pm