Instructive endgame challenge.

Technical questions regarding Openings, Middlegames, Endings etc.
Jonathan Rogers
Posts: 4656
Joined: Tue Nov 18, 2008 9:26 pm

Instructive endgame challenge.

Post by Jonathan Rogers » Sat Aug 18, 2012 10:27 pm

Consider the following position (sorry, I don't know how to produce a diagram, but maybe someone can help? I'd be very grateful if so).

White: pawns g2, g4, h4: king d2
Black: pawns a3, a2, g6, h7: king a1

It is White to move and clearly his choices are between 1. Kc1 or 1. Kc2. Your questions are:

1 Show why after 1 Kc1, Black draws with 1...h5 or 1...h6 but loses with 1 ...g5

2 Show why after the superior 1 Kc2, Black now loses with 1...h5 as well as 1...g5 (but can still draw with 1...h6)

(The problem arises from an earlier position given in the Times, where Keene presumably only had space to say about the position above that after 1 Kc1 Black draws with ...h6. I think the position is worth looking at more closely, if only to see why Kc2 is probably a better practical bet than Kc1)

Clive Blackburn

Re: Instructive endgame challenge.

Post by Clive Blackburn » Sat Aug 18, 2012 11:02 pm

Here is the position:


Matthew Turner
Posts: 3604
Joined: Fri May 16, 2008 11:54 am

Re: Instructive endgame challenge.

Post by Matthew Turner » Sun Aug 19, 2012 5:35 am

The answer takes quite a few steps, but I think I've got it.

Richard Bates
Posts: 3338
Joined: Fri Nov 14, 2008 8:27 pm

Re: Instructive endgame challenge.

Post by Richard Bates » Sun Aug 19, 2012 8:13 am

Jonathan Rogers wrote:...

(The problem arises from an earlier position given in the Times, where Keene presumably only had space to say about the position above that after 1 Kc1 Black draws with ...h6. I think the position is worth looking at more closely, if only to see why Kc2 is probably a better practical bet than Kc1)
Hmmm. Did Keene have the K on c1 or c2? Easy to see the difference after 1...h5, but less easy to see why there's no difference after 1...g5. With K on c2 white has Kb3 at the key moment, but struggling to find the win with the K on c1...

David Sedgwick
Posts: 5249
Joined: Mon Apr 09, 2007 5:56 pm
Location: Croydon

Re: Instructive endgame challenge.

Post by David Sedgwick » Sun Aug 19, 2012 9:29 am

Richard Bates wrote:
Jonathan Rogers wrote:...

(The problem arises from an earlier position given in the Times, where Keene presumably only had space to say about the position above that after 1 Kc1 Black draws with ...h6. I think the position is worth looking at more closely, if only to see why Kc2 is probably a better practical bet than Kc1)
Hmmm. Did Keene have the K on c1 or c2? Easy to see the difference after 1...h5, but less easy to see why there's no difference after 1...g5. With K on c2 white has Kb3 at the key moment, but struggling to find the win with the K on c1...
I think Kd2 does the trick. It also works with the K on c2, although Kb3 is then quicker.

Jonathan Rogers
Posts: 4656
Joined: Tue Nov 18, 2008 9:26 pm

Re: Instructive endgame challenge.

Post by Jonathan Rogers » Sun Aug 19, 2012 11:08 am

Thanks to Clive for providing the position.

David is right about the 1 Kc1 g5 line.

But I still want to see some moves showing how Black draws with 1...h6. You should find that there is one plausible variation where Black would lose if (but only if) White had started 1 Kc2. In other words, there is only one way to draw after 1 Kc2 h6, ie, Black would have more margin of error after 1 Kc1 h6.

I will give the answers later with fuller explanation for those who might currently find themselves rather lost.

Paul McKeown
Posts: 3735
Joined: Thu Apr 12, 2007 3:01 pm
Location: Hayes (Middx)

Re: Instructive endgame challenge.

Post by Paul McKeown » Sun Aug 19, 2012 12:46 pm

Thanks for this Jonathan, good way to get my brain working on a Sunday afternoon. If 1. Kc2 then 1... h6 2. g3!? g5 =, however here 2... h5?? loses to 3. gxh5 gxh5 4. g4 as there is a check at the end of the pawn race, with immediate mate. If the King was on c1, this variation would not work for White, as Black would promote with check.

Jonathan Rogers
Posts: 4656
Joined: Tue Nov 18, 2008 9:26 pm

Re: Instructive endgame challenge.

Post by Jonathan Rogers » Sun Aug 19, 2012 12:57 pm

Paul McKeown wrote:Thanks for this Jonathan, good way to get my brain working on a Sunday afternoon. If 1. Kc2 then 1... h6 2. g3!? g5 =, however here 2... h5?? loses to 3. gxh5 gxh5 4. g4 as there is a check at the end of the pawn race, with immediate mate. If the King was on c1, this variation would not work for White, as Black would promote with check.
and to complete things, after 1 Kc2 h6 2 g3 (best) g5 3 Kc1, the draw is ...?

AustinElliott
Posts: 665
Joined: Tue Mar 01, 2011 11:01 pm
Location: North of England

Re: Instructive endgame challenge.

Post by AustinElliott » Sun Aug 19, 2012 1:37 pm

The key point seems to be that, in several different variations involving a mutual pawn race down the K-side, the King on c1 can be checked by a Queening Black pawn, while on c2 you get denouements like ...g1=Q (NOT with check) h8=Q mate. White has to allow a pawn race, since exchanging off the K-side pawns to leave just one White pawn, or even TAKING the last Black K-side pawn, is an automatic stalemate.

Would be hard to pick the square (c2 or c1) 'on calculation', though, since some of the lines are *!**dy long. Or is that just the 150s patzer in me speaking? Probably...

Paul McKeown
Posts: 3735
Joined: Thu Apr 12, 2007 3:01 pm
Location: Hayes (Middx)

Re: Instructive endgame challenge.

Post by Paul McKeown » Sun Aug 19, 2012 1:39 pm

Ha! I see your little joke! After 1. Kc2 h6 2. g3 g5 3. Kc1, 3... gxh4 is rather humorously a blunder, because of 4. gxh4 h5 5. Kc2! So Black must play 3... h5 to draw, when 4. Kc2 is met not by 4... gxh4?? but by 4... hxg4.

AustinElliott
Posts: 665
Joined: Tue Mar 01, 2011 11:01 pm
Location: North of England

Re: Instructive endgame challenge.

Post by AustinElliott » Sun Aug 19, 2012 4:17 pm

Yes, never cease to be impressed how many tricks apparently simple pawn endings can contain, esp involving pawn 'breakthroughs' and stalemate themes... quite apart from the dreaded opposition.

PS Could we have a chessical Monty Python sketch where a bunch of men with cushions run in and shout:
"NO-ONE expects the distant opposition!"

Jonathan Rogers
Posts: 4656
Joined: Tue Nov 18, 2008 9:26 pm

Re: Instructive endgame challenge.

Post by Jonathan Rogers » Sun Aug 19, 2012 6:19 pm

I am gald that people seem to have enjoyed this - it kept me distracted for a day or so while I was discovering some of the nuances alluded to above. The answers seem to be

1...h5 loses after 1 Kc2 because of 2 gxh5 gxh5 3 g4 hxg4 4 h5 g3 5 h6 g2 6 h7 g1=Q 7 h8=B (so one would hope). this cannot work if White had played 1 Kc1 because Black would have queened with check. Everyone seems to have got this.

1...g5 will lose no matter whether White has played Kc1 or Kc2. Play then continues 2 hxg5 h5 3 g6 hxg4 4 g7 g3, and now White does not have time to promote because of stalemate and must release the Black king first. If he had played 1 Kc2, it is easy - he just plays 5 Kb3 Kb1 6 g8=Q a1= Q 7 Qg6+ and 8 Qc2 mate. But as David thought, he can still win even if he had played 1 Kc1. Then he would play 5 Kd2 Kb2 6 g8=Q a1=Q 7 Qg7+ and he can zigzag forwards bey giving diagonal checks until he mates on c2. Eg 7...Ka2 8 Qf7+ Kb1 9 Qf5+ Kb2 10 Qe5+ Kb1 11 Qe4+ Kb2 12 Qd4+ Ka2 13 Qc4+ Kb1 14 Qc2 mate (12...Kb1 13 Qd3+ Ka2 14 Qc4+ and Qc2 mate).

1..h6 is a quick draw after 1 Kc1: 2 g3 h5 leads to the first line with White's king on the wrong square. 1 Kc2 is tougher because after 1...h6 2 g3 Black has to play 2...g5 and then has to think a bit after 3.Kc1. He draws with 3...h5 4 gxh5 g4! and his stalemate bid comes too early for White to be able to pre-empt it. But not 3...gxh4 4 gxh4 h5, as Paul noted, because now White plays 5 Kc2 again and transposes to the right side of the first line above!

So 1 Kc2 is a better bet for two reasons. First, it means that Black cannot draw at once with 1...h5. But more importantly even if Black plays the right first move 1...h6 he has to be accurate again a bit later on move 3, when deciding between 3...gxh4 and 3...h5. And that's how people like Keith win drawn endings - by finding the ways that put their opponents to their test as often as possible. Opponents tend to find one good/only move in a row, when one is all that is needed. They are much, much less good at finding a series of only moves.