## Monday, February 06, 2006

### I still don't get it.

If logic is your guide, life can be a burden sometimes.
I want to play better chess. Recent analysis of my games at Corus revealed that endgame study would benefit 44% of my games. So I started endgame study.
That's logical.
Since I don't want to do a half job, I started with the mother of all endgames: pawn endgames.
Because any endgame can only be well-evaluated by the underlying pawn endgame, when all pieces are traded off.
That's logical.
At this moment I can say I master endgames with only one pawn.
So I moved on to endgames with two pawns on the same file.
And now I'm lost. I use SOPE, which has a lot of exercises with explanations.
Much to my surprise I read in about 25% of the cases that a grandmaster didn't play that specific exercise well in an official game and lost unnecessary. Mind you, we are talking about only two pawns!

Take for instance the diagram below.

White to move and draw.

It is part of the same problem I showed you thursday. Only two moves later in the line.
If white plays 1. Ke7 here he can hold the draw.
If he plays 1. Kf5 the Nalimov tablebase says: lost by mate in 20.
Actually any other move loses too.
Why?
If I would get such a position in a game and would play it by intuition, I would give it 70% to 30% in favour of 1. Kf5. Inspired by some sort of reasoning as "this has nothing to do with opposition, it is a kind of a Reti manoeuvre with two targets: both pawns. If I stay in the middle between them that must be good enough."
If I toss a coin my chances are better than playing by intuition.
But this line is about 11 moves long! Tossing a coin 11 times results in a chance of 0.00048828125 to bring the line to a good end.
I looked at this position for hours and hours. I did a lot of similar problems that revealed much mysteries about such endgames. I know for instance a lot more now about the impact of pawn moves in such positions.
But I still don't quite get it.
I want to find the rationale for the choice between 1. Ke7 and Kf5.
If I have to analyse such lines in time trouble OTB there must be a reasoning which I can trust.
If I would be wise I probably would move on and accept that you can't understand everything. At this pace I'm at Lucena's position in 2078.
But that's wise and not logical.
So I'm going to use my time to reveal the mystery of this diagram.
How can I ever understand more complex endgames if I don't understand this?
That's logical.
After all I've seen people busy with much more silly occupations like solving Sudoku's or throwing darts:)

1. I can sympathize with your sincere goal to play better endgames ( like us all ) but it seems like this approach is likely to sap the fun out of the whole enterprise after all is said and done.

1) Study some of the great classic endgames.

2) admire the genius of ( Capablanca? Pillsbury? ) as they sacrifice pawns that should be worth their weight in gold ( I quote Chernev )

3) Hope that some of this knowledge (if not genius) rubs off on us by osmosis.

2. hey. i would say Ke7 makes sense because you need to get to the pawns, not just get between them. Otherwise black gets his king to your pawn with no trouble.

3. >Much to my surprise I read in about 25% of the cases that a grandmaster didn't play that specific exercise well in an official game and lost unnecessary. Mind you, we are talking about only two pawns!
...
I think you have just come up with the way to beat masters :-)!

This is the only line to a draw:
1. Kg5 Ke4 2. Kf6 Kd5 3. Ke7 Kc6 4. Ke6! b6 5. Ke5 Kc5 6. Ke4 Kc4 7. Ke3 b5 8. Kd2 Kb3 9. Kc1 Ka2 10. b4! Kb3 11. Kb1 =

How about this experiment. Solve this example at this diagram at white to move at 11.? should be easy right. Then move it out one further Black has moved 9...Ka2 was is whites move 10? If you can solve that move it out to blacks move 8 and solve for whites move. tell me if this helps

5. Funky,