I couldn't find last weeks solution so I tried a computer (I admit I should have tried harder). Surprisingly there seems to be no win.

1.f5 Kd3!! (Rybka evaluates 0.00)

Some lines
2.e6 fxe6 3.g6 Nc3+ 4.Ka5 Nd5 5.g7 Nf6 6.fxe6 Ng8 7.Kb5 Kd4 8.Kc6 Ke5 9.Kd7 Kf6
2.g6 fxg6 3.fxg6 Nc3+ 4.Kb4 Nd5+ 5.Kc5 Nf4 6.g7 Ne6+
2.Kb5 Nd4+
2.Ka3 Nf4 3.e6 fxe6 4.g6 Nh5 5.g7 Nf6 6.fxe6 Ng8 7.Kb4 Kd4
2.Ka5 Nf4 3.e6 fxe6 4.g6 Nh5 5.fxe6 Ng7 6.e7 Kc4 7.Kb6 Kd5 8.Kc7 Ke6 9.Kd8 Kf5
2.g6 fxg6 3.fxg6 Nc3+ 4.Kb4 Nd5+ 5.Kc5 Nf4 6.g7 Ne6+

without endgame tablebases, Rybka can't find the correct solution, and will give a bad evaluation of the position.

In this case pierre rybka is right.

LOL

I have runnig tablebases at the background, so I think this endgame study is really cooked.

and chessvibes doesn't check a study before publishing it?
What do they have to answer, to the fact that the study is a draw? Nothing? do they care that it was wrong? or are they going to do the same mistake once again?

@JustBe

"Ask you money back!"

fuck your nice, tight, and sweet little ass hole.

@Willem

"This study was composed long before tablebases were invented"

That's no excuse. And before that fat jurk of Yochanan Afek gives a study to think about, he should check.

If it's wrong, he should apologize, and say this will not happen again.

This study was composed long before tablebases were invented.

This happens now and then.

Tablebases are max 6 pieces. I count 7?

Yea, this is SHOCKING news!

Ask you money back!

hey Pierre watch the language

the other problem is computers are for the most part HORRIBLE at studies since they dont consider certain situations as won if they arent on their calculation horizon,

and people dont auto check everything with computers. thats for people who cannt think.
you can still enjoy a study for the idea even if its cooked with out getting insulting

@Ron i thought it was 6 pieces plus 2 kings table bases could handle.

@Pierre there is no good reason to use that strong language about Afek.

The initial position is not a tablebase position, but a there are many ways to reach one when anything is captured. In that case the engine evaluates no further, but returns the evaluation of the database. So I am not 100% sure the studie is cooked, but the variations after 1.f5 Kd3 look convincing.

I hope Afek will keep up the good work of sharing the beauty of chess composition.