Endgame Query

[John Upham]

Help!

I'm stuck on a King and Pawn ending from Larry Evans "Chess Endgame Quiz"

It is Diagram 2(! : this is not promising), page 19 by Sackmann from a tournament game :

White : King on e5, Pawns on g2, h3 and g4

Black : King on h4, Pawns on g3 and g7

White to Play and Win (allegedly)

I assume the correct move is 1. g5! Kg5: 2. Ke4 Kh5 3. Kf3 and Evans gives Black's moves here as 3..Kh4 (or 3..g6) and now 4. Kf4 g6 5. Ke3 Ksomewhere 6. Kf3 Kh4 7. Kf4 g5+ 8. Kf3 and wins which is true.

However, Black can play 3..Kg5 which seems to draw trivially.

What have I missed?


I check the position in Averbakh and Maizelis (Diagram 594) and the starting positions agree and Averbakh makes the same error.

Thanks for any help.

[Roger de Coverly]

However, Black can play 3..Kg5 which seems to draw trivially.

What have I missed?

Nothing I think. The position (white to play) with a King on f3, pawns on g2 and h3 against a Black King on g5 and pawns on g7 and g3 is a draw according to the Shredder tablebase
http://www.shredderchess.com/online-che ... abase.html

With 6 man endings, that site is becoming very useful. We used it to predict the Jones game in the Olympiad where he had queen against rook and two pawns.

[IM Jack Rudd]

The tablebase gives the position after 1...Kxg5 as a draw.

[John Upham]

Thanks Jack. There cannot be too many examples of Averbakh getting and King and Pawn ending wrong!

But surely the drawing move is not difficult to see? Seems odd to me.

Possibly Evans assumed Averbakh was correct?

[Maxim Devereaux]

Short answer: yes, it looks like they're both wrong.

Tablebases agree that Kg5 draws (not exactly trivially - Kg5 Kxg3 Kh5! is an important zugzwang position).
And if White plays 3.Kf4, then 3. ...Kh4 4. Kf3 Kg5! (also ...g5) leads to the same.

[John Upham]

Short answer: yes, it looks like they're both wrong.

Tablebases agree that Kg5 draws (not exactly trivially - Kg5 Kxg3 Kh5! is an important zugzwang position).
And if White plays 3.Kf4, then 3. ...Kh4 4. Kf3 Kg5! (also ...g5) leads to the same.

Thanks Max. Do you think Averbakh missed the Kh5! move? I guess we will never know!

Maybe the endgame specialists knew about this error years ago!

I've checked in Lamprecht and Muller and cannot locate this endgame in their two books.

[Kevin Williamson]

Bloody tablebases - I had an opponent announce mate in 43 against me last year in a correspondence game as soon as we got down to six men on the board.

I did think I might have been holding it, but then I’m no Averbakh!

[Sean Hewitt]

Bloody tablebases - I had an opponent announce mate in 43 against me last year in a correspondence game as soon as we got down to six men on the board.

Isn't that, ahem, cheating?

[Kevin Williamson]

Bloody tablebases - I had an opponent announce mate in 43 against me last year in a correspondence game as soon as we got down to six men on the board.

Isn't that, ahem, cheating?

Not in the ECF County & District Correspondence Chess Championships. There’s no mention of computer assistance in the rules, which is very sensible since it’s impossible to prevent.