So maybe White should meet 1...Rh2 by 2 Rf3. Then if 2...b2 3 Rf6 and Rh6 is in time and has destroyed this zugzwang, and if 2...h3 3 Rxb3 is adequate. At least I think it's adequate - yes of course it is; the black king's still on e7 so just Kf7 4 Rg3 defends the g-pawn. Maybe that's what you meant by 'without White's cooperation' and you're way ahead of me, Dink?

Kc1 allows ...Rc2+ and h2

I think its a genuine reciprocal zugzwang, as Black to move can't win.

It's not so much that as just that it's not zugzwang, surely? White can play 1 Kc1. You can go 1..Rc2+ and Rc3 if you like, but that's back to the game

Yes, this was one of a number of extremely bizarre observations Yermolinsky made (sweere). Either he'd been refreshing himself extensively or someone else was using his server, if you ask me.

I didn't understand 1...Rh2 in this diagram at all. White plays 2 Rh6 and sits there - whatcha gonna do? Black can go 2..h3 3 Kb1 b2 4 Rf6+ Kg8 5 Rg6+ Kh7 6 Rh6+ Kg7, which looks like it might be zugzwang for a moment, but it isn't - 7 Ka2= (another GM, Tigrano, was overlooking a similar Ka3 saver with the rook on the third rank). Basically it's just a complete draw; Black's king can't join in, his rook can't leave the h-file, and he can't put White into zugzwang. Of course Black can play ...Rh1 and push the pawn to h2. Then he can put Black into zugzwang and win the g-pawn, but the resulting position is a total draw; White just keeps the rook on the h-file and Black's king can't do anything on either side of the board.

I certainly thought at the time Rf4 was drawing as well: 1 Rf4 Rh2 2 g6 h3 3 g7 Rg2 4 Rh4 seems to do the trick. But Rf6 is more relaxing somehow.

Micawber wanted to play 72...Kd5 in the Linares thread and I agree that's closer than Topalov's quaint trip to the a-file: I was trying 72...Kd7 which at the time I thought was superior for some reason, although I don't seem to be able to put my finger on it just now......

Thanks, Willempie!



Muhahaha!! Now the world is mine...

(generated via Apronus, which seems much more user-friendly than your link (especially, if like me, you're linguistically challenged   )

Yip...no way for White to reach the Vancura position 

BTW, could some kind soul please explain how to insert a diagram..

IMJohnCox wrote on 03/05/07 at 13:47:56:

Chesslive's database seems to be down so I cant quote the game, but I was thinking of a Korchnoi-Karpov (or vice versa) game in which Karpov managed to draw an endgame by going for a R+K vs R+K+2 split pawns, which was apparently a theoretical draw. It is in Timman's book (art of analysis) and I spent hours on this game without understanding half of what is going on.

IMJohnCox wrote on 03/05/07 at 12:52:52:

Not going into the lines themselves yet, but is losing the g-pawn such a problem? If losing it buys time to ship the king towards the b-pawn and the rook has the opportunity to get near the h-pawn (assuming the white one gets lost in the process of moving the king) I think it is still a dead draw as both pawns cant progress.