This is an actual position in game 5 between Leko and Carlsen in their match in Miskolc, Hungary. It is White to move. Leko could have tied up the match if he would have found the right continuation. How should he proceed?
Chess Daily News from Susan Polgar
How about Rxh3 then march the pawns?
How about Rxh3 then march the pawns?
Creative idea, but I can’t get it to work:
1.Rxh3+ Kxh3
2.c5 Rg4+
3.Ka5 Rc4
… looks secure for Black. Don’t take this to the bank, though.
Something else that doesn’t seem to win:
1.c5 Rg4+ 2.Ka5 Rh4 3.Rxh4 Kxh4 4.c6 h2 5.c7 h1=Q 6.c8=Q Qb1+
… and (I think) Black has a perpetual.
I think 1.Rx3 would lead to a draw.
The immediate
1.c5 seems to draw as well,
1…Rg4+
2.Ka5 Rh4
3.Rxh4 Kxh4
4.c6 h2
5.c7 h1Q
6.c8Q leads to draw according to tablebases.
“1.Rxh3+ Kxh3
2.c5 Rg4+
3.Ka5 Rc4
… looks secure for Black. Don’t take this to the bank, though.”
After that white plays 4.Kb6 and this is winning for white. So ,yes 1.Rxh3+ is the right move. Imediate c5 is a mistake because of
1.c5 Rg4+
2.Ka5 Rh4
”After that white plays 4.Kb6 and this is winning for white.”
According to tablebases after 4.Kb6 this is draw.
One line (maybe the only one) that seems to win is :
1.c5 Rg4+
2.Ka5 Rh4
3.Rd7! h2
4.Rd1 h1Q
5.Rxh1 Rxh1
6.b6 wins for White.
Pharaoh
Very pretty, Pharaoh. I’d never have found 3.Rd7!
1. Rh8! h2 2. Rxh2!! only now!
and advanced white pawns.
The sacrifice in h3 or h2 makes all diference!
This is very difficult; the result hangs by a thread. I think I would try 1.c5, allowing 1…Rg4+ 2.Ka5 Rh4. Now White can try 3.Rd7 with the idea 3…h2 4.Rd1 h1=Q 5.Rxh1 Rxh1 when White should win with 6.b6.
Thoughts?
Whoops! Didn’t see that Pharoah had scooped me on that. I also like the line given by anonymous a few minutes ago: 1.Rh8, when 1…Rg4 fails because 2.b6 puts the pawn too far forward: 2…Rh4 3.Rxh4 Kxh4 4.b7 and White will win Black’s new Queen with a skewer.
This endgame is too difficult for me so I fed it into my computer (Toga 1.3.4). The
program (currently at 21 plies + 5 piece table bases) enthousiastically recommends
something else….
Excuse me, but what about simply 1. Ka5 and 1..Rg4 doesn’t work because of 2. b6 ?
1. Rh8! h2 2. Rxh2!! only now!
and advanced white pawns.
The sacrifice in h3 or h2 makes all diference!
Sorry, I don’t see it. Why can’t Black draw with 2…Kxh2 3.c5 Rg4+?
1.Rh8, when 1…Rg4 fails
1…Rg4? does indeed lose to 2.b6. That’s why Black plays 1…h2 instead, and draws.
Excuse me, but what about simply 1. Ka5 and 1..Rg4 doesn’t work because of 2. b6 ?
You guys and your 1…Rg4 fetish 🙂 Yes, 1.Ka5 Rg4? loses after 2.b6, but Black draws with 1…Rg5.
White has 3 winning moves here. 1.Rh8 (Rf6 2.Rh5!) as favored by Chessbase’s analyst wins well enough. And 1.Ka5, which was my own (sans-computer) choice, works as indicated here. But neither of them is the “clearest” move, which is what I think Susan had in mind…