Hou Yifan - Humpy Koneru
Robert asked some relevant questions about the game of the previous post. Let's see what we can learn.
This is the position after move 17. ... Kf7
 |
| White to move |
3r3r/1pp1bkp1/2p1p3/p3P2p/P7/1P6/1BP2PPP/3R1RK1 w - - 0 18
If you think away the rooks and the bishops, only white can try to play for the win. Black must hold the draw. White can try to create a passer on the kingside by converting his pawn majority there, while black cannot convert his queenside majority into a passer due to his doubled pawns. "Rule": one holds two.
The general "rule" for this position is that white must trade pieces, not pawns, while black must trade pawns, not pieces.
The open d-file is a line of attack where both parties strive for dominance. No side can permit the other to get the upperhand on this file and invade. So the most likely outcome is to trade the rooks along this file, which plays into the hands of white.
A few moves later we have this position after move 25.f4
 |
| Black to move |
8/1ppr1kp1/2p1p3/p3P2p/P1P2P2/1Pb3P1/4K2P/2B2R2 b - f3 0 25
Humpy makes a mistake here by playing 25. ... g6?
That is way too passive, and prevents her King from reaching g6-f5-e4 anytime soon. The next step for white is to trade the rooks. But not on d1, where her king would be passive, but on d3.
Another important move is 32.g4 where white makes her own life unnecessarily complicated.
 |
| Black to move |
5b2/2p2k2/1p2p1p1/p1p1P2p/P1P2PP1/1P2BK2/7P/8 b - - 0 32
Better had been 32.h3 here. Why? No idea yet, and I don't like parroting the author without understanding it myself. Maybe something for a future post when you guys are interested?
According to Stockfish, there is nothing wrong with 32.g4. It is even better than 32.h3. So I have to think for myself again. what is going on?
ReplyDeleteTemposchlucker asks:
ReplyDelete"Maybe something for a future post when you guys are interested?
I'm certainly interested!
I've had GM Stockfish working for over 3 days. In the position after 17...Kf7, he 'thinks' there were 8(!) different continuations that maintained the balance [0.0] of position. I'm still working through the analysis trying to figure out an overview explanation that makes HUMAN sense—at least, to ME.
Reminder: GM Nimzowitsch, My System, Part 2, Chapter 2, Doubled pawns and restraint has a lot to say. In general terms, Black's structure is statically strong but dynamically weak. In plain language, Black should not advance the pawns unless absolutely necessary so as to avoid the dynamic weaknesses that ensue. In the position after 17...Kf7, White cannot penetrate to the queenside with just his King; there is insufficient weak white squares. In the position after 31...Bf8, look at the “super highway” along the white squares into the queenside. Also consider how the pawn structure has transformed White’s bishop into a “good” bishop while Black’s bishop has been transformed into a “bad” bishop. Hard to believe that a 2600-rated player did this to herself.
Dvoretsky’s Endgame Manual, 2nd Edition, has the following general advice [“rules”!] in Chapter 6: Bishops of the Same Color:
ReplyDeleteThe Bad Bishop
A vital principle [“RULE!”] of chess strategy (which is applicable to more places than the endgame) requires us NOT to place our pawns on the same color as our own bishop.
In the first place, pawns that are fixed on the same color as the bishop limit its mobility—this is why such a bishop is called “bad.”
In the second place, a bad bishop is unable to attack the enemy pawns (which are usually placed on the opposite color squares), which dooms it to passive defense of its own pawns.
And third, since both pawns and bishop control only one color of squares, there will be “holes” in between those squares that the enemy pieces will occupy.
All of these applicable “rules” indicate that Black should be doing her best to activate her king along the white squares, while avoiding placing her pawns on black squares UNLESS needed as part of her counterattack. She should have “attacked” by advancing the g-pawn against White’s kingside majority. Perhaps she “forgot” that the mobility of the pieces (including the king) is vitally important in endgames!
In the original position, White has only one pawn on the same color as her bishop. After 31...Bf8, White has all of the kingside pawns on the same color as her bishop. The vital difference is that Black’s two pieces have become extremely passive, without dynamically attempting to prevent an incursion by White on the white squares. In spite of the fact that White has the formally “bad” bishop, Black’s “good” bishop is being squeezed into a smaller space and has no viable targets—it only has 3 squares to shuffle back and forth on. The Black queenside pawn structure blocks any movement of the Black bishop in that direction and is an open invitation to invade on the white squares if the bishops disappear and White manages to get a passed e-pawn.
As for 32. g4 vice 32. h3, my preference would have been driven by general principles: Steinitz (following Philidor) advocated utilizing a pawn phalanx in such a way as to be able to recreate the phalanx as it advances. There is also Rubenstein’s Rule: DO NOT HURRY! I suspect that White anticipated breaking into the kingside on the white squares, and so was not worried about rebuilding the pawn phalanx as the pawns advanced. After all, Black doesn’t appear to be doing much of anything to prevent White’s obvious ideas.
"After 31...Bf8, White has all of the kingside pawns on the same color as her bishop"
ReplyDeleteA bishop can only be bad when the pawns are fixed. When they are mobile, it is only temporarily.
Good point: that may be why White played 32.g4 immediately.
ReplyDelete