< Earlier Kibitzing · PAGE 2 OF 5 ·
|Jun-23-11|| ||thegoldenband: What's mystifying about this is that even if Gelfand didn't see the key 22nd move, the position is still very good for Black even if he doesn't play 22...Ba5: pinned Knight, active Rooks and Bishop, and after ...Rxf5, a material advantage. I didn't see 22...Ba5, but I would've played into the ...Bxd5 line even so. Am I missing something?|
|Jun-23-11|| ||SamAtoms1980: Spent 15 minutes on this one, and I'm definitely glad I took the time to work through the moves. Of course, this was a lot harder for Gelfand to do over the board at rapid.|
On Monday, we got an Arabian Mate; on Tuesday, we swung north and westwards with a Greek gift; and yesterday we returned to the Middle East with a variant of Dilaram's double rook sac. Today, we're keeping it right here in Persia, and mixing the Greek back in with Alexander the Great against the Persians at the Battle of Gaugamela. Alexander won the encounter in 331 B.C. by advancing at an angle, then going right for the king.
And so it is here: Black topples the cavalry with 19....Bxd5! White must recapture, and yet cannot recapture, for either way is doom. If: 20.cxd5 Qxc1 21.Rxc1 Rxc1+ 22.Ne1 Ba5, and White's material corps will be wiped out. Or: 20.Rxd5 Qxd5! 21.cxd5 Rxc1+ 22.Ne1 Ba5 meets the other path.
Although Gelfand let this one get by him, he moved on anyway and will face off with Anand for the title.
|Jun-23-11|| ||Morozov: 4/4. First time ever!|
|Jun-23-11|| ||SimonWebbsTiger: a better puzzle (or question?) would be:
how many people screamed take on d5 at their computer screen and then screamed "OMG!" when it didn't happen when this was broadcast live?
|Jun-23-11|| ||notyetagm: Game Collection: PINS! PINS! PINS! PINS! PINS! PINS! PINS! PINS!|
|Jun-23-11|| ||dzechiel: <fightclub: Hello Chessgames. Just signed up>|
Allow me to be the first to say "Welcome aboard."
|Jun-23-11|| ||David2009: Kamsky vs Gelfand, 2011 Black 19...?|
White has a back rank weakness: 19...Bxd5 20.Rxd5 Qxd5 21.cxd5 Rxc1+ 22.Ne1 Ba5 0-1. Or 20.cxd5 Qxc1 transposes. Time to check:
click for larger view
Trust your opponent?
|Jun-23-11|| ||Old Wolf: Should be a Monday problem. Two moves to win material, I would think that looking at this position the majority of players would play Bxd5 without any thought|
|Jun-23-11|| ||Once: <fightclub> And I can I be the second to say hello? You've picked a great handle by the way. The first rule of chessgames.com is that we do not talk about you. |
And the second rule is that we DO NOT talk about you...
|Jun-23-11|| ||perfidious: This one was easy-while the variation was long, it was a forcing series of moves, so not at all difficult to spot.|
|Jun-23-11|| ||Once: Odd - we mortals find it in a heartbeat, but stellar creatures like Gelfand and Kasmky missed it OTB. And the usually infallible phony doesn't see it. So what gives?|
Well of course, it's a puzzle so we are on our guard for tactics and tricks. In the real world if our bishop is attacked we move it away. But in puzzleworld we throw it further into the fight.
In the real world we take pawns when they are offered. In puzzleworld we ignore them as mere chaff while we press on for mate.
But there's more to it than that. When the puzzle opens, the white back rank seems very securely defended. Two rooks, a queen and a knight all have a say on a part of the back rank. So it seems counterintuitive that a back rank tactic would be "on".
And then we have the sneaky Ba5 move. We need to see that to bring home the point.
Interestingly, it doesn't matter much whether white recaptures with 20. Rxd5 or 20. cxd5.
I suppose the learning point is to be on guard for tactics even in positions that seem fairly well nailed down.
|Jun-23-11|| ||lost in space: Got it, I guess.
19... Bxd5 20. Rxd5
(20. cxd5 Qxc1 21. Rxc1 Rxc1+ 22. Ne1 transposing to mainsline)
20... Qxd5! 21. cxd5
(else losing the rook)
21...Rxc1+ 22. Ne1 Ba5 23. Kf1 Rxe1+ 24. Qxd1 Bxd1 25. Kxe1 Rxf5
click for larger view
|Jun-23-11|| ||lost in space: <Well of course, it's a puzzle so we are on our guard for tactics and tricks. In the real world if our bishop is attacked we move it away. But in puzzleworld we throw it further into the fight.>|
Exactly. I am sure that I would miss this OTB.
|Jun-23-11|| ||TheBish: Kamsky vs Gelfand, 2011|
Black to play (19...?) "Medium"
I got this right away, like 15-20 seconds. Not bragging, just gauging the difficulty. It's natural to look at exchanges, so that greatly simplifies this one, as that's mostly what it is.
White's last move was obviously 19. exf5, expecting 19...Bxf5. Surprise! Now there are two ways to go, but they are similar.
A) 20. cxd5 Qxc1! 21. Rxc1 Rxc1+ 22. Ne1 Ba5!
The point. Black gets the queen (as moving away allows mate, e.g. 23. Qf3 Rxe1#), so White has nothing better than 23. Kf1 Rxe1+ 24. Qxe1 Bxe1 25. Kxe1 and Black is up a rook.
B) 20. Rxd5 Qxd5! 21. cxd5 Qxc1+ 22. Ne1 Ba5!
We have just transposed to variation A).
There are two other possible continuations: playing on a piece down after 19...Bxd5, and resigning!
It looks like White could have improved with either 19. h3 or 19. b4, but I like the two bishops against two knights.
Wow, Gelfand missed this! It's one thing to blunder and allow a combination, but it's another to thing to study the board after the blunder has been made, and miss it with the possibility staring you in the face. Amazing that two GMs missed this -- gives me some hope!
|Jun-23-11|| ||Once: <TheBish> I didn't think to see a second example of the ~~ phenomenon quite so quickly!|
When you type a line of ~~ it makes the page go into panoramic widescreen. It might look okay on your browser but it causes some folks lots of problems reading the text.
If you are still in the magic hour, can you please edit your post to take out the line of ~~~?
That way the screen will revert to normal. Thanks.
|Jun-23-11|| ||gofer: Black can create a simple back-rank mate threat, which wins a piece.|
<19 ... Bxd5>
White cannot accept as this only loses more material. At the moment
white has lost a knight. If white takes back white loses a rook. If
white takes back again it loses the rook and the match.
20 Rxd5 Qxd5! 21 cxd5 Rxc1+ transposes into the line below...
<20 cxd5 Qxc1>
<21 Rxc1 Rxc1+>
<22 Ne1 Ba5>
<23 Kf1 Rxe1+>
<24 Qxe1 Bxe1>
So, black is a rook up. As white cannot go down this route, white
cannot re-capture on move 20 and so black wins a knight for nothing
and probably white resigns immediately. Time to check...
|Jun-23-11|| ||cocker: Any Brits might like to know that this position was featured a month ago in Jon Speelman's excellent column in the Independent. Makes it even easier when you've seen it before!|
|Jun-23-11|| ||NewLine: Of course Gelfand didn't miss it! It is impossible! He is the WC challenger! |
He just thought the rules of the tournament said "Only draws allowed!", - OR - He didn't want to break the incredible draw streak.
|Jun-23-11|| ||CHESSTTCAMPS: In this early middlegame position, black has two bishops in exchange for two knights and is temporarily down a pawn. The critical positional considerations are the doubling of majors on the semi-open c-file and back rank weakness for white. It seeems that white has walked into a simple trap by capturing at f5 and leaving the Nd5 insufficiently protected. Black spurns the automatic recapture at f5 and plays |
19... Bxd5 and now
A) 20.Rxd5 Qxd5 21.cd Rxc1+ 22.Ne1 Ba5 wins
B) 20.cd5 Qxc1 transposes to A.
Was Kamsky under the weather? Time for review.
|Jun-23-11|| ||eblunt: Got this one in a couple of minutes, having spotted the themes - back rank weakness, pinned pieces etc - or so I though until I saw 1/2-1/2 at the top as the result. |
At this point I cursed and went back to the puzzle again and wasted 20 minutes trying to find out where I had gone wrong. Thank you Boris for wasting my time like that :)
|Jun-23-11|| ||CHESSTTCAMPS: I can't top <Phony Benoni's> comment, but it's nice to know that these world class players weren't cheating by using computers.|
|Jun-23-11|| ||nisarg1: <Morozov: 4/4. First time ever!>
|Jun-23-11|| ||jussu: Real mystery that neither of them saw it...|
|Jun-23-11|| ||DarthStapler: I got the first 4 moves but after Kf1 I didn't see that black could win material by exchanging|
|Jun-23-11|| ||goodevans: <fightclub: ... although, black is a rook up and should win, I couldn't find any forcing lines.>|
Welcome aboard, <fightclub>.
cg.com's puzzles are different to a lot of sites. They don't always end in mate. Sometimes the objective is to save a seemingly lost position or simply to find a move that gives a bit of advantage. The only thing they have in common is the object is always to find the "best" move, although every now and then they even throw in a "spoiler" to trap us.
In case you hadn't already noticed, the puzzles get harder as the week progresses and by the weekend we often get positions where the lines are not forcing and the analysis is really complex. I, for one, am always delighted on the odd occasion when I solve one of these!
The analysis you gave with today's puzzle shows you solved it completely even if you found one of the lines a bit frustrating.
< Earlier Kibitzing · PAGE 2 OF 5 ·