|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
| Aug-27-16 | | patzer2: Today's Saturday solution 21...Nd3! recognizes the most pressing positional issue for Black is the elimination of White's dangerous passed c-pawn. My flawed solution 21...Bxh4 22. Qxe5 Bf6 allows White to exploit the passed pawn with 23. Qf4 as play might continue 23...Qxa2 24. c6 a5 25. c7 Rc8 26. Qd6 Qd5 27. Qxd5 exd5 28. Ra7 d4 29. Bf4 g5 30. Bd6 Rfe8 31. Rxa5 Kg7 32. Kg2 Be7 33. Re5 Kf6 34. Re4 Bxd6 35. Rxe8 Rxe8 36. c8=Q Rxc8 37. Rxc8  (+1.92 @ 20 depth, Deep Fritz 15). Black's game goes down hill after 20. Rb7? allowing the pawn capture 20...Nxe5  . Much stronger is protecting the e-pawn with 20. Qc7  (0.87 @ 29 depth, Stockfish 7). |
|
| Aug-27-16 | | patzer2: If, instead of 27. Ng7, White tries 27. Kf4 (diagram below)
click for larger viewBlack has a mating combination starting with 27...g5+ (diagram below): 
click for larger viewFrom here (diagram above) play might go 28. Ke5 f6+ 29. Kxe6 Qxe4# (diagram below) 
click for larger view |
|
Aug-27-16
 | | al wazir: I was clueless. I didn't see the strength of 21...Nd3, and when I first played through the game I couldn't believe that black's moves were best or white's replies were forced. But after going over it a couple of times I became convinced. |
|
| Aug-27-16 | | Steve.Patzer: I guessed what the first move would be, but had no clue how to continue. |
|
| Aug-27-16 | | lost in space: as <al wazir>. |
|
| Aug-27-16 | | AlicesKnight: Likewise. It's a 'won game' but drawn out and black must safeguard his back rank. |
|
| Aug-27-16 | | agb2002: The material is identical.
White threatens Qxe5, Rxe7 and c6-Rxa7.
Black has several moves:
-21... Nd3, to save the knight while forking rook and pawn. -21... Bf6, to protect both pieces.
-21... Bxh4, to solve the threats with tempo.
-21... Ng6 loses to 22.Nxg6.
-21... Rd7 seems to hand the initiative over to White after 22.Rxd7 Nxd7 23.c6. -----
In the case of 21... Nd3:
A) 22.Rd1 Nxf2 wins decisive material.
B) 22.Rcb1 Nxc5 23.Bxc5 Bxc5 wins a pawn.
C) 22.Rb3 Qa3
C.1) 23.Rcb1 Nxc5 wins a pawn.
C.2) 23.Rc2 Nxc5 24.Bxc5 Bxc5 25.Rxc5 Rc1+ 26.Kg2 Qf1+ 27.Kf3 Qh1+ C.2.a) 28.Ng2 Rg1 seems to recover thed piece with advantage. C.2.b) 28.Ke3 Re1+ wins decisive material.
C.2.c) 28.Kf4 g5+ 29.Ke5 (29.Kxg5 Qxe4 wins) 29... f6+ wins. -----
I don't have time for more today. |
|
| Aug-27-16 | | gofer: I decided that things weren't going to work out for black on the kingside and really the only weakness for white was Rc1.
So in the end I chose the first move for both black and white... <21 ... Nd3>
<22 Rb3 ...>
At that point, I was lazy and a little uncertain that I was even right this far, so I went for a combination that might be
winning, but also might be losing...
<22 ... Qxb3>
<23 axb3 Nxc1>
<24 Bxc1 Rd1+>
<25 Kg2 Rxc1>
click for larger viewOkay, this isn't so great, but its not a complete disaster, perhaps the two rooks can combine quickly enough to sweep off the b and c pawns,
but perhaps they can't!
I certainly didn't see <22 ... Qa6> with the awsesome follow up <25 ... Qf1+>. Another weekend puzzle where the combination was beyond my vision. |
|
| Aug-27-16 | | mel gibson: Why did White resign?
According to DR4 64 bit whites next move is
29. Ke3 score -1.36 depth 17.
That's only really a pawn down! |
|
| Aug-27-16 | | Honey Blend: <mel gibson: Why did White resign?
According to DR4 64 bit whites next move is
29. Ke3 score -1.36 depth 17.
That's only really a pawn down!>
After 29. ♔e3 ♖xg2 30. ♖b1 ♕h2 31. ♖c2 ♖g1 32. ♖xg1 ♕xg1: 
click for larger viewThat's two pawns down and it looks like after an inevitable trading off of ♕'s it's a ♖ and ♙ ending with White on the ropes. |
|
| Aug-27-16 | | mel gibson: <White on the ropes.> Yes - but you don't resign because you're a pawn down - or even 2 pawns -
human error can easily make up for 2 pawns down. |
|
| Aug-27-16 | | RandomVisitor: <patzer2> mentions 20.Qc7
click for larger viewKomodo-10.1-64bit:
<+0.85/32 20...Qa4 21.c6 Bxh4 22.gxh4> Nb6 23.Bxb6 axb6 24.Qxb6 Rc8 25.c7 Qxa2 26.Qb7 Qa4 27.Qb4 Qxb4 28.Rxb4 f5 29.exf6 gxf6 30.Rxe4 Kf7 31.Kg2 Rfe8 32.Rec4 Re7 33.Kf3 Rd7 34.Ke4 Ke7 35.Ke3 Kd6 36.Rc6+ Ke7 37.f4 Rd6 38.R6c2 Rd7 39.Ke4 Kd6 40.Rc6+ Ke7 41.R6c3 Rd2 42.Rg1 Re2+ 43.Kf3 Rxh2 +0.91/32 20...g5 21.Ng2 Rfe8 22.Rd1 Qxa2 23.Rbc1 Qa4 24.Ne1 Nf8 25.Rxd8 Bxd8 26.Qb7 Re7 27.Qa8 Rd7 28.c6 Rd1 29.Rxd1 Qxd1 30.Kg2 h6 31.Qb8 Qd5 32.c7 Bxc7 33.Qxc7 a5 34.Nc2 Ng6 35.Qb8+ Kg7 36.h3 Kh7 37.Nd4 Nxe5 38.Qf8 Qb7 39.Nc2 Nf3 40.h4 e5 41.hxg5 hxg5 42.Bc5 a4 43.Ne3 Qd7 44.Be7 +1.46/32 20...Qxa2 21.c6 Bxh4 22.cxd7 Be7 23.Qc6 h6 24.Rd1 Qa5 25.Qxe4 Bc5 26.Rb7 Bxe3 27.Qxe3 a6 28.Qd4 f6 29.Qc4 fxe5 30.Qxe6+ Kh7 31.Qc4 Qa3 32.Qc2+ Kh8 33.Rb6 Qe7 34.Rbd6 e4 35.Qd2 Rf7 36.Qd4 Rf6 37.Rd5 Rf7 38.h3 Kg8 39.Rd6 a5 40.Rd2 a4 41.Qxa4 e3 |
|
| Aug-27-16 | | RandomVisitor: <mel gibson>Perhaps play would continue with white exchanging queens and obtaining a passed pawn - perhaps a reason to play on - but maybe not
click for larger viewKomodo-10.1-64bit:
<-1.59/26 29.Ke3 Rxg2 30.Rb1 Qxh2 31.Qh4 Rxf2 32.Qxh2 Rxh2 33.Ra5 Rc2 34.Rb7 Rc3+ 35.Kf2 a6 36.Rxa6> g5 37.Raa7 Kg7 38.g4 h6 39.a4 Kg6 40.a5 Ra3 41.a6 f5 42.gxf5+ Rxf5+ 43.Ke2 Re5+ 44.Kd2 Kf5 45.Rb6 h5 46.Rh7 Rd5+ 47.Ke1 h4 -1.96/26 29.Qc2 Qxg2+ 30.Ke2 h6 31.Rb1 Rxb1 32.Qxb1 Rd8 33.Rb5 Qxh2 34.Rb8 Qh5+ 35.Kf1 Qh1+ 36.Ke2 Qd5 37.Rxd8+ Qxd8 38.Qe4 Qd7 39.a4 g6 40.Qb4 Kh7 41.Kf1 Qd1+ 42.Kg2 Qd5+ 43.f3 Qd1 44.Kf2 Qc2+ 45.Ke3 Qc7 46.Kf2 Qe5 47.a5 Qd5 -2.26/26 29.Rb4 Qxg2+ 30.Ke2 Qf1+ 31.Ke3 h6 32.Qd4 Qh3 33.Qc3 Qxh2 34.Rg4 g6 35.Rc8 Re1+ 36.Qxe1 Rxc8 37.Rh4 Qg2 38.Qh1 Rc3+ 39.Kd4 Qxh1 40.Rxh1 Rf3 41.Rxh6 Rxf2 42.a4 Kg7 43.Rh3 Ra2 44.Kc4 Rxa4+ 45.Kb5 Ra3 46.Kb4 Ra2 47.g4 |
|
| Aug-27-16 | | JimNorCal: Wait. It's Black to move? Dag nab it .... |
|
| Aug-27-16 | | YouRang: I think the key point to this puzzle is how efficiently the knight buys time. After <21...Nd3>
click for larger view
- Black can't take the Be7 because his Rc1 is attacked. - Then after ...Nxc5, black still can't take Be7 because his Q is attacked. - After white finally takes the N with Bxc5, then ...Bxc5, and there's no more Be7 to be taken. Hence, the N attacks extricate black from an otherwise nasty double-attack, in addition to taking away white's great asset: the passed c-pawn. |
|
Aug-27-16
 | | Sally Simpson: I'm with Mel Gibson,
We have all seen (and played) 100's of human blunders in so called 'won' positions.' at all levels. There is no such thing as a won position till either your opponent resigns or you have mated him. All you need to do is plant one seed of doubt to unsettle the player. Sometimes you need not even do that. All players are perfectly capable of screwing up a game. Filling up the thread with Komodo vomit is meaningless. It does not have a clue what a seed of doubt is. Here is where White resigned:
click for larger viewIt was his game. White can resign whenever he wants. He has finished with it. We can take over. How about.
29.Ke3 Rxg2 30.Rb1 Qxh2 31.Qf3 Cannot see that shot mentioned and yet: 
click for larger viewIt threatens to win the Queen with Rh5.
Don't look at what a computer would play here. What is a human response. 31...g6
We have all seen a lot worse moves in a 'won position' from humans. You have to remember it is a human v human game. White needs a blunder to back into the game keep plugging away till all hope is gone. 32.Rc7
click for larger viewAnd the game is far from the resigning stage. Suddenly the Rook on g2 is looking a bit silly. Black has to get his hands dirty. White has Rb8 threats. (seeds of doubt.) 32...Qh6+ 33.Ke2 Qg7
click for larger viewGood move. Black has woken up.
If now 34.Qxg2 Qe5+ and 35...Qxc7.
You can run a Komodo over every 'won game' that has ever been thrown away
and it can show the win. But it could never anticipate the human blunder that turned the win into a loss. And yet it happened....the things are useless for looking at what 'may' happen in a human v human game. |
|
| Aug-27-16 | | mike1: hmmm. If white replaces 23. Rcc3 with Rc4 then Black's idea does not work. |
|
| Aug-27-16 | | YouRang: I kinda agree with the sentiment that white resigned prematurely. It's complex enough to hope for a mistake by black. Of course, white could also make the mistake (maybe resigning *was* the mistake).
click for larger view
You might at least hope that <29.Ke3> is met by 29...Qxg2?, which actually leaves white a bit better. Even after the correct <29...Rxg2 30.Qf3>, black has to extricate his rook and queen.
click for larger view
Humanly, one might expect <30...Qe1+ 31.Qe2>
click for larger view
This practically forces 31...Qg1. If instead, black plays 31...Qxe2? 32.Kxe2 Rxh2 33.a4
click for larger view
Black is up a couple pawns, but at the moment, white's rooks are much more active and they will be able to take black's a-pawn and make trouble with the passed a-pawn before black can make trouble with his extra pawns. Assuming black goes with the best move, <31...Qg1>
click for larger view
Now <32.Rbc3 Qxh2 33.Qf3>
click for larger view
Black must now find the difficult 33...e5! to maintain his advantage. If instead (say) <33...Qh6+? then 34.Ke2 Rh2 35.Rc8 Qh5 36.Rxf8+ Kxf8 37.Rc8+ Ke7 38.Rc7+>
click for larger view
White will win black's a-pawn and have a nice passer to work with, and the game is decidedly undecided. |
|
| Aug-27-16 | | YouRang: <mike1: hmmm. If white replaces 23. Rcc3 with Rc4 then Black's idea does not work.> After 23.Rc4
click for larger view
It's a good try, but black does have some replies that work. The best is 23...Qxa2! 24.Rxd3 Qb1+ 25.Kg2 Qxd3 wins the exchange. Next best is 23...f5, which more complicated, but at worst it also ends up winning the exchange. |
|
| Aug-27-16 | | mel gibson: <Aug-27-16
Premium Chessgames Member Sally Simpson: I'm with Mel Gibson,We have all seen (and played) 100's of human blunders in so called 'won' positions.' at all levels. There is no such thing as a won position till either your opponent resigns or you have mated him. All you need to do is plant one seed of doubt to unsettle the player. Sometimes you need not even do that. All players are perfectly capable of screwing up a game. Filling up the thread with Komodo vomit is meaningless. It does not have a clue what a seed of doubt is.> Thanks Sally -
you never give up if you're one pawn down - it's so silly. |
|
Aug-28-16
| | Sally Simpson: Hi YouRang,
First the header. '? (2015) '
The game was played in round 7 of the American Continental Classical Championship 2015. At the time both players were tied 2nd= on 5pts along with four others. http://chess-results.com/tnr169788.... All we have is the bare score. No notes, no clock times, nothing. For all we know White may have continued but the score was not electronically recorded. GM resignations are sometimes baffling till you look just a bit deeper and it's understandable. But not with a computer. They cannot look for counter play or moves that may induce a blunder. In most cases they tell us via a number what we already know. With best play Black wins and we also know winning a won game can be tough, we also know positions humans are apt to blunder in and most of the reasons why won games get blown.(the winning player tries too soon to clarify the position and walks unaided into a trick, or they get caught napping and relax. Computers cannot predict this.) But here...It does appear all White saw was Black threat of Qxg2+ which wins the Queen or the c5 Rook.
click for larger view1...Qxg2+ 2.Kf4 (any other moves drop the White Queens) 2...Qxf2+ and Qxc5. But wretched as it appears White can muddy the waters and if Black clears them and makes the win beyond doubt, then resign. My yardstick is, could I beat a world champion if I had my opponent's position. If the answer is yes then why am I playing on, it's time to resign. But others differ.
Re: Computers and 'may have happened.'
Take this one. Black to play
click for larger viewIf the players agreed a draw here and some chappie wants to verify it with a computer fair enough. Off you go. Chances are it will batter on for 20-30 moves reaching some drawn game. But there is no way it will actually guess at what was played. Even in simple positions like this humans can blunder. White played 50.Qe4 threatening mate on h7.
Black missed this one move threat and played 50...Qf7. 
click for larger viewWhite now missed he had mate in one and played 51.Ne6. A comedy of errors. They played on for another 25 moves before, as they use to say, the game was abandoned as a draw. And people think computers can predict human moves. Two very good players. One (Black) was a brilliant chess player. F Bellini vs V Gashimov, 2002 |
|
| Aug-28-16 | | YouRang: Hi <Sally Simpson>. Speaking of human blunders, your last position reminded me of Kramnik's famous flub vs Deep Fritz:
click for larger view
Kramnik played <34...Qe3??> The computer didn't overlook the mate in one.
Deep Fritz vs Kramnik, 2006 |
|
Aug-28-16
| | Sally Simpson: Hi You Rang,
I don't think the computer expected it either.
Must be something about the Nf8 - Qh7 mating pattern. It's a human blindspot. Ala poor Kramnik. in 50 years time they will have forgotten what a great players he was and only this will remain. By then of course computers will have cracked chess and started re-designing humans. Man invents chess.
Man becomes good at chess.
Man invents computer.
Computer becomes good at chess and can beat man.
Computer invents man.. |
|
| Aug-29-16 | | Honey Blend: <you never give up if you're one pawn down> Human problems - can we really quantify the limits of when to resign, in contrast to a chess engine that puts a number to every possible chess position? Maybe not in the next century or so, so apparently we should spew a whole list of rules of thumb on when to and when not to give up, like, "Never resign when you're down the exchange because there's still hope." Or better yet, "It's fun to play ♕ vs ♖ games to the bitter end, either merely to call their endgame knowledge a bluff or hope they blunder their ♕ to a skewer." |
|
Aug-29-16
| | Sally Simpson: Hi Honey,
The most common method of messing up a KQ v KR ending is when the KQ player is moving in for the kill and gets caught in a stalemate trick. Anton Shomoev - Dmitry Jakovenko Russian Cup, 2004. http://2700chess.com/games/jakovenk... White to play.
click for larger view108.Rb6+
Black spots the stalemate but instead of 108....Kc7 Black played 108....Kc5 109.Rc6+ ½-½.
click for larger viewThere maybe a wee story behind this. White had been trying since move 80 to win. Some may consider this embarrassing for a GM but it's not easy. After demonstrating this is not his favourite ending and rather than plod on with 106...Kc7 he walked into the stalemate. From the game this position arose with Black to play. 
click for larger viewHas a move than can pick up the Rook in a handful of moves no matter what White plays. I won't tease (and save electricity from the engineers) it is 98...Qf7. 
click for larger viewIt is good practise busting all the tries White has to avoid a loss of the Rook. Now you have read this (who ever you are) book up on this ending because fate decrees (Morphy's Law) you will be involved in one very soon. |
|
 |
|
< Earlier Kibitzing · PAGE 2 OF 2 ·
Later Kibitzing> |
Kibitzing
sac: 23...Nxc5
