< Earlier Kibitzing · PAGE 3 OF 3 ·
|May-23-07|| ||realbrob: <Marvol> No, I'm not British, I just saw the avatar and I chose it because I liked it.|
Of course, if you're playing a real game OTB it's a very big satisfaction to reach a position like that of this game. It means you played much better than your opponent and that you're winning without getting big tactical blunders. What I meant was that today is wednesday, and usually wednesday puzzles involve more decisive moves... In addition, there are many winning moves.
|May-23-07|| ||Bears092: <dzechiel: Needless to say, after 22 Nxh6 Kxh6 23 Qf5 is not check.>|
As noted in my earlier analysis, 23. Qf4+ Kh7 24. Qf5+ is the line I was discussing.
|May-23-07|| ||Gilmoy: I fixated on Nxh6 -- too much Topalov vs Sasikiran, 2007 :) It wins, so I stopped looking. My idea was standard g-deflection. Usually this allows a Q+B mate at h7 -- here White has the other B, so the g7-hole allows a triangle mate at h8 instead. Anyways, gxh6 gets mated (Qf5+ Qf6), Kxh6 soon drops the Rook.|
|May-23-07|| ||Bears092: gxh6 doesn't quite get black mated. He has to give up the queen.....but not the king.|
|May-23-07|| ||Kneesocks: Wow! My first puzzle on Chessgames.com. I was looking for a checkmate, so I didn't get it.|
|May-23-07|| ||Jack Kerouac: After 18 beers (plies) I came up with
|May-23-07|| ||yalie: i saw re1 within seconds and nxd6 after a few more seconds. Devoid of silicon assistance, I decided nxd6 was the cleaner continuation.|
|May-23-07|| ||kevin86: This puzzle looks like an advanced math problem-with several solutions. I saw Nxh6 before the Nd6 text move. My move has the same threat of a fork-and also opens black's king to attack.|
The text is obviously most forcing and simplest. It is a fork leading to another fork.
|May-23-07|| ||YouRang: cg.com threw us a curveball today! After 2 successive days where the solution was Qh4, today it's not even a legal move. They are forcing us to think! :-(|
Fortunately, I didn't have to think long. :-)
With the exposed king and unguarded rook both vulnerable to attack on light squares, the queen move to f5 with double attack quickly became apparent.
Of course, I have a knight there, so it's time for a clearance sac. I debated a while between taking g7 or d6. By taking g7, I bust the king's defense, while taking at d6 gives me a nice passed pawn on the 5th rank.
I opted for d6, since I figured that winning the exchange would probably win the game by securing an endgame advantage; and the passed pawn would be decisive. But I don't think it matters much.
|May-23-07|| ||nimzo knight: For 22. Nxh6 Kxh6 discussion. 23. Qf4+ g5 24 Qf6+ ..mate to follow. move of king on 23 allows the rook to be forked.
Only thing is Kxh6 is not forced and as pointed out earlier 22..b4 is possible. But I still think after 23. axb4 cxb4 24. Bd2. Black might soon loose queen side pawns.
I will like to see what engines say for this.|
|May-23-07|| ||MostlyAverageJoe: <Counterpoint> and <GHOST19: Navara's line is the best just because it causes resignation in only two moves>|
By that reasoning, 1. d3 is best move ever, because it causes resignation in one move: Pillsbury vs Von Bardeleben, 1895.
No, wait 1. e4 is just as good: E Safarli vs T L Petrosian, 2007 and look who resigned! Cannot complain about 1. g3, either: Votava vs P Vavra, 2006.
No, wait, the best strategy is not to move at all, and have the black resign: T Roussel Roozmon vs S Kulic, 2006.
OK, so I rely on computer analysis extensively. But I try to see the underlying reasons why the moves evaluate highly. This is not always successful, as some of my notes in this game show, but I don't follow the advice blindly. The positions where the computer indicates a hard choice is to be made are usually quite educational.
|May-23-07|| ||newton296: simple idea really , white is winning and he just needs to pick up the loose rook at c8 with a Q check at f5. I saw this and realized i could sak the knight. either nxh6! or the more forcing nxd6! ; Nxd6 speaks for itself . What about nxh6!? 1st) If ... Kxh6 and black loses to Qf4+ ...g5? Qf6+ ...Kh7 Q f5 + and white wins the rook. If ...GxH6 Qf5+ wins the rook. Best defense against nxh6 is probably ...Qd7 protecting the rook and now Q d3+ ...kh8 and it seems white's attack is out of steam !? probably improvements for white in this line though>|
|May-23-07|| ||newton296: nxg7 seems the same as nxd6|
|May-23-07|| ||Justawoodpusher: Looked at Re1, Nxg7 and Nxd6. Decided on Nxd6 because it is most straitforeward to calculate.|
Even it may be not the strongest move, OTB it's probably the best choice as one can be reasonably sure that it works. Probably that's why it was choosen by Navara.
|May-23-07|| ||newton296: best play after nxh6 seems to be ...Qd7!? than Qe4+ ...kh8 ( not... kxh6 due too bd2+ leads to mate) Qh4 and white's knight is safe , thx to pin of g7 pawn to king, plus white is up a pawn and has a discovered check with next move to win more material!|
|May-23-07|| ||MiCrooks: The idea that Re1 is a better move is a bit far fetched. Basically if you take any of the three pawns with the Knight you are winning with the d6 and g7 pawns being almost equal, but the d6 pawn is absolutely clear.|
To play Re1 you have to see all the way to the end of a long combination to realize that the position with QN vs RNB is easily won.
Yes, it has a high evaluation as you will eventually pick up one of the other pieces but the other is so clear-cut that you'd be a fool not to play it over the board. Evaluation +999.99 because your opponent will likey resign immediately, as happened in the game. Down a pawn, the exchange, and position no reason to play on.
|May-23-07|| ||fm avari viraf: White's text move 22.Nxd6 is strong & precise as it wins the exchange & the vital d6 pawn. Any other Knight move looks 2nd best.|
|May-23-07|| ||T Ciddasselepoh: This one was way too easy.|
|May-23-07|| ||WarmasterKron: Nowhere near. I tried too hard to find something clever, but 22.Nxd6 is simple and effective. It may not objectively be the strongest move if one goes by computer evaluation, but a win is a win, right? Why make it harder than you need to?|
|May-23-07|| ||MostlyAverageJoe: I see I contributed to quite a controversy here :-)
First, I agree that Nxd6 is definitely easiest to find and analyze. No matter what, white wins an exchange and a pawn. I have no doubt I could find this OTB (and most likely would play it). It is also attractive in that it offers a chance to completely disassemble black's queenside, except after Qd7 defense (I guess this was the black's reason for resigning after seeing the imminent loss of the rook and a possible slaughter of the pawn structure). I am curious whether the proponents of Nxd6 have considered the complications introduced by the Qd7 response ...
As in my very first post, done without computer analysis, I would still argue that Nxg7 is better; the Qd7 defense falls flat in this line (consider Nxg7 Qd7 Nh5!). The black has really no choice other than Bxg7, and now Qf5+ is just as forceful as in the Nxd6 line, without relying on the minor blunder Bxd6 (see also the analysis by <Counterpoint>).
Finally, Re1 (mistyped as Ne1 in my first post) seems to offer immediate win of a minor piece, also without having to analyze too deeply.
Bottom line: any of the 4 top-rated moves is playable and, IMO, about equivalent as a solution for this puzzle. I am not going here to argue superiority of Re1 just on the basis of 1.5 pt advantage after more analysis than anyone would have time to do OTB.
It is interesting to note that computer analysis finds Nxd6 veru easily, drops its value after deeper consideration, switches to Nxg7, then, much later, settles on Re1 (see my forum for supporting details). This is very much like the process I followed myself (which is why I decided to spend more than usual time on the analysis - about 8 CPU hours worth of it, to be exact).
|May-23-07|| ||gBizzle: i thought Bxg7 followed by Nxd6 but this is better|
|May-23-07|| ||playground player: I saw Nxd6, then Re1, and then settled for Nxh6. Unfortunately, Kxh6 makes this attack problematic--but gxh6 means curtains for Black (23. Qf5+, Kg8 24. Qf6, and there's no way for Black to escape checkmate). Since Black in these puzzles always does the right thing, Nxd6 is the way to go.|
|May-23-07|| ||Crowaholic: I considered 22. Re1 with the probable response ..Qd7, but had trouble finding truly convincing follow-up moves.|
Then 22. Nxd6 occurred to me and I figured that again, Qd7 is the best response because of the obvious queen fork in case of 22. ..Bxd6?.
I then figured that White is winning fair and square after 23. Nxc8 Qxc8 24. Qd3+ f5 25. g4 and decided to leave it at that.
And then I looked at the solution and couldn't believe that Bitansky had missed the fork...
|May-23-07|| ||Sololoy: I analyzed Nxh6 but black is not forced to take the Knigth, they could play Qd7 and I don't see a clear way to win for White...|
|May-24-07|| ||crwynn: I got 22.Re1 with the idea of 23.Bg7 winning a pawn (or the queen) - didn't even see the check picking up c8. But oddly enough Re1 is best.|
Also the queen trap 22.Re1 Qd7 23.Bg7 Bg7 24.Re7 is really neat...much more fun than crudely grabbing a rook.
< Earlier Kibitzing · PAGE 3 OF 3 ·