|Sep-23-07|| ||not yet a patzer: First? Where is everyone?|
|Sep-23-07|| ||dzechiel: White to move. "Insane" <sigh>. Material is even. White has control of the e-file with his doubled rooks, but his knight on g3 is under attack. The black king is much more exposed than the white king (which is nestled snugly behind his pawns).|
Looking for candiate moves I see such things as:
- 19 Nf5+
- 19 Nh5+
- 19 Bh6+
- 19 Qxf6+
- 19 Bg5
- 19 Ne4
- 19 Nxf7
- 19 Nxg6
Well, of all these moves, I like 19 Bg5 and 19 Nxf7 the most, but I don't see anything concrete.
One line could go something like
19 Bg5 hxg3 20 Ng4 gxh2+ 21 Kh1 Nbd7 22 Bxf6 Nxf6
But this is going nowhere. I'm throwing in the towel tonight, time to check.
|Sep-23-07|| ||TrueBlue: Bg5 is insane, I wanted Nf5+|
|Sep-23-07|| ||MostlyAverageJoe: I looked at the position for the allocated 5 minutes, figured out that IF there is anything to do here then it would start with Bg5, found no convincing continuation, and gave up. OTB, I'd play Nf1.|
Time to start the (silicon) engines ...
|Sep-23-07|| ||Takya Kotov: Tough puzzle, I would not have got it. Like <MostlyAverageJoe> I would have played Nf1. However, I don't think it should be rated "Insane". Once you see the plan, there is a certain sensible logic behind it.|
I always think "Insane" means that even both players don't know what the hell is going on. A really good player probably would have spotted this.
|Sep-23-07|| ||patzerboy: At first I thought that 19.Nxf7 followed by 20.Bg5 worked as well as the other way around, but if 19.Nxf7, then|
21.Re6 Bf8 and Black seems to have the resources to defend.
19.Bg5 is more accurate because Black cannot play 19...Nd7, since he would lose a piece to Nxd7. Anticipating the sac with 19...Rf8 would simply lose the h-pawn while leaving White with tremendous pressure. And 19...BxN is no good after 20.dxe5, so 19...hxg3 is as good as he has. Get what he can and hope White messes up.
Interesting position. I don't think I've ever seen this opening variation before, either.
|Sep-23-07|| ||Some call me Tim: I'm with all of you and specifically with <true blue> in looking at Nf5+ as the best move. I saw Bg5 and like others saw nothing concrete because I did not see the brilliant Nxf7. I find this an amazing combination. In a blitz game I'd play Nf5.|
|Sep-23-07|| ||MostlyAverageJoe: Indeed, insane. The best defense for black seems to be:|
19. Bg5 hxg3 20. Nxf7 Kxf7 21. Re6 gxf2+ 22. Kxf2 Rf8 23. Rxf6+ Qxf6 24. Bxf6 cxb3 25. axb3 Bb4 26. c3 Rxc3 27. Qf4 Rcc8 28. Rf1
which evaluates at (+2.26). So, while Bg5 is winning, it is not as soundly winning as it was played in the game.
I am not very sure that the above line is really optimal, The breadth of searched combination is so huge, that my 1.5GB of cache gets recycled very quickly and does not retain all the information it needs to avoid repeated dives into dead lines after backtracking (by this, I mean that, e.g., after descending into alternate variant and backtracking, I need to descend into the primary line again to recompute its valuation).
It would be too time consuming to do reliable analysis (and I'd need to buy more RAM first). Enough. Just a couple comments on proposed moves.
<dzechiel> Your line loses for the white:
19 Bg5 hxg3 <+2.26>
20 Ng4?? <-3.31> gxh2+
21 Kh1 Nbd7
22 Bxf6? <-6.18>
White would likely play Nxf6 above - you need to save B for a pin.
<TrueBlue:> I also thought for a while about Nf5+, but could not find a way to get any threats on the opened g-file. Hiarcs tells me that the resulting lines are drawish (or slightly better for the black), and require very accurate game from the white (and Bg5 has to be the second move). Again, I am not 100% sure this is a correct assesment. Side lines are even less tractable than the mainline.
<patzerboy> In your line, the last move, Bf8 is inaccurate. 21...Rc6 probably wins for the black. The entire line:
19. Nxf7 Kxf7 20. Bg5 Nd7 21. Re6 Rc6 22. Bxf6 Nxf6 23. Rxf6+ Qxf6 24. Qxd5+ Kg7 25. Qxc6 hxg3 26. Qd7+ Kh6 27. hxg3 Rf8 (-1.87)
<Some call me Tim> 100% agree. Nxf7 is a beauty.
|Sep-23-07|| ||SwitchingQuylthulg: Got this one. At first I liked the move order 19.Nxf7 and 20.Bg5 but then noticed it didn't work. Then I remembered that 19.Bg5 immediately was also a sac and thus an acceptable CG.com puzzle answer :)|
|Sep-23-07|| ||vibes43: I went with the brilliant Nxf7 - but on move 19.|
|Sep-23-07|| ||thorndeux: I mainly considered 19.♘f5+, 19.♘xf7 and 19.♗g5. Like many, I looked at 19.♘xf7 first, and after 19...♔xf7 20.♗g5.|
But somehow I still decided 19.♗g5 hxg3 20.♘xf7 ♔xf7 21.♖e6 was the best try, although I expected gxh2+ (at some point) rather than gxf2+.
I saw that 21...♘bd7 leaves the bishop uncovered but I have to admit that I didn't calculate any further. Still I'm pretty happy I got the ideas right...
|Sep-23-07|| ||znprdx: znprdx: White had this too easy – the only thing <insane> about this is asking us to believe he saw the 5 ply Re6 in the text line since it clearly was not forced and which I'd rejected at the time I analyzed my first choice of Ne4 - just assuming that Bxd5[N] would hold. Perhaps it doesn’t when d4 recaptures – but then is it really all over? At least now N[b]d7 would be of some purpose. Of course the shock value of leaving a piece en prise can be quite unnerving. As mentioned by others Nf5+ as the key would be so pretty and usually works in similar positions. OTB I’d choose the less spectacular Nf1>e3>g4.|
|Sep-23-07|| ||King mega: OMG! i dont notice the knight at g3 is threatened!|
|Sep-23-07|| ||xKinGKooLx: I've been feeling sick this weekend, so I didn't bother to push myself and look too hard. I looked at this puzzle for a while, but I didn't see anything. If I looked longer and harder I *might* have seen the solution. I don't really care too much, since it's very unlikely I would have seen this OTB. Anyway, it's Monday tomorrow, which means an easy puzzle!|
|Sep-23-07|| ||whiteshark: I've been looking for a q-sac, when I noticed that today is only sunday... :(|
|Sep-23-07|| ||fm avari viraf: Definitely, a hard one but after a little thought one can assess the position & can conclude that White is far better than Black. But how to exploit, whether with the tempting 19.Nf5+ or 19.Bg5. The former leads to nowhere but the latter puts more pressure on f6. Hence, I would like to go for my 19.Bg5 hxg3 20.Nxf7! this allows the White Rooks to penetrate through e6 exerting more pressure 20...Kxf7 21.Re6 Nbd7 22.Rxd6 gxf2+ [ or...gxh2+ 23.Kh1 ] 23.Qxf2 Re8 then 24.Rxd7+ Qxd7 25.Qxf6+ & wins. If 23...Rf8 then 24.Ree6 |
|Sep-23-07|| ||patzer2: White solves today's extremely difficult puzzle with 19. Bg5!!, initiating a deep pinning combination. As noted by <Patzerboy>, and explained in more detail by <MostlyAverageJoe>, attempting to transpose with 19. Nxf7?? (expecting 19...Kxf7 20. Bg5 hxg3??) fails to 19...Kxf7 20. Bg5 Nd7! to when Black turns the tables and steals the win from White.|
In the followup, 21. Re6! begins the process of overloading on the pinned knight at f6, forcing the obstruction 21...Nbd7 (breaking the Queen's communication or protection of the Bishop) and allowing 22. Rxd6 (gaining back one of the two temporarily sacrficed pieces and setting up a second decisive pin on the Knight at d7). Finally, 24. Ree6! completes the overloading process, leaving Black hopeless lost after 25. Rxf6+ .
|Sep-23-07|| ||willyfly: Material is even and both of White's ♘s are under attack. At the insane level I know I haven't got a chance. Most of what I've explored in this position has been close but no cigar. So here is a possible line just to guess at the general idea|
19 ♗g5 ♖e8 20 ♘xf7 ♔xf7 21 ♖xe8 ♕d7 22 ♕xf6#
and now to see if I was even close.
got the first move -that's good enough for a Sunday
|Sep-23-07|| ||pawnofdoom: Is all this play forced? That was insanely diffucult, and there was no way I could have gotten it. I gave up after 10 minutes. There were so many possible sacs, but I coudln't calculate any wins from them. Because Bg5 indirectly sacs a knight|
|Sep-23-07|| ||Marmot PFL: Quite a tough combination, the idea of which I found but found many of the calculations too tough. In some lines I was worried about gf2+ followed by Ne4 or Ng4 but the pin after Nxf7! solves that and controls e6 as well. very strange opening play by black aagainst the usually harmless b3 Sicilian. Giving up castling without need was bad judgemant.|
|Sep-25-07|| ||kevin86: It looks like the keystone to this problem was the knight at f6. White sacrifices much material to effect an over-attack on the poor resident of f6-the black stallion.|
|Oct-22-07|| ||DoctorD: Pogats in later life was well-known as a composer of long helpmates and helpstalemates, with a book in 1995: Hundert Hilfsmatt-und Hilfspattmehrzu'ger.|
|Jul-12-09|| ||Amarande: Let's see ...
Black's King is somewhat exposed, and the Knight at f6 is vulnerable. In fact, it's effectively pinned, since if it moves in the present situation, Qxf7 would be mate.
At first I did consider 19 Bg5, as this pin theme indicates (it not only increases the pressure but also pins the Knight a second way), but discarded it as I did not see the strength of the followup Nxf7. Thus, I thought the move to be inadequate, not only because the Knight at g3 was attacked but also because I saw what appeared to be the reasonable defensive try 19 ... Be7. (In fact, 19 ... Be7 is refuted also by 20 Nxf7, whereupon follows 20 ... Kxf7 21 Rxe7+ Qxe7 [forced, otherwise Qxf6 gives a speedy mate] 22 Rxe7+ Kxe7 23 Bxf6+, and White wins.)
OTB, thus, I would probably have played 19 Nf5+ to open new attacking lines, as 19 Nf1 or 19 Nh1 appear far too pusillanimous for such an aggressive position.
|Jul-12-09|| ||zenpharaohs: This computer analysis will take a long time, but so far Rybka 3 likes|
19 Bg5 Bb4
20 c3 hxg3
but I can tell there are lots and lots of lines that are not ruled out yet.
|Jul-12-09|| ||zenpharaohs: After much crunching, Rybka 3 ends up with:
19 Bg5 hxg3
20 Nxf7 Kxf7
After that, it's still thinking.