< Earlier Kibitzing · PAGE 3 OF 3 ·
|Aug-09-07|| ||Creg: <Sneaky> Excellent post! I have never liked seeing posts of computer analysis. What does it achieve? It's no different than imagining being on a math website where everyone has to figure out today's math problem. Then someone comes along and says that my calculator solved the problem. Does one become better at math this way? Heck no!|
Chess programs are glorified calculators as the game is broken down mathematically, so it's nothing more than numbers to the computer.
I have nothing against <Random Visitor> either, his posts appear sincere, but I feel he, along with many others out there have become, in a way but not entirely by definition, addicted to the chess engine.
They do not realize that this is not how humans learn, and that it does not explain things in a way the human mind works. It's sort of like eating the wrong foods, say cake instead of vegetables. It fills you up temporarily but without the nutrients it is unable to sustain you for a longer period of time.
|Aug-09-07|| ||Jack Kerouac: That rabbit's dynamite!|
|Aug-09-07|| ||EyesofBlue: Stop the world, I want to get off. I got one right?!|
|Aug-09-07|| ||patzer2: For today's puzzle solution, the demolition of pawn structure with 18. Nf6+!! destroys any illusion Black may have had about the safety of his King's weakened castled position.|
The 10-move mating variation given by <Random Visitor>'s Rybka is entertaining and worthy of study.
|Aug-09-07|| ||MostlyAverageJoe: <thorndeux: <mkrk17><MostlyAverageJoe> Doesn't moving the king instead of capturing the knight simply lose the exchange?>|
Of course it does. But it does not give white a forced mate in 10, as is the case in the capture.
<Sneaky: Now when it comes to 18.Nf6+ Kf8 and 18.Nf6+ Kh8, I don't see any point in looking at these lines any further. When I saw the puzzle I summarized these variations very quickly, thinking this: "If Black has nothing better to do then move his king and surrender the exchange, then the combination is valid." >
OK, I buy that, and hereby retract my previous opinion that presence of much better defense than what was played in the game <reduces the appeal of today's puzzle>.
|Aug-09-07|| ||MostlyAverageJoe: <Creg: I have never liked seeing posts of computer analysis. What does it achieve? It's no different than imagining being on a math website where everyone has to figure out today's math problem. >|
Do you also dislike knitting frames ?-)
Are you familiar with the 4-color theorem in graph theory, one of the most famous/celebrated/difficult problems in math? Guess what - the proof involved HEAVY computer use.
Having been involved in software for about 30 years now, I find good software to be as appealing to my aesthetic sensibilities as best art.
BTW, I always try to interpret the computer-generated lines in terms of "what would I need to notice to come up with this idea using just the brain". See the analysis of 34. Be3 move in another puzzle, referenced below.
<Sneaky> and <Creg>: Sometimes, interesting insights can be obtained from the lines that are not the best. Here's a recent example: Carlsen vs H Harestad, 2003
|Aug-09-07|| ||psmith: <dhotts>
19...Bb8 20. Bxe6 Bf4+ 21. Qxf4 Rxe6 is winning for White after 22. Rh3 Kf8 23. Rh8+ Ke7 24. Rxc8.
But the move given by many already -- 19...Bb8 20. Rg3+! is superior. It leads to a forced mate.
(By the way, I got this one too, with all bells and whistles.)
|Aug-09-07|| ||RandomVisitor: <Sneaky><Creg>The computer analysis is posted to help humans understand what might be best play in a position. The computer might be wrong. It is up to humans to interpret what are, essentially, heuristics which sometimes result in lines of best play. |
As far as the computer offering "little light" on the position, I think that Sneaky's long post indicates that much insight can be obtained, even if it means that secondary moves are inferior.
|Aug-09-07|| ||Brapp: Well, the knight appears extremely dangerous for move 18. First off, the knight is under attack, provoking a move. Secondly, the Knight has access to THREE squares forking heavy pieces. Have I ever seen such a well positioned knight? or such an unfortunately arrayed defense? Start with checks.|
The royal fork 18. Ne7+ leads nowhere. The rook takes with knight with no repercussions.
Ah, but 18. Nf6 is interesting. Black must forfeit the rook on e8 or else face an unbeatable attack by white's queen and rooks.
An easy Thursday, really, but an excellent demonstration of how to attack a castled king!
|Aug-09-07|| ||Creg: <MostlyAverageJoe> I love a good debate and I too work in the software end of computers. I am as pleased with the final outcome of one of my programs as you are. But <Sneaky> and I refer to the art of learning, not just the answer. |
Oh, I'm certain that technology is helpful in so many ways, and the 4-color theorem in graph theory needing a computer to confirm it's analysis is most likely intriguing. I am always impressed with people who can even conceive of such things, but do we start with the 4-color theorem when teaching math? I don't think so.
Learning/teaching/getting an answer are different things. Since you and I are into writing programs I'm certain you've come across many a beginner in your time. Have you ever tried to explain a simple array to that person? Sure, you can just show them how it's done, but explaining how it works is far more effective.
I know, this is chess and seeing the move may feel like an end to a means. Please note the last paragraph of my previous post. I also have a training background, and I don't believe that knowing the move itself truly helps someone grow.
It's possible you haven't reached a level of chess training that allows you to understand the broader concepts involved that the computer is incapable of showing you, thus you may not fully appreciate what I am saying. This too can take us back to the cake and nutrients in my previous post. How many times has a parent told a 6 year old he must eat his broccoli, when all he wants his cake?!
The 6 year old only knows it fills him up, but he doesn't understand that it will not help him grow.
Just a reminder I see this as a discussion/debate not an angry tirade with one another. <MostlyAverageJoe> if you want we can discuss this outside the forum if you wish.
|Aug-09-07|| ||RandomVisitor: By the way, it might appear to the many that read my computer posts of analysis that I am "addicted" to chess engines. It might surprise you to know that I always try to solve puzzles first before I load up the machine. Sometimes I get the puzzle and sometimes I don't. In any case, I always end up learning something that I didn't know before.|
|Aug-09-07|| ||Crowaholic: After missing yesterday's best move, it's comforting that I saw 18. Nf6+ quickly although it wasn't the first move I considered. Then after 18. ..gxh6 19. Qxh6, I first considered the Bg4 defense with 20. h3! and now White threatens to clear the h file allowing for Qh8#. I then analyzed 19. Bb8 and convinced myself that mate is unstoppable, e.g.:|
20. Rg3+! Bxg3 21. hxg3 Bh3! 22. Qg6+ Kh8 23. Bxf7 Qf5 24. Qh6+ Qh7 25. Qxf6+ Qg7 26. Rxh3#
I then analyzed Black refusing the ♘, didn't see a mate, but the won exchange together with Black's cramped position should be enough to secure the win.
18. ..Kf8 19. Nxe8 Qxe8 20. Bxe6 Qxe6 21. Qc7 - looks anything but pleasant for Black.
18. ..Kf8 19. Nxe8 Bxc4 20. Nd6 Qc7 21. Rf3 Be6 22. Rd1 also not good for Black.
PS: I immediately saw 20. Rg3+ after 19. ..Bb8 (or ..Bxf2 which I missed) because I had already analyzed 19. ..Bg4 20. h3! with the h file clearance threat. So it was obvious that White can just sac the rook on g3 and proceed to mate. The really interesting question is why Black even accepted the ♘ sac. If he did see up to 20. Rg3+, it makes no sense at all. If he didn't, I would at least have expected him to play on until this move. Probably he missed it at first and then it occurred to him as he was playing 18. ..gxf6 or shortly thereafter.
PPS: Black got his due for allowing White 4(!) pieces in the center ready to attack the almost undefended kingside. And the mating theme with doubled f, pawns, hostile queen on h6 and hostile rook ready to attack on the g file is so common, it should ring all alarm bells. This is probably because the position is straightforward to achieve, and because it's so powerful. With the king on g8 and the enemy queen on h6, the king has no flight squares. Vacating f7 often takes too much time because f6 must be cleared first. And finally the doubled f pawns also make it impossible for Black to assist the king along the 6th and 7th rank. So Black is forced to find a defense on the diagonals - in this case, a rook sacrifice nicely refutes that possibility, too, depriving black of all hope.
|Aug-09-07|| ||Crowaholic: <Gilmoy: [A] 18.Nf6+ gxf6
[A1] 19.Qxh6 f6>
19. ..f6 is not a legal move in this line.
<dhotts: 19...Bb8 protecting against Rh3+. The right response is 19.Qxh6 Bb8 20. Bxe6!!>
Black has the clearance sacrifice 20. ..Bf4+! 21. Qxf4 Rxe6 and it seems the king can escape, although I think White is still going to win some material and the endgame.
<Sneaky>: Thanks, that was an exhilarating comment. I also noticed that Rg3+ is present in almost all of the winning lines that RandomVisiter posted, so it wasn't very enlightening. Don't give up on computers so quickly, though. They just do what they were told to do. If they don't give the most useful answers, then maybe they weren't asked useful questions. It would be helpful to make chess engines accept some sort of move constraints, e.g. "Do not play Rg3 for the next five moves". This is all possible - if it hasn't been done before it's not the fault of the computer but the ones who were programming it. This would allow for a sort of "advanced analysis", as Kasparov might call it, where the human operator tries to find the best questions and the computer tries to find the best answers.
|Aug-09-07|| ||MostlyAverageJoe: <Creg ... I see this as a discussion/debate not an angry tirade with one another.>|
Of course. I don't get angry on internet any more :-)
I'll reply later in the evening to your post. The debate should be probably moved to Kibitzer's Café.
<RandomVisitor> Your last post perfectly reflects my approach to puzzles.
|Aug-09-07|| ||MostlyAverageJoe: <Crowaholic: I also noticed that Rg3+ is present in almost all of the winning lines that RandomVisiter posted, so it wasn't very enlightening.>|
I disagree. What would be a better indication that Rg3 is the center of the entire combination?
|Aug-09-07|| ||Billosky: <There is a huge jump in difficulty level between M-TR and Sat-Sunday puzzles, at least has been so in the past 2-3 weeks.> I agree. This puzzle was very easy for a Th., but the weeked puzzles have been very tricky. But then again I suppose it's no small task to grade every puzzle and assume that everyone else will find it exactly as difficult as you happen to have found it.|
|Aug-09-07|| ||fzoozle: <It looks like everyone missed the puzzle as 19.Qxh6 is a winner even after 19...Bb8 protecting against Rh3+. The right response is 19.Qxh6 Bb8 20. Bxe6!! and either Qxe6, Rxe6 or fxe6 followed by 21.Rh3 with mate to follow. If 21...Qxh3 the rook on h1 will mate.>|
How does white mate after <19. ...Bb8 20. Bxe6 fxe6 21. Rh3 Kf7> ?
|Aug-09-07|| ||fm avari viraf: I think, 18.Nf6+ is obvious & crushing. Hence, instead of the text 18...gxf6 I would better go for the exchange down line 18...Kf8.|
|Aug-09-07|| ||RandomVisitor: After 18.Nf6+
1: Jean-Marc Degraeve - Jean-Pierre Boudre, Ch France Royan (France) 1987
click for larger view
Analysis by Rybka 2.3.2a mp up:
1. (2.89): 18...Kh8 19.Nxe8 Bxc4 20.Nd6 Bb8 21.e5 Bxd6 22.Qxc4 Bxe5 23.Qxf7 Qf8 24.Rf3 Bf6 25.Qxb7
2. (2.96): 18...Kf8 19.Nxe8 Bxc4 20.Nd6 Qc7 21.Rf3 Be6 22.e5 Qe7 23.Nf5 Bxf5 24.Qxf5 Qg5+ 25.Qxg5
3. (#11): 18...gxf6 19.Qxh6 Qd7 20.Rg3+ Bg4 21.h3 Bb8 22.Rxg4+ Qxg4 23.hxg4 Bf4+ 24.Qxf4 Rad8 25.Qh6 Rd1+ 26.Rxd1 a5 27.Qg6+ Kf8 28.Qxf7#
|Aug-09-07|| ||Crowaholic: <Crowaholic: Black has the clearance sacrifice>|
D'oh! It's a deflection sac of course, not a clearance (what was I thinking when I wrote that?).
<MostlyAverageJoe: I disagree. What would be a better indication that Rg3 is the center of the entire combination?>
Hmmm, you do have a point here. But it's not very enlightening on what one might have played if not for Rg3+ at some time.
<fzoozle: How does white mate after 19. ...Bb8 20. Bxe6 fxe6 21. Rh3 Kf7 ?>
Hmmm, maybe 22. Qh7+ Kf8 23. Qg6 Ke7 24. Rd1 Qd7 (..Bd6 25. Qg7+ Kd8 26. Rxd6+ Qd7 27. Qxd7#) 25. Rh7+ Kd8 26. Rdxd7+ Kc8 27. Qxe8#
If 23. ..Qd8 then 24. Rh7! Qd7 (..Qe7 or ..Re7 25. Rh8#) 25. Qxf6+ Kg8 26. Rh8#
|Aug-09-07|| ||Skylark: I got this one pretty easily. I saw that after Qxh6 black couldn't stop the dual threats of Bxe6 followed by Rg3 or Bxe6 followed by Rh3. The only other possibility is .. Bg4 but then h3 makes this move laughable. I thought this puzzle was rather simple for a Thursday O_o|
|Aug-09-07|| ||Zorilla: Heres to Broccoli!
|Aug-09-07|| ||thorndeux: <MostlyAverageJoe: <thorndeux: <mkrk17><MostlyAverageJoe> Doesn't moving the king instead of capturing the knight simply lose the exchange?>|
Of course it does. But it does not give white a forced mate in 10, as is the case in the capture.>
Well, the puzzle's caption doesn't read 'White to move and force a mate in ten,' but simply 'White to move.' I think the puzzle was great - they nearly always are - and as described in the Daily Puzzle F.A.Q. the first move is not necessarily the solution to the puzzle.
In any case Nxf6 was the best move whether black captures or not. Only because one line (capturing) is much worse than the other that doesn't make it a bad puzzle in my opinion. It makes it a harder puzzle.
|Aug-10-07|| ||SpecialK: Knf6+ seems to win it for White!|
|Jun-15-19|| ||tpstar: 9. c3 a6 10. d4 scores well for White = Opening Explorer|
If 19 ... Bg4 to block the g file, White has 20. Qg6+ Kh8 (20 ... Kf8? 21. Qxf7#) 21. Bxf7 threatening 22. Qh6#.
< Earlier Kibitzing · PAGE 3 OF 3 ·