< Earlier Kibitzing · PAGE 6 OF 6 ·
|Mar-02-12|| ||maxi: I am reminded of some posts aroung last midyear. I analyzed some spectacular loss by Capablanca. After carefully going over the variations with and without the computer, I wrote in the game's kibitzing that, from a general point of view, Capa's mistake had been to misplace the Queen in the center where there was just too much opponent piece activity. Somebody (that fortunately seems to have disappear from the map for good) called me a beginner and many other names, something to the effect of, "When are beginners going to stop insulting our intelligence with trivial statements." You see, this guy always gave long variations that you could tell were still warm from the computer's oven, as if that meant that he was understanding the piece play.|
Yes, we can use computers to improve variations. No, we still cannot have computers implanted in our heads! We still are humans... thinking.
|Mar-02-12|| ||Penguincw: And black will be up a full piece.|
|Mar-02-12|| ||King Death: < Everett: ...Computer evaluations seem to destroy the human ability to understand pragmatism under the stress of competition...>|
What seem to be endless criticisms of even top GMs in play by Monday morning quarterbacks that have one silicon helper or another aren't much of a turn on either.
|Mar-02-12|| ||LoveThatJoker: What a game from Morozevich! As Kramnik once said of him, "Morozevich's games are very entertaining. Always forward!"|
|Mar-02-12|| ||LoveThatJoker: PS. This is the exact quote - it is actually located in Moro's CG bio page:|
"Morozevich is a bright player; I like how he plays. This is active chess: only forward! Sometimes luck is on his side, sometimes it is not. It is not boring to watch his games." – Kramnik
|Mar-02-12|| ||Everett: <Once: <Everett> Nicely put!>
You inspire me!
And Thanks <maxi> and <KingDeath>, for not getting carried away with the computers as well... I for one feel very lucky to have learned the game before the software got too good, and my own poor use of it has actually diminished my ability to calculate on my own. Oh well, let's get back to learning how Moro created a beautiful game with the help of Sokolov.
|Mar-03-12|| ||AylerKupp: Sigh . . . here we go again. Why do so many insist on thinking that finding better moves than the player played OTB constitutes a criticism of their play? It's no crime to identify the best move, regardless of whether it originates from a human or a computer. The best move in a given position is . . . the best move. Period. Finding a better move than what was played OTB by a human under the stress of competition and a ticking clock, whether done by a computer or a human after the game is played in the comfort of one's home (or enclosure) and with all the time in the world, doesn't diminish or detract from the talent of humans that can find near-optimum moves (and often optimum moves) under difficult circumstances.|
Mathematical precision is not something to be scorned and is not incompatible with either the human element nor art. It is a form of beauty all on its own. And computer evaluations not only don't destroy the human ability to understand pragmatism under the stress of completion but can actually enhance our admiration for those who are able to come oh so close to perfection under difficult circumstances. At least it does to some of us. As far as using computers, I find that it has increased and not diminished my ability to calculate. Perhaps it’s the poor use of computers that has caused the contrary. One should learn to properly use the tools available before criticizing them.
Yes, lets get back to learning how Moro created a beautiful game with the help of Sokolov. It has nothing to do with whether computers increase or diminish our capacity to admire the game.
|Mar-03-12|| ||Once: <AylerKupp> You have a point and I agree ... up to a point.|
<Once> a game of chess has been played, it changes status. It stops being a struggle between two people and becomes public property. Then it's perfectly legitimate to analyse it, discuss it, find better moves.
It's the same for a work of art - a painting, a book, a film. Once it is out there, the public will take over ownership of it. They will form theories about it, discuss it, may even write a sequel or two.
George Lucas doesn't seem to understand this when he constantly changes the storyline to star wars. Han shot first - that belongs to the fans now. And no amount of fancy editing in the later versions will change that.
And, as you say, there is nothing wrong with using computers as tools to aid our analysis of a chess game.
What I object to is the use of computers to criticise someone who did not have the benefit of computers at the time. In objective terms 32...d2 is a better move than 32....0-0-0 because it wins more material and mates more quickly.
But I still maintain that 32....0-0-0 is a more pragmatic move to play OTB. Its gains might not be as great as 32...d2, but it needs less calculation. And what we don't know is how much time Moro had to make that move.
And if we want to learn and improve our own play, it sometimes pays to try to understand why a strong player didn't play a move that the computer recommends.
So I don't subscribe to the school of thought which says that computers are bad. But nor do I believe that we should slavishly trust in them. We need to think for ourselves, using computers where they are heplful.
|Mar-03-12|| ||lupko: All this discussion about using computers in order to find mistakes or criticize moves played by grandmasters started when i commented the move 32. I think it lead to some unnecessary and pointless debate. |
When i went through the game, on move 32 I just spotted the move d2 in a second, thinking why Moro didn't play it, and then took another about 15 seconds to calculate that it was completely safe and couldn't harm his king before he mates him (with his two (or one) queens).
I really didn't expect that this innocent comment could provoke so many accusations and grumbling.
I even don't know how to use computers for analyzing chess games. It was just a pure human remark.
|Mar-03-12|| ||Once: <lupko> The reason for the debate is that this is a long-running issue on this site. |
There is one group of people who think that GMs should always play the objectively best move and that we can criticise them if they don't. Then there is another group of people who think that GMs should play a reasonable second best move if it is a pragmatic choice.
Then there are a couple of side issues to spice things up. Some folk don't like computers. And some folk don't like it when we are a little too free with our criticisms of players who are much stronger than we are.
So when you asked your interesting question you sparked off a few of these old debates. Nothing directed at you, just unfinished business.
There's another point here. One of the raison-d'etres of this site is that we want to learn and improve our play. So when someone with the strength of Moro passes over 32...d2 (which I am sure he would have seen) and plays 32...0-0-0 instead, we really ought to ask why.
That was the excellent question you posed, and what I tried to answer. The rest of the debate was about the issue and not about you.
|Mar-03-12|| ||King Death: < Once: <AylerKupp> You have a point and I agree ... up to a point...>|
So do I but only to a point.
<...And, as you say, there is nothing wrong with using computers as tools to aid our analysis of a chess game...>
I agree with both of you on this though I don't use one.
<...What I object to is the use of computers to criticise someone who did not have the benefit of computers at the time...>
This is where I also have a problem, I see kibitzers that tear strong GMs to pieces on different pages and it's almost funny because somebody like me (that actually managed to draw with and beat some GMs in active play) can't touch these players overall.
<...I don't subscribe to the school of thought which says that computers are bad. But nor do I believe that we should slavishly trust in them. We need to think for ourselves, using computers where they are heplful.>
A good habit here and in real life. I wonder if over time relying on the engines might hurt somebody's ability to work things out for themselves.
|Mar-03-12|| ||Everett: Yes, many thoughtful points. Just to bring it full circle, the original question by <lupko> was this:|
<: Instead of 32...0-0-0, why black didn't play d2? It seems even more efficient, and still safe for his king.>
To which <Once> replied:
<Probably because Moro isn't a computer.> and goes on to discuss the human elements of chess, like a ticking clock, issues of clear calculation, an easy calculated win (which conserves energy over long tournaments, or at least allows one to enjoy winning a won game without losing your mind)... His response was by and large as perfect as could be. No insults, just strong reasons why a super-GM would take an easier/slower win than a quicker/harder-to-calculate win.
Of course mood and style leave an indelible mark on move choice. So many things the human element brings to chess....
|Mar-03-12|| ||AylerKupp: <Once> And I agree with your points. I was remiss in not stating that I wholeheartedly agree with your reasoning that 32...0-0-0 is a more pragmatic move than 32...d2 over the board and that I think that Morozevich was correct in playing it. Particularly since, as you pointed out, after such a complicated game both players were probably short of time and Morozevich was correct in choosing the safest move to win. Why take a chance in losing a won game, particularly one as attractive as this one, due to a possible and understandable error in working out the possibilities? Leave that to those of us who are less talented and have both the time and tools to analyze the game after the fact to see if they can find better moves. So I was not disagreeing with Morozevich's choice of moves, not at all.|
The reason for my diatribe (and I readily acknowledge that's what it was) is that several people seem to think that finding alternate and possibly better moves after the game is over, particularly with the help of computers, necessarily constitutes adverse criticism. That may be the right interpretation for some post-facto analysts but not all, and hopefully not the majority of them. Certainly not for me. Many ultra-strong players in the past (Lasker and Tal immediately come to mind) did not necessarily play the strongest move in certain positions, even though they probably saw it, and that does not reduce our admiration for their play. In fact, it probably increases it. It certainly increases our enjoyment of their games.
|Mar-03-12|| ||AylerKupp: <lupko> Yes, you triggered some additional discussion in what is a recurring and probably unfinishable debate. But I wouldn't call the subsequent discussions accusations and grumbling (well, perhaps grumbling is an appropriate description in my case but I was trying to address a somewhat different subject), nor would I categorize the subsequent debate unnecessary and pointless. It was all rather mild compared to some of the other discussions that occasionally take place on this site. And, heck, I thought it was interesting to see others' thoughts on the subject.|
<Once> explained everything very well, as he always does. It's tough to follow <Once>'s explanations and attempt to improve on them! Time to think of him as the "explanation computer". :-)
|Mar-03-12|| ||Once: I think it is about <how> we say it. |
Post match analysis - with or without computers - will often unearth better moves than those played. Incidentally, that's one of the reasons that prevents me from posting any of my feeble OTB efforts on this or any other site!
And this searching for truth is a good thing. It helps us all to learn and gets us closer to "chess truth" about a position and a game.
But when we find something that was missed OTB we should really resist the temptation to think (or claim) that we are better than the players or that they were eejits for not spotting it. We operate in a very different environment to the players.
So I'm cool with spotting and pointing out the improvements. I'm also comfortable with the sensible use of computers. It's what <King Death> calls monday morning quarterbacks (lovely phrase!) that I feel uneasy.
What I would really love would be for the players themselves to visit these pages. Let us know what they were thinking and why they played this move or that. But given the rough treatment we sometimes give their games, I think I know why they don't visit often.
|Mar-03-12|| ||AylerKupp: <<King Death> I wonder if over time relying on the engines might hurt somebody's ability to work things out for themselves.>|
You have a good point there. Those of us that are old enough to remember life prior to ubiquitous electronic calculators are probably not nearly as good now with simple arithmetic as we were before the advent of calculators. The same may happen with chess engines. I know that I am often reluctant to trust my instincts and calculations until I've checked them out with a computer but in my case that lack of confidence is well founded since I am not, never was, and never will be a strong player. So for me using computers to provide meaningful analysis (at least I hope that some of them are meaningful!) is a necessity and not a choice. I envy strong players like yourself and others on this site who are capable of providing insightful analysis without the help of computers.
But I think that using computers and reviewing their analyses in detail has actually improved my chess capabilities since I come up with ideas based on my human evaluation of the resulting positions and more and more often (although, alas, not too often) I find that my ideas turn out to be correct. So maybe (hopefully) your fears are unfounded. And, equally hopefully, my perception of my improved chess capabilities is not unfounded.
Let's also remember that use of computers to analyze chess is a relatively new phenomenon since only until fairly recently have sufficiently powerful computers and sufficiently capable chess engines been available at a reasonable price to the chess playing masses. I suspect that we are a long way from figuring out the best way to use computers in this task and hopefully we will do that without reducing our ability to work things out for ourselves.
But I still intend to continue using my calculator when adding and subtracting deposits and checks from my bank balance!
|Mar-03-12|| ||AylerKupp: <Once> We may have to wait until a book about Morozevich's best games appears and hope that this game is included, along with his discussion of 32...0-0-0 vs. 32...d2.|
|Mar-03-12|| ||maxi: Is the best move in a given position the one that the computer suggests? What is the best move? What can you define as the best move?|
Chess is finite game. The number of moves of any game is finite. So you can, in principle, check out for yourself if a given move can force a win or not. If the position is such that there is no way to force a win, then there is no intrinsic advantage to the move the computer suggests. (Remember the computer evaluates on the basis of an evaluation function that is defined by the programmer, a man. Its parametrization comes from our chess lore and the practical experience programming chess engines the programmer may have.)
If the position is such that it is possible to force a win, then the question becomes this: is the line chosen by the human master a winning one? If it isn't then it is a bad move, a terrible move. If it still wins, even if it isn't the shortest, then there is no objective reason to put it down. The question becomes one of esthetics, because human being have this knack for beauty. The best move becomes the most beautiful one, or, at least, the most interesting one.
|Mar-04-12|| ||AylerKupp: <maxi> Good and difficult questions to answer, my answers being somewhat vague and subjective, and raising some philosophical issues.|
My answer to your first question is "Not Necessarily". A lot depends on the chess engine used, the depth that it was allowed to search, and the "cleverness" of the engine's heuristics in selecting which branches of the search tree it searches. And, regardless of the circumstances, any engine is quite capable of making mistakes.
Chess engines are different and some are simply better than others in evaluating the position resulting from the candidate moves they investigate. All things being equal, the deeper an engine is allowed to search the more confident we should be that it will find the "best" move since it will have looked at more potential positions. But all things are never equal. No chess engine, regardless of how powerful a computer it runs in, can look at all the moves in its search tree at the depth required to establish confidence in its evaluation in a reasonable amount of time. So engines restrict (prune) the branches of the search tree that they look at in depth according to some heuristics (a.k.a. "guesses"). The more aggressive an engine prunes its search tree, the deeper it can search in the same amount of time (and therefore increase our confidence in its final evaluation of the position). But at the same time it increases the chances the branch of their search tree that contains the "best" move will not be looked at, so the quality of its search depends on the quality of its "guesses". So search engine developers must perform a number of tradeoffs and determine what approaches work best in the majority of positions.
Oh, yes, the majority of positions. Chess engines currently tend to excel in tactically complex open and semi-open positions, and they tend to do less well in strategically complex closed positions and endgames. To make it even harder to generalize, if two engines are used to analyze a position, the first one might find the "best" move and the second one may not. And, if they are given a different position to analyze, the second engine may find the "best" move and the first one may not. So whenever possible, I try to analyze a position using at least 3 different engines, and average their evaluations to reduce one engine's possible evaluation bias.
To try to answer your second and third questions, IMO the "best" move is the one that will cause the player on the move to reach the position's most likely conclusion in the smallest number of moves. I phrased my answer that way because as you pointed out in some positions a win can be demonstrated, in some positions a win is likely but not capable of being demonstrated, and some positions are just dead drawn and it's not reasonable to downgrade a human or a computer for their inability to find a winning move in that position, assuming equally good play from their opponent.
But I'm not really satisfied with that definition myself. If two moves give the same result, but one of the moves causes that result to be delayed one or more moves, is it proper to consider one move "better" than the other? After all, the end result was the same. And, if you enjoy making your opponent suffer (yes, there are some chess-playing sadists out there), you might even be able to argue that the move that prolongs the inevitable result the longest is the "best" move.
As you also pointed out the criteria for determining the "best" move can also include esthetics And as <Once> pointed out, the criteria can also include practicality. So Morozevich's 32...0-0-0 was arguably better practically than 32...d2 even though the latter may have lead to a win in a smaller number of moves. Not implying at all that Morozevich might have a sadistic streak in him; no, not at all. He was just being practical.
And what about a player who in a lost position plays a move that leads to complications that are the most likely to lead his opponent astray even though it's not the "best" move in the sense that the "best" move would delay the otherwise inevitable loss the longest (the flip side of my definition of "best" move)? Kind of like the horizon effect affecting chess engines. And what if as a result of that less than "best" move one's opponent goes astray and a certain loss is transformed into a draw or even a win? Who would then argue that the move actually played wasn't "better" than the objectively "best" move?
Sorry for my ramblings but as I said in the beginning, there are no easy answers to your good questions. And I am not known for my brevity.
|Mar-04-12|| ||Once: I'm not known for my brevity either! But I'll try...|
There is more than one game of chess. First and (probably) foremost there is the struggle. Two people over the board trying to beat each other. And here the best move is the one that wins. That might not be the absolute best move in theoretical terms. But chess is a struggle, a fight, a sport.
If you are winning, keep it simple. If you are losing, look for ways to complicate. And don't care too much about the "best" move.
Then we have chess as art or science. Writing books, kibitzing, theorising. And here the "best" move does become important. There is no point in adjusting your play to suit your opponent when there is no opponent.
|Mar-04-12|| ||King Death: <Once> The phrase "Monday morning quarterback(s)" isn't mine, it's a common one over here.|
|Mar-04-12|| ||Once: First time I'd heard it, and very apt it is too.|
|Mar-04-12|| ||maxi: <AylerKupp> Nice long exposition... I basically agree with all you say. My aim here is not so much to theoretize about chess engines but to show the dangers of trusting them too much.|
Just because some Monday morning expert gives a long line (actually from an engine) that does not mean he understands what he is quoting (what I called "piece play" before). If you don't understand the reasons for a line then your chess ain't improving that much.
|Apr-04-12|| ||BigEasy1203: If a computer can analyze a game and find a line that changes the outcome ... in those cases, I think it can be important and helpful. Both players could study the analysis and maybe improve their overall game by learning from it.|
However, in situations like this, I don't see the point. A win is a win. Does anyone really care if there's a mate in 3 and it's determined that it could have been a mate in 2?
|May-19-13|| ||4play: a5!!.. Super awesome game|
< Earlier Kibitzing · PAGE 6 OF 6 ·