1. SubscriberPaul Leggett
    Chess Librarian
    The Stacks
    Joined
    21 Aug '09
    Moves
    113497
    02 Feb '10 02:431 edit
    Originally posted by Ajuin
    ahum

    I was joking of course.the explanations are obviously meaningless

    I did like
    5.Bb5+ gives a check
    13.g4 is ok

    it made me smile 🙂
    I have a really funny example of this- in a recent OTB tournament game that I shared in another thread, there was a move where black played Nc3 (I was white), with a white a4-b3-c2-d3 pawn structure. The knight was at a great outpost on the sixth rank, but it was frozen in position, and could not participate in the black king's defense, and I ultimately won with an attack.

    Fritz 12 (which a friend ... "lent" me) assigned a ? and added the commentary "A beautiful square!" I laughed out loud.

    That said, I have multiple versions of Fritz, and I have used the "full analysis" mode to analyze my games after I play them. The function starts at the last move and moves backwards through the game, and it has provided me with valuable tactical feedback, and shows me often where I missed winning tactical shots or better continuations. (EDIT: In the above game, I won in 50 moves, but but Fritz showed me how I could have done it in 35 or so.)

    It's not as good as a human, but I have printed out the games and taken them to my club to share with other players, and the analysis makes a great starting point for our "human" analysis of the game.

    It's not perfect, but it is pretty good IMO.

    Paul
  2. Joined
    10 Oct '09
    Moves
    3027
    02 Feb '10 02:51
    Giving a move a question mark and then annotate it with an exclam is sheer brilliance.

    So these things even have a sense of humor 😀

    And yes,obviously they can be very helpful.But not to explain things with words
  3. Standard memberthesonofsaul
    King of the Ashes
    Trying to rise ....
    Joined
    16 Jun '04
    Moves
    63851
    02 Feb '10 02:55
    Originally posted by Paul Leggett
    I have a really funny example of this- in a recent OTB tournament game that I shared in another thread, there was a move where black played Nc3 (I was white), with a white a4-b3-c2-d3 pawn structure. The knight was at a great outpost on the sixth rank, but it was frozen in position, and could not participate in the black king's defense, and I ultimatel ...[text shortened]... "human" analysis of the game.

    It's not perfect, but it is pretty good IMO.

    Paul
    Fritz may seem to help by pointing out missed tactical oppertunities, but in the end doesn't help at all because the program does nothing to help you to understand why you missed those tactics.

    The best way to learn from your games is to analyze them with depth yourself, or even better find a collection of games with decent (human) annotations, but then write down (with one of those old fashioned ink delivering devices) your own in-depth comments before reading the author's comments. Then it's only a matter of hard work figuring out why your thinking is so messed up after comaring the two sets of notes.

    In short, knowing what move you should have made does nothing--closing the barn door after the bull is on the ice, one might say. To improve, a player needs to fix his way of thinking. As I have yet to see a computer program that can accomplish this feat, there is nothing but dark skies ahead for lazy chessplayers.
  4. SubscriberPaul Leggett
    Chess Librarian
    The Stacks
    Joined
    21 Aug '09
    Moves
    113497
    02 Feb '10 03:01
    Originally posted by thesonofsaul
    Fritz may seem to help by pointing out missed tactical oppertunities, but in the end doesn't help at all because the program does nothing to help you to understand why you missed those tactics.

    The best way to learn from your games is to analyze them with depth yourself, or even better find a collection of games with decent (human) annotations ...[text shortened]... m that can accomplish this feat, there is nothing but dark skies ahead for lazy chessplayers.
    I don't disagree with you completely, but I think to suggest that it doesn't help at all is incorrect. A person with a tactical oversight may never even see the error unless someone or something else points it out to them. And a computer that catches EVERY tactical oversight helps you see patterns in what you miss, which make your studying easier.

    I'm not even sure a human could help one understand why we miss tactics, except to point out the same patterns. I think that we ultimately have to do that for ourselves, and discovering the missed opportunities, no matter who or what points them out, is a good first step.

    This may just be semantics, but I can definitely say that it has helped me.

    Paul
  5. Standard memberthesonofsaul
    King of the Ashes
    Trying to rise ....
    Joined
    16 Jun '04
    Moves
    63851
    02 Feb '10 03:351 edit
    Originally posted by Paul Leggett
    I don't disagree with you completely, but I think to suggest that it doesn't help at all is incorrect. A person with a tactical oversight may never even see the error unless someone or something else points it out to them. And a computer that catches EVERY tactical oversight helps you see patterns in what you miss, which make your studying easier.

    step.

    This may just be semantics, but I can definitely say that it has helped me.

    Paul
    I don't mean to be rude, but before I could ever be convinced that chess engines help humans to play better I need some concrete evidence. What patterns have you found with computer aid?

    For years I was convinced the same way you are now, printing out various computer analysis and pouring over them, until fairly recently, when I asked myself for the same evidence I requested above. And guess what? I had none. Not once shred of evidence that my game was improved even one iota. So I just started studying games, writing my own thoughts down so I could see how empty they were. And guess what? Well, you may not call it results per sae, but in the past year my OTB rating has jumped over 250 points and does not show any signs of stopping it's climb.

    Lesson learned: if it's easy to do, you probably don't learn anything. Actually work to achieve a goal and don't let a computer (or some author) think for you.... Well, all those people out there who think, "a new shiny computer program is what I need to improve my game" really should take this advice to heart or be satisfied with the skill they were born with.

    On the flip side, knowing some alternate moves via computer is not a bad thing, really--but players wishing to improve, especially low rated players, need to be very careful with how the computer generated analysis is used.
  6. SubscriberPaul Leggett
    Chess Librarian
    The Stacks
    Joined
    21 Aug '09
    Moves
    113497
    02 Feb '10 04:00
    Originally posted by thesonofsaul
    I don't mean to be rude, but before I could ever be convinced that chess engines help humans to play better I need some concrete evidence. What patterns have you found with computer aid?

    For years I was convinced the same way you are now, printing out various computer analysis and pouring over them, until fairly recently, when I asked myself for the ...[text shortened]... rated players, need to be very careful with how the computer generated analysis is used.
    I have two concrete examples, if we go beyond just chess playing programs and include databases, and one additional one that is very general for me.

    For tactics, the first pattern I discovered is that I miss backward knight moves. I had a vague sense that I had issues with them because I lost a game or two to opponents that way, but the computer showed me lots of examples where I had the opportunity to "do it to the other guy", but never saw it. I think it's easy to find our tactical mistakes that lead to obvious harm for us, but the computer will show you good moves that you overlooked, which are much harder to find. When you are playing the same openings and doing it to your own games, patterns start to emerge. You still have to do the work to find them, but the computer makes them much easier to find.

    Second. Chessbase has a function that will break down all of your games (assuming you have "moused them in" and created a database) based on the types of endings you end up with. From that, I learned that I had a very particular weakness in queen and pawn endgames. I have dramatically improved my play in those endings since I discovered what a big whole I had in my knowledge there.

    In the broad sense, chessbase helped me correct a major misconception about the state of my game. I had always pictured myself as a good opening and middlegame player, but weak in endings. After crunching a database of my OTB games (254 at the time), I learned that my performance rating in games that made it to the endgame stage was about 300 points HIGHER than my overall rating. Basically, I was winning lots of endings, even against higher players, while many of my losses were occurring before the game made it to an ending.

    I think computers are just like books, in that they are garbage in/garbage out, and it depends on how you use them. In the very broadest sense, the computer is great at "Dude, look what you missed here!" type stuff, and it is best at the concrete tactics part, where no explanation is needed- either I saw the tactic, or I didn't, and I don't need a deep explanation to help me understand that I keep missing pins, etc.

    As an aside, I do not interpret your question as rude at all (and aren't we all too sensitive online!), and I hope I have given it the answer it deserves, even though I went beyond the engine parameters a bit.
  7. Joined
    30 Aug '06
    Moves
    28651
    02 Feb '10 04:44
    Originally posted by Wulebgr
    Fritz is very good at explaining how your moves endanger thin ice from heavy bovines and leave cats treed.
    Not only did it call me thin, it threatened to endanger my cat. I challenge it to boxing. Meet me later for details sir. 😠
  8. An airport near you
    Joined
    21 Apr '04
    Moves
    12247
    02 Feb '10 10:52
    Originally posted by thesonofsaul
    I don't mean to be rude, but before I could ever be convinced that chess engines help humans to play better I need some concrete evidence. What patterns have you found with computer aid?
    I find my older version of Fritz pretty good to help with analysis - I don't go over my games anywhere near enough, but I use Fritz's ability to graph the game evals to show me where it's eval [on 60 seconds thinking] changes by more than 0.5, which gives me a quick view of where missed tactical opportunities are. I then investigate those positions myself, before coming back to look at what Fritz thinks best. I then do the same with any positions that change by more than 0.3. The key for me is to do this whilst I can recall how I felt during the game - for example, was I looking for something in the position, or not exerting effort and just playing rote (a large source of tactical mistakes). If I was trying to look deep, for what and with what aim - then compare this with what I missed and try to understand why my thoughts were heading in the wrong direction.

    Then if I'm feeling conscientious I look at planning around pawn structures, using a Chessbase search and something like Sokolov's middlegames book [often will have come up above, but Fritz is less good at the time limit I set it to on understanding things like the d5 break in an IQP - this I like to investigate myself]. However that part usually gets 'postponed' due to the need to earn a living 🙁
  9. Account suspended
    Joined
    30 May '09
    Moves
    0
    02 Feb '10 13:241 edit
    Originally posted by thesonofsaul
    I don't mean to be rude, but before I could ever be convinced that chess engines help humans to play better I need some concrete evidence. What patterns have you found with computer aid?

    For years I was convinced the same way you are now, printing out various computer analysis and pouring over them, until fairly recently, when I asked myself for the rated players, need to be very careful with how the computer generated analysis is used.
    if after a certain level of positional understanding, analyzing with a good engine doesn't help you improve your game, then you don't know how to analyze with them. it's as simple as that.
  10. Standard memberwormwood
    If Theres Hell Below
    We're All Gonna Go!
    Joined
    10 Sep '05
    Moves
    10228
    02 Feb '10 13:371 edit
    Originally posted by philidor position
    if after a certain level of positional understanding, analyzing with a good engine doesn't help you improve your game, then you don't know how to analyze with them. it's as simple as that.
    you'll improve much more by doing the same without it. you don't learn tactics by observing an engine point them out to you. you learn them by doing lots of tactical problems. the process of figuring it out is much more important than the answer. it's about the journey, not the destination. that's how simple it is. 🙂
  11. Account suspended
    Joined
    30 May '09
    Moves
    0
    02 Feb '10 14:034 edits
    Originally posted by wormwood
    you'll improve much more by doing the same without it. you don't learn tactics by observing an engine point them out to you. you learn them by doing lots of tactical problems. the process of figuring it out is much more important than the answer. it's about the journey, not the destination. that's how simple it is. 🙂
    I really don't understand the idea of improving by relying on your analysis. OK, analyzing your games without engines first is fine by all means, but unless you go through the game including your post mortem analysis with a good engine to see how awful, how useless, how pathetic both tactically AND strategically your analysis actually is, I think most of the work goes to waste.

    it's like entering a test on a subject you know very little about (unless you are kramnik or anand with a lot of time on your hands), and trying to check your answers by yourself. it doesn't matter, your analysis will still be really, really weak. Even if you are 2000+, even if you are 2200+, and sometimes even if you are 2700+.

    as a side note, modern engines (rybka 3, shredder 12, stockfish etc) can tell much more about a game than just tactics. if you have some concepts and ideas stored up, you can interpret what plan it advises you to take, what strategical idea you have missed etc.

    as another side note, I think without engine checking all my variations, my tactical studies wouldn't be nearly as beneficial too.
  12. Joined
    25 Apr '06
    Moves
    5939
    02 Feb '10 14:131 edit
    Fritz's analysis is just as unworthy and useless dude. If your worthless analysis is enough to take out the other guy, then it's enough. Just keep your pieces & king protected and apply some pressure here and there, make some threats, screw the other guy. That's all that matters. I have never won a game because of my Fritz at home.
  13. Account suspended
    Joined
    30 May '09
    Moves
    0
    02 Feb '10 14:163 edits
    Originally posted by heinzkat
    Fritz's analysis is just as unworthy and useless dude. If your worthless analysis is enough to take out the other guy, then it's enough. Just keep your pieces & king protected and apply some pressure here and there, make some threats, screw the other guy. That's all that matters. I have never won a game because of my Fritz at home.
    that's a way too simplistic view. Kramnik just scored +3 in a world class tournament, if caruana won his winning ending against carlsen, kramnik would even be 1st place, but after the tournament he said he was very disappointed with his openings and he has to make serious changes etc. so it's never enough. you can't improve with an attitude like that.

    of course you won't win any games because of your fritz at home if you hate it this much. you need to show some affection, some caring. it's all about l-o-v-e.
  14. Joined
    25 Apr '06
    Moves
    5939
    02 Feb '10 14:20
    Originally posted by philidor position
    that's a way too simplistic view.
    I'm sorry my chess isn't up to your standard 😕
  15. Account suspended
    Joined
    30 May '09
    Moves
    0
    02 Feb '10 14:22
    Originally posted by heinzkat
    I'm sorry my chess isn't up to your standard 😕
    please don't distort what I said there. it should be obvious it wasn't your chess level I'm talking about.
Back to Top

Cookies help us deliver our Services. By using our Services or clicking I agree, you agree to our use of cookies. Learn More.I Agree