Please turn on javascript in your browser to play chess.
Only Chess Forum

Only Chess Forum

  1. 13 Mar '11 18:46
    I'm a 1200 player looking to improve my game...I've tried to use CM9K to analyze my games and have gotten little out of the long variation lines and scant commentary it produces...I've recently picked up Fritz 12 and it seems a little less intuitive to use...I'll have to mess with it more... I need more verbal commentary and less playthough of deep variation lines...am I expecting too much, or is there a better way to learn from my games? How does one use analysis programs to get better? Maybe I'm just not there yet?

    Thanks
  2. Subscriber Ragwort
    Ex Duris Gloria
    13 Mar '11 19:18
    Originally posted by Regiscyde
    I'm a 1200 player looking to improve my game...I've tried to use CM9K to analyze my games and have gotten little out of the long variation lines and scant commentary it produces...I've recently picked up Fritz 12 and it seems a little less intuitive to use...I'll have to mess with it more... I need more verbal commentary and less playthough of deep variati ...[text shortened]... How does one use analysis programs to get better? Maybe I'm just not there yet?

    Thanks
    If you need "more verbal commentary" then ask a (stronger) human.

    Computers do not "think" about the game the same way a human does and therefore their verbal commentary is largely useless for improvement purposes.

    They are deemed helpful analysis partners however because they will never forget to take one of your pawns or pieces if you leave them hanging when you put in trial lines, or if you play against them. If you input one of your games and play it through, keep an eye on the evaluation reading. Anytime it swings by a point or more then someone has probably made a mistake and the computer has "seen" a line where it can win material.

    What I have found interesting is putting through games I play against stronger over the board players. Even though I have been soundly beaten and never felt I had a chance, the computer will usually highlight how their game was far from perfect and what opportunities I should be looking out for in the next game.
  3. 13 Mar '11 19:24
    Always interesting to see how others use these programs.
  4. Subscriber Ragwort
    Ex Duris Gloria
    13 Mar '11 19:48
    Originally posted by Regiscyde
    Always interesting to see how others use these programs.
    If I am honest I don't believe that a computer engine analysis can ever really help you play markedly better chess. They are very good at digging into positions and analyzing them to death but almost none of this "skill" is passed on to the operator. You have to learn to do that for yourself, then come back to the computer (if you must) to have your work assessed.

    An analogy. Which improves your arithmetic? Stretching your brain to do the sum or turning on the electronic calculator?
  5. 13 Mar '11 19:57 / 1 edit
    Originally posted by Regiscyde
    Always interesting to see how others use these programs.
    The way to use a computer is to go through your games looking for the large increases or decreases in the computers assessment of the position and then working out for yourself, why it assess the position as it does, apart from that, they are pretty useless, for they have no strategical conceptual awareness, none whatsoever.
  6. 13 Mar '11 20:12
    I think Raqwert and Robbie have the right idea. Here is what I do, which is a combination of their methods. I use Chess Assistant to review my games. I put in what I 'saw' during the game and why I made the decisions I did. I also try to identify tipping points or critical moments where the initiative switches from one side to the other. Then I try to figure out why the shift occurred.

    Then I let the program look at my work. Sometimes I avoided a 'threat' that I could have simply refuted. Sometimes I miss a tactic. Then like Robbie said, I use the evaluation graph feature to identify the big shifts and focus on the lines at those points.

    I'm pretty inconsistent due to playing at work or playing while drinking boilermakers, but I find this method of analysis invaluable and fun. For me, having the computer check my thoughts during the game has improved my chess more than anything outside of my rivalry with a certain human player.
  7. 13 Mar '11 21:47
    Hi king killer. I looked through a few of your games.

    You seem to have an aggressive style of play and I saw some nice attacks in the games I saw. These attacks didn't always go as planned, of course, but you'll get there.

    If you'd like some advice about your chess play:
    1. You need to develop your pieces. Allowing a minor threat to exist is better than having half of your minor pieces on their starting squares. Get all of your minors developed and your queen off the back row (safe and centralized).
    2. You need to calculate the outcome of an attack after all of the captures will be done. Include the value of the pieces in these calculations-IE consdier that a player likely won't drop a queen for a knight. I've seen cases were you should have captured and didn't and cases where you should not have attacked yet did.
    3. When you are attacking, keep the focus on getting checkmate. You are aiming to deny escape routes to the enemy king and deliver a final check. If you can close a route while delivering check all the better.

    If you focus on 1 in the early game and 2 in the middle game you'll see your rating skyrocket.

    btw, I annoted these games:
    Game 8233889
    Game 8180579
    Game 8217524
  8. 13 Mar '11 22:18
    Originally posted by Regiscyde
    I'm a 1200 player looking to improve my game...I've tried to use CM9K to analyze my games and have gotten little out of the long variation lines and scant commentary it produces...I've recently picked up Fritz 12 and it seems a little less intuitive to use...I'll have to mess with it more... I need more verbal commentary and less playthough of deep variati ...[text shortened]... How does one use analysis programs to get better? Maybe I'm just not there yet?

    Thanks
    I learned the basics and a bit of theory and mainly self discipline from my brother when I was laid up with casts on both legs, for a few weeks. I wouldn't recommend it, but having a personal trainer is the idea. This site could use something like a learning center where if you can go sign up. Players who have both a high rating and a desire to train others, could volunteer there, and you could play them without effect on your, or their, win/loss records or rating. Your trainer could comment to you on the position,how to analyze it, your options, possibly (with some programming) they could let you take back moves, etc. You could set up positions such as to practice end game theory. You could rate your trainer so the better ones would get more business. Your training games would be part of your public games but marked as learning center so your regular game opponents could check to see if you are getting help that way.
  9. Standard member Thabtos
    I am become Death
    13 Mar '11 22:44
    The best use of an engine you can have right now is to check for blunders. Moves that you made that changed the evaluation + or - a LOT, like by 3 or more.

    Like it's been said, computers don't make plans, which is exactly what a good human player does.
  10. 13 Mar '11 22:45
    Darax, I can't thank you enough. I don't realize what I don't see. Invaluable. Usually I don't understand what weakens me until I blunder...I may try to annotate a few ongoing games and then revisit to examine the faults in logic. I've always wished I could play someone and then discuss the game in depth with them later. Again, thank you for taking the time, I will benefit from this!
  11. 13 Mar '11 22:54
    Originally posted by Regiscyde
    Darax, I can't thank you enough. I don't realize what I don't see. Invaluable. Usually I don't understand what weakens me until I blunder...I may try to annotate a few ongoing games and then revisit to examine the faults in logic. I've always wished I could play someone and then discuss the game in depth with them later. Again, thank you for taking the time, I will benefit from this!
    You should join a club, that will give you all the OTB analysis you need.
  12. 13 Mar '11 22:56
    I think joining a club will provide me with many new and exciting ways to lose.
  13. 13 Mar '11 23:16
    Originally posted by Regiscyde
    I think joining a club will provide me with many new and exciting ways to lose.
    maybe so, to start with, but I guarantee you will improve more quickly than you realise.
  14. 13 Mar '11 23:26 / 1 edit
    Originally posted by michael liddle
    maybe so, to start with, but I guarantee you will improve more quickly than you realise.
    Can i ask you in all honesty Michael why this should be the case? for i have always understood that the greatest benefits come from independent study, but then again, its not always easy to be truly objective with regard to ones play. I just wondered how a club will help in this regard (please note that i am not disputing that it will, just wondering how)
  15. 14 Mar '11 00:04
    Originally posted by robbie carrobie
    Can i ask you in all honesty Michael why this should be the case? for i have always understood that the greatest benefits come from independent study, but then again, its not always easy to be truly objective with regard to ones play. I just wondered how a club will help in this regard (please note that i am not disputing that it will, just wondering how)
    The OP has stated that he wants to improve, as we all do. In my experience, OTB play, and subsequent analysis with the opponent, is more beneficial than studying masters games. Unless it is match night, I regularly get at least 4 games in against my club colleagues. They mostly play different styles and I have gained a lot from the experience. Nothing wrong with other methods at all, just that I feel there would be a quicker, more marked improvement with regular OTB play.