Is computer analysis really any good?
Sort:
[COMMENT DELETED]
I find the computer analysis very usefull for correct assesment of a position and for pointing out mistakes. I see that the latter doesn't interest you but the former is still important for seeing what was going on in your game.
Jul 11, 2010
0
#4
to analyse games very well for your benefit i would suggest analyzing the game with someone else! And even better analyse then with someone who is better than you! They will point out many tactics and often subtle moves that you would not see yourself either due to your chess level or linear thinking of not looking at some types of moves. By analyzing games on your own you most likely will miss a lot of the great moves that a better player would find and look at!
tommynomad, take the analysis result with several grains of salt. It certainly will give you some directions of what might not have been good (and what may have been better), but it may well be that it won't see some clever plan of yours especially if that plan involves sacrifices
Compute analysis with only lines of variation and symbol annotations are only helpful if you have a deep chess understanding of tactics and strategy, but computer analysis with written annotation is very helpful because you know what the complute is basing variation on. For example, like ChessMaster computer chess software
I am certainly no GM, IM or NM!
But for the player interested in improving their chess, who does not have access to a chess club where there might be good players willing to help and is playing at a level below, say ELO 1800 (I don't really know where this might be), using a good chess program such as Fritz or Chess Assistant with Rybka can help you. But don't let it become a crutch! You do need to study tactics, strategy, endgames and openings as well as use analysis.
I do not believe that you should use a chess engine while you are playing. What is the point of this? My engine is better than yours? I can buy the best and most current that there is? My computer is faster than yours? Etc. Morally, you cheating yourself from a good mental exercise.
But after the game, I highly suggest downloading the PGN file and replaying it using a chess program. If you have the time, (I find it takes probably 4 to 5 times the length of time that I played the game,) you should first try to replay the game with the engine off. Try to understand every move that you played and comment on it. What was your plan? What did you see at the time of the game and what to you see now? Do you think you found the best move?
Then you should engage the engine and again go through the game. Does the engine produce the same moves? If not, why not? Don't expect the engine to explain this to you: you need to observe what line the engine took and then think about it. The value of this is that you see what are generally good moves in the situation although they may not be the best moves. The big differences will occur when you blunder: study these and understand why you blundered! But even the small differences can give you a different perspective on the game.
Here is a game that I played rather poorly in where I used this procedure. I used Chess Assistant 10 with Deep Rybka 3 after I did my first pass through without using the engine.
Most of the comments were made before I used the engine. The engine put several references games that had a similar opening pattern for study. Then, as one can see, I was surprised a number of times by his moves: clearly, I have to work to do. But now, I have seen or have been exposed to patterns of play that were better than mine!
See
http://www.chess.com/video/player/when-and-when-not-to-use-computer-chess-engines
Excellent video for engines!
This is a genuine question. I've analysed a tonne of games I've played and naturally, as I'm not very high rated at all, the computer always finds a load of blunders and mistakes in my games. The thing is, often these 'blunders or mistakes' are sort of decoys or the beginnings of some vague trap I have in mind (or they're just cock ups :P) But how does the computer know? And is it worth really studying the computer analyses?
Any GMs out there...when/ if you computer analyse your games, does the computer ever tell you you've made mistakes and blunders where you know the game wouldn't have worked as it did without that certain move?