Analysing by yourself... without an engine... doesn't achieve much.
I beg to disagree : my personal experience is that in my 20's my OTB rating jumped by 200 pts when I started analyzing my games by myself, and there were no software at that time 
Could you imagine how fast you would improve if . . .
Sort:
Aug 16, 2009
0
#22
Aug 16, 2009
0
#23
RainbowRising wrote:
butdid it jump from 1200 to 1400 or 1700 to 1900?
1700 to 1900 OTB at this time
If you think analysis using an engine makes your own analysis skills better, do you also think doing math problems with a calculator makes you better at math ? I see the engine as a crutch in many instances, just like the calculator.
Aug 16, 2009
0
#26
RainbowRising wrote:
Then, like with most maths, perhaps the optimum way to do it, time allowing, is to analyse the game yourself, as fully and as best you can, and then run an engine?
IMO, the very best analysis consists of 4 stages :
- first with your opponent,
- second by yourself,
- third with coach or stronger player to get human insights/ideas
- fourth analysis with computer to detect tactical mistakes (in the game and in the 3 other stages of analysis) and very unusual ideas (as Borgqueen mentioned, and I fully agree with him on this point)
This way, the analysis and corrections of stronger players and computer sticks better, as you have already spend some time working by yourself before
Scarblac wrote:
I know this is the best way to improve, and still I don't do it.
When I try, my brain sort of stops. I can't concentrate anywhere close to the concentration I get while playing. I get bored and I can't find anything non obvious. Especially after a major mistake has been made, then the rest of the game just doesn't seem as interesting to analyze (even though both players still had to play hard chess, there's no reason I couldn't have played those positions better).
So what happens is after an OTB game we do a quick post mortem, at home I just enter it into Scid and sometimes I let Rybka analyze it. I look up where we diverged from opening theory and what the blunders were according to Rybka, and that's it. I'm missing out on the benefits of analyzing my own games.
Is there anybody who had the same problem and managed to get out of it? Should I just stop being lazy?
Ha ha! Yes... I know the feeling when you make a critical error ... then I think "why even continue to analyze after this gross blunder"
Reb wrote:
Imagine.....IF you spent half the time you use for posting in forums on the internet actually studying and/or playing chess !
I feel pretty sure that had the internet been available back when I started I would never have reached NM.
lol...
BorgQueen wrote:
I've been analysing my games just as you suggest ever since engines were available. It is very useful and I agree it should be done by anyone who is serious about improving.
As soon as the game is over, I get the scoresheet (or the PGN) and get home to see how I did -- regardless of the outcome. Sometimes I lose games, but played quite well and other times I played absolute crap but still got home with a win. The outcome does NOT give any clear idea of how you did! Only analysing your own games (with a mentor or a good engine) will do that.
There are a couple of problems with engine analysis though:
1. Engines might be able to show this move is better and by how much, but they are hopeless at explaining why. Some positions, I have spent hours on trying to figure out why my move was 1.50 worse than move A and at the end of it I was none the wiser and a whole lot more frustrated.
2. It can take a hell of a lot of time. Frequently, there are positions that the engine just cannot decide on what's best - sometimes my move is in 9th place (so I then start looking to find out why). When I look back at the engine my move is now 5th. I give it more time. 3rd. Hmmm.... Give it even more time. Hey it's first! Finally the engine has figured out that my move was best. Bye bye time!
3. The reasons behind why that move is best are completely and totally different to the reason I choose the move as best. For example, in my mind, I saw line a and line b so therefore I determined move z was correct. When in fact, lina a is absurd as it includes an illegal move, line b is a blunder for the opponent, so the analysis I did during the move was VERY VERY WRONG... even though it made me "blunder" the best move on the board, for reasons quite possibly beyond my comprehension!
Having said all that, own game analysis is the best resource I have to help me learn and I will continue to put up with these downsides as I still believe it is the #1 thing that helps me improve.
Yep. That's why I still post my analysis to the internet so at least some fresh objective eyes can look over it. It's also very good for someone to explain the positional aspects. The computers pawn decimal system (i.e. +.43) can be really deceiving. The computer is very materialistic and goes after material pretty much always. But there are some positions where a computer will say you are even, but you are busted from a positional standpoint . . . and I don't think you can separate positional and tactical chess.
hicetnunc wrote:
I don't think the group is really necessary, as the Forum section 'game analysis' just fits the bill nicely (unless you want to keep all of this a little more private, of course...)
The only reason I thought of a group is really to control the quality. The game analysis section doesn't really get a lot of comments of interest in my experience. I think a group of folks who are dedicated to commenting on each others games may do the trick.
RainbowRising wrote:
Then, like with most maths, perhaps the optimum way to do it, time allowing, is to analyse the game yourself, as fully and as best you can, and then run an engine?
Ding ding ding! Correct for 200 dollars Rainbow! lol
RainbowRising wrote:
Political, want to start one xD ?
Yes. I'm down. I will start it and make you a superadmin with me. Do you think we should stay away from vote chess and matches? Sometimes they can be a distraction from a group's purpose lol
hicetnunc wrote:
fourth analysis with computer to detect tactical mistakes (in the game and in the 3 other stages of analysis) and very unusual ideas (as Borgqueen mentioned, and I fully agree with him on this point)
RainbowRising wrote:
Then, like with most maths, perhaps the optimum way to do it, time allowing, is to analyse the game yourself, as fully and as best you can, and then run an engine?
IMO, the very best analysis consists of 4 stages :
first with your opponent, second by yourself, third with coach or stronger player to get human insights/ideasfourth analysis with computer to detect tactical mistakes (in the game and in the 3 other stages of analysis) and very unusual ideas (as Borgqueen mentioned, and I fully agree with him on this point)
This way, the analysis and corrections of stronger players and computer sticks better, as you have already spend some time working by yourself before
Hice,
You just wrote the introduction for the new group. Thanks! lol
RainbowRising wrote:
if I buy a program I reckon I could do the last 3 playing online - I could post the game in a forum for the human element. Maybe I should invest in an engine! Are there mixed opinons about which is the best? I recall one of the GM (forgive me - can't remember your name!) saying in one of his videos that Rybka was the best because it appreciated development more than Fritz. What are your views?
I think Rybka 3 is the superior program. It has smashed all the others in the chess computer championships. I am still using Fritz 11 because I'm waiting for someone to sell Rybka to me for the cheap lol.
66_Mustang wrote:
Politicalmusic wrote:
If we had the patience to annotate every chess game we played? How much better do you think your game would be if you reviewed each game you played? What about for a month? A year? A week?
I'm bringing this up because I've read sooooo many posts on how to improve.
We hear
Under 2000, study tactics tactics tactics Know your endgame it will make or break you (I'm guilty of this proselytizing). Get Jeremy Silman's Books Get a study partner Play stronger players Analyze with stronger players Pick a couple of solid openings and stick with them Get a coach Go over GM games If you didn't learn when you were 5, you can forget about it.Now all of these steps are good! I'm don't think they have merit... but I don't see a lot of "annotate, analyze, and review your own games."
I love blitz. But I know that I can repeat the same mistakes over and over and over again... which will make it harder to break. I found out recently I had been playing a refuted line in the King's Gambit for YEARS. The moral of the story, good slow players are good blitz players.
Many of the things you mentioned DO have merit. When you become a strong player and end up playing those with ratings beyond 2400, you will find that you may hold your own through the opening but in the middle game these masters will out play you strategically and tactically. The winning of every game ends with tactics. If you are not extremely well educated in tactics you will lose over and over again until you make the time to be well educated. That requires a lot of tactics study.
Blitz chess, while error made all the time, is very good for speeding the process of noticing tactical possibilities and taking quick steps to negate them. This kind of training can be an asset in regular tournament games - you see the tactic quickly so you can spend more time planning how to appropriately deal with it.
Yes, annotating and analyzing one's games is very important, especially the loses! I actually annotate my in progress chess games as I go because I am planning several moves ahead to see that my line of play is good me. I'm not always successful, but then since I didn't get it during the game, I have Chess.com's computer analyze the game, and then I use Rybka 3 to again analyze the game, move by move. Chess.com's chess engine is only about 2500 in strength, while Rybka 3 is in the 2800-2900 range. Manytimes what Chess.com's engine calls a blunder, Rybka 3 calls the best move for the position, not a blunder at all.
Now the whole point of using a chess engine to analyze my games here, is as I said, I already annotate the game as I play it, so if I lost I do not know where I went wrong. I'm playing some very strong players, so understanding why I lost is not so easy. It's not like playing a low rated player and making a simple tactical or strategical mistake that one can go back to to understand why I lost. Very strong players play very differently than club players.
Good point Ed about analyzing throughout the game. I guess I could REALLY remember why I made a certain move
Politicalmusic wrote:
hicetnunc wrote:
I don't think the group is really necessary, as the Forum section 'game analysis' just fits the bill nicely (unless you want to keep all of this a little more private, of course...)
The only reason I thought of a group is really to control the quality. The game analysis section doesn't really get a lot of comments of interest in my experience. I think a group of folks who are dedicated to commenting on each others games may do the trick.
I have to agree, the game analysis forum is *with exceptions* pretty disappointing. The problem with online game analysis is that it is very hard to balance computer analyzed lines with non computer analyzed lines. Both, imho, are valuable and add to the analysis, but as soon as someone offers computer analysis it tends to shut down other contributors. I want to hear how masters would play a position... engine play be damned. But even masters tend to be intimidated by Mr. Rybka.
We all take computer analysis as gospel, and in many ways it is... but some of it simply isn't _playable_, no human, not Anand nor Carlsen nor Kasparov would quite play some of those lines, let alone a class-player patzer like myself... the "best" line as computer certified, often isn't what's most helpful to me; a theoretically less strong line, but one that has clear, winnable, goals, is going to be more playable and thus more practical.
In my online games here at cc I use a database to improve my play... I recently added to my database a sizeable (+150,000 games) amount of strong computer vs. computer games. I'm starting to regret it. The positions these computers play can be ubersharp -- and if I'm not careful when the game leaves database/book I can suddenly find myself in a position that I have NO IDEA how to play. It's like, "uh, I suppose Rybka would know what to do from here... but uh, I uh, geez... why did I sac the exchange four moves ago?"...of course to get there my opponent was probably using the same games as me, and there's a fighting chance he's just as bewildered as I am. 
I've been analysing my games just as you suggest ever since engines were available. It is very useful and I agree it should be done by anyone who is serious about improving.
As soon as the game is over, I get the scoresheet (or the PGN) and get home to see how I did -- regardless of the outcome. Sometimes I lose games, but played quite well and other times I played absolute crap but still got home with a win. The outcome does NOT give any clear idea of how you did! Only analysing your own games (with a mentor or a good engine) will do that.
There are a couple of problems with engine analysis though:
1. Engines might be able to show this move is better and by how much, but they are hopeless at explaining why. Some positions, I have spent hours on trying to figure out why my move was 1.50 worse than move A and at the end of it I was none the wiser and a whole lot more frustrated.
2. It can take a hell of a lot of time. Frequently, there are positions that the engine just cannot decide on what's best - sometimes my move is in 9th place (so I then start looking to find out why). When I look back at the engine my move is now 5th. I give it more time. 3rd. Hmmm.... Give it even more time. Hey it's first! Finally the engine has figured out that my move was best. Bye bye time!
3. The reasons behind why that move is best are completely and totally different to the reason I choose the move as best. For example, in my mind, I saw line a and line b so therefore I determined move z was correct. When in fact, lina a is absurd as it includes an illegal move, line b is a blunder for the opponent, so the analysis I did during the move was VERY VERY WRONG... even though it made me "blunder" the best move on the board, for reasons quite possibly beyond my comprehension!
Having said all that, own game analysis is the best resource I have to help me learn and I will continue to put up with these downsides as I still believe it is the #1 thing that helps me improve.
There is a difference between analyzing by yourself and analyzing with an engine. The former improves our chess skills, while the latter helps you spot your tactical weaknesses.