Warning: Do not use computers to look at your games!
Sort:
0
#2
well its true.. You'd need an engine to find the complex tactics you missed in your games.. But when a begginer starts using an engine to analyse his games.. He might tend to believe whatever the engine score tells him.. Without even bothering to understand the position.. To make it simpler... It would be like vishy anand analysing a novice player's game in 'GM language' which the novice would have no idea what vishy is talking.. And would have his own assumptions!! Which is 'bad for a beginner' to improve his chess.
Actually I was thinking the other day this may be good advice (to stay completely away from computer analysis) for players around my skill level too. Didn't come to agree with that wholeheartedly, but the use for computer analysis is definitely limited.
Partly because nearly all the "advice" it will give is useless in terms of practical results for a player like me.
As for missing the tactic, I suppose a computer is handy for pointing out a combination you missed, but honestly this is only instructionally useful if the tactics is both a pattern you've never seen and not too deep to have been impractical to see given you skill.
Otherwise it does come down to concentration, practice, etc. And you will get a lot more out of your games if you go over them thoroughly yourself first. (This means more than once :)
Musikamole wrote:
What about reason 11? I missed a tactic that would have won material. How am I going to find these material winning tactics without using a chess engine?
First, you can see some of them by yourself in post-mortem if you pay attention. Second, you should analyze with stronger players. There are many ways to do that : your opponent, friend, club mate, game analysis forum, coach...
Human insights are what is going to improve your game the most. Computer analysis should only be used as double-check features.
The worst part of engine analysis is that it encourages lazyness... 
I mean : imagine someone playing a guitar work corrected by a computer for accuracy in tempo and heights...
0
#5
hicetnunc wrote:
Musikamole wrote:
What about reason 11? I missed a tactic that would have won material. How am I going to find these material winning tactics without using a chess engine?
First, you can see some of them by yourself in post-mortem if you pay attention. +1
Second, you should analyze with stronger players. +1
There are many ways to do that : your opponent, friend, club mate, game analysis forum, coach...
1. Human insights are what is going to improve your game the most. Computer analysis should only be used as double-check features.
2. The worst part of engine analysis is that it encourages lazyness... I mean :
3. imagine someone playing a guitar work corrected by a computer for accuracy in tempo and heights...
Your comments are priceless and always get me back on track. Thank you!
1. A strong player here at chess.com has annotated a few of my games several months back, including comments and continuations, without engine analysis - along with what I should pay attention to in the future. It's so much better than what Fritz 12 can do for me. Human analysis is priceless. I need to look into having this done again, especially after jumping 200 points in tactical strength!
2. It makes me more than lazy. I actually start to get bored and sleepy, since most of the computer continuations are so terribly long, complex and impossible to remember. It's not interactive, and it's just not fun.
3. Oh my! I started playing the guitar at age 7, and now I am 52! I've spent more hours with the guitar than with any other musical instrument, and the guitar was my major instrument of study in college, and over a lifetime, spending as much time on the guitar as someone one spend to become a Grand Master in chess, if not much more. Geez!
A metronome can help with a developing musicians tempo, but when it comes to correcting/developing musicality, that requires another set of human ears.
Again, thanks.
0
#6
waffllemaster wrote:
Actually I was thinking the other day this may be good advice (to stay completely away from computer analysis) for players around my skill level too.
Didn't come to agree with that wholeheartedly, but the use for computer analysis is definitely limited.
Partly because nearly all the "advice" it will give is useless in terms of practical results for a player like me.
As for missing the tactic, I suppose a computer is handy for pointing out a combination you missed, but honestly this is only instructionally useful if the tactics is both a pattern you've never seen and not too deep to have been impractical to see given you skill.
Otherwise it does come down to concentration, practice, etc. And you will get a lot more out of your games if you go over them thoroughly yourself first. (This means more than once :)
I am astounded! You have the playing strength of a professional chess player compared to me! I would think that computer analysis might help you with refining things like your opening repertoire, since players of your strength are playing more book lines.
Also, you could appreciate better than me a combination or series of exchanges you missed five moves ahead that would have won a pawn. You have the ability to calculate that far ahead, perhaps more. I don't.
Yep. I find it fascinating thay you have entertained the idea of staying away from computer analysis.
Of one thing I am certain, which is to stay away from long strings of computer analysis, which I can't remember anyway, and using only Blunder Check (for the big mistakes) after going over my games alone or better yet, with a stronger player.
waffllemaster wrote:
Partly because nearly all the "advice" it will give is useless in terms of practical results for a player like me.
As for missing the tactic, I suppose a computer is handy for pointing out a combination you missed, but honestly this is only instructionally useful if the tactics is both a pattern you've never seen and not too deep to have been impractical to see given you skill.
Otherwise it does come down to concentration, practice, etc. And you will get a lot more out of your games if you go over them thoroughly yourself first. (This means more than once :)
I aggree.
When I analyze my games, I first try to find blunders and improvments, without a computer.
After that I start the enginge, but I move the engine window so I only see the green/yellow/red indicator, and then step through all moves. If the indicator indicates yellow or red for a move, I put a ? or ?? in the notations, without looking at the specfic line.
I after I have marked the whole game, with the bad moves, I try again without the engine to see why the moves were bad. I create sidelines, and let the engine tell what it thinks about these. Sometimes if I don't find the best line, I let the computer show the best line. But sometimes this line a very complex tactical line.
But I'm not sure if the above procedure is helpful for my improvment. But it is very easy to by lazy and not spend enough time on manual analys. Sometime I spend 2-3 hours on a game, write notes etc, but not always.
I also think your on (long) games is a gold mine of information, about your own chess ability (or lack of).
0
#8
actually GM Roman said to use Rybka program to analyze your games to spot opening errors. He suggested in a beginners video. He said Rybka understand opening principles and understand to punish opening errors.
http://www.chess.com/video/player/no-exceptions
I think you should analyze the game without a computer to start with and afterwards use the computer to see if it agrees with you.
memories is a process of the mind where good and bad memories are stored and learned by our mind. there is what they call trained memories and untrained memories.the trained memories maybe ugly sometimes but is is easier for recall depends on the careful observation and continous study of what it is stored for the mind in its learning process and stage of development thinking for further improvement. as the memory process keeps on improving with its learning process the better it become so its speed and done without consuming so much time.computer assisted learning maybe quite difficult as it is more of the advance stage of a variation which the novice cannot fully grasp for the moment but longer exposure to chess like playing 1000 games on line may teach our memory function to eventually get better.,for untrained memories to take progress.
msiipola wrote:
When I analyze my games, I first try to find blunders and improvments, without a computer.
After that I start the enginge, but I move the engine window so I only see the green/yellow/red indicator, and then step through all moves. If the indicator indicates yellow or red for a move, I put a ? or ?? in the notations, without looking at the specfic line.
I after I have marked the whole game, with the bad moves, I try again without the engine to see why the moves were bad. I create sidelines, and let the engine tell what it thinks about these. Sometimes if I don't find the best line, I let the computer show the best line. But sometimes this line a very complex tactical line.
But I'm not sure if the above procedure is helpful for my improvment. But it is very easy to by lazy and not spend enough time on manual analys. Sometime I spend 2-3 hours on a game, write notes etc, but not always.
Interesting, though it sounds like a lot of work! At the time being I mostly just check my games post mortem for possibly better lines in the opening which I could use in future games. For this I'm perusing the books in my library rather than Rybka or Fritz, though (which, honestly, I would not really trust in evaluating opening lines). But maybe I'll try out your "red lights" technique sometime.
Dec 21, 2011
0
#11
I don't use an engine a lot at this point, mostly just checking some individual positions that occurred in my games to see if I was actually better or worse than I thought. I plan on using one more in post game analysis as soon as I get my backlog of games into a database.
I don't have the money to hire a coach and while there are a few local players that are stronger than me we don't go through game analysis more than just a few variations during or after a casual game.
My thought on the issue, in most cases, is that an engine can be an adequate substitute for a stronger player if you don't have good access to one. If I was unable to see a positional or tactical problem in my game OTB, when I had plenty of time, I am unlikely to see the issue in analysis either. There are some exceptions, especially when the problem was exploited, but for the most part I think that holds, especially at my current level.
My plan moving forward is to use the engine, post-game of course, to provide a move-by-move evaluation (no variations/lines) that I will then use to try and guide me to what I did wrong and what would have been a better plan. In addition, I will use the engine to tell me once the game left book, so I can also work on my openings.
Once that analysis has been completed, I plan on using it to see if there were problems with my analysis and finally to provide some concrete analysis/lines in areas I still don't understand.
I really think that how an engine is used is the most important part. Hunting down minor centi-pawn improvements in the game are useless to me, especially at the level I'm at. Pointing out missed opportunities and major issues is more important and where I will derive the most benefit.
If you can study chess intelligently, surely you can use a chess engine in a similar fashion.
Use a chess engine to play over your OTB games, especially the ones you lose. It provides a clean and quick check on what when wrong, and shows you what you "should have done" inside your game.
At the end of a tournament day it allows you to sleep easy that night. Instead of being mentally restless, and thinking about what moves you might have made to avert your loss. It quickly points out these errors to you.
I (typically) don't bother analysing "my wins" with a chess engine. In these games, my opponent "made mistakes" that led to the loss, and mistakes usually stand out.
Engines are just a tool. Good for novices and stronger players alike. They are an aid to your game. Use them freely. But not as a crutch.
Computers are also useful for playing out the tactics of thematic middlegames, in the opening you have chosen to play. Where your opening book leaves off, usually around move #20, turn on your chess engine. It helps with "what's next."
Playing against strong chess engines is not recommended, unless you want die from a thousand cuts. What's the point?
And if you want to read an "endless argument" about why to use chess engines (or not) versus using "chess knowledge" and the advice of stronger players, then check out the long running thread "...On the Two Knights Defense" in the Main Forum on "Chess Openings." It's quite a battleground for opposing assertions about chess engines, versus FIDE "expert level advice." Nearly 350 posts, begun by @Sloughterchess.
0
#13
kborg wrote:
I (typically) don't bother analysing "my wins" with a chess engine. In these games, my opponent "made mistakes" that led to the loss, and mistakes usually stand out.
it may not be that simple, Kasparov adviced to analyze the game you win just as hard, because you may have gotten away with an error early in the game. remember the player making the last mistake loses and he may take advantage of an error you don't know you make next time. http://www.youtube.com/watch?v=B2KKfOGaR_w
0
#14
msiipola wrote:
waffllemaster wrote:
As for missing the tactic, I suppose a computer is handy for pointing out a combination you missed, but honestly this is only instructionally useful if the tactics is both a pattern you've never seen and not too deep to have been impractical to see given you skill.
I agree.
When I analyze my games, I first try to find blunders and improvments, without a computer.
After that I start the enginge, but I move the engine window so I only see the green/yellow/red indicator, and then step through all moves. If the indicator indicates yellow or red for a move, I put a ? or ?? in the notations, without looking at the specfic line.
Extremely clever! The chess engine indicates the mistake, but we must find out why it's a mistake. I like it. It gives me a chance to think and not be lazy, and it sounds fun. Thank you.
0
#15
--a wrote:
Danny should change his tune, doesn't Chess.com offer computer analysis to it's paying members? Buy more subscriptions ! ! !
I never checked my games with an engine until recently. I like it. It's not the end all be all of course, and it's my final form of review.
I have a bazillion games analyed by that stupid 2500 strength engine. It's done me no good, as I don't have the tactical strength to make sense of the insanely long computer continuations.
I did discover one, and only one possible good use for chess.com engine analysis as a beginning chess player, and that is for the purpose of what I call mistake stats.
| Score: | 0-1 | Date: | 12/12/2011 |
| Time: | 15|10 | ||
| Opening: | C20: King Pawn Game: Leonardis Variation | ||
Computer Analysis (~2500 strength)
Inaccuracies: 3 = 6.2% of moves
Mistakes: 17 = 35.4% of moves
Blunders: 2 = 4.2% of moves
Even after 3 inaccuracies, 17 mistakes, and 2 game losing blunders - I actually won. So, in a 49 move game, I made 22 bad moves. That is just crazy!
Of the top 10 reasons a chess player loses, I do hope it was for this reason: I did not try my hardest (lack of focus).
0
#16
paul211 wrote:
The downsides of chess computer software analysis are:
Chess move analysis gives a string of moves, just like in music a partition is a string of notes and if one cannot read the notes in the written key, and I cannot, then they do not understand the music written.
A chess engine does not talk nor does it tell you why the string of moves is better than another one.
A chess engine does not give you a plan either, how to develop and what to do for the long term.
I do agree with many of the suggestions posted and let me add:
1. Chess engines are not always good in the opening phase of the game as they always look for an advantage such as a pawn win and do this even at the cost of a poor if not bad overall structure or position of the pieces such as destroying completely the center which is the corner stone of a game of chess, they d this to get the +/- score advantage in a game that makes all of us fell better when reading the score.
Why? Because the programming or algorithm used is build to give the player an advantage and without a constant advantage it is difficult to win a game.
2. Middle game analysis is much better thought by chess engines as it can spot most good tactical moves to play.
3. Chess engines rarely see sacrifices in the opening and even middle games as probably computation of moves exceed the time required to give the best move(s) to play. In other words speed of computers today though we think is great is not enough to compute the results of move analysis fast enough. Perhaps a new engine created that can use up to 24 hours to analyze a move could give the best move to play with a sacrifice and a depth of 20 moves ahead but I think that most people would not buy it as we are all in the fast track and need an answer now.
I have analyzed several grand masters games with chess engines and in some games none could detect the value of the sacrifice in other words none would suggest to play the move the GM played though he won the game.
4. The other weakness of a chess engine is end game analysis and mates, not in a few moves obviously say less than 5 or 7 but a winning combination of say 15 moves.
Not sure why this is the case as a retrograde analysis is not more intensively used, starting from the end of the game and working backward to the initial position. There are a limited number of smaller chess engines free on the web that can solve end games with a limited number of pieces on each side and they do a better job than larger chess engines.
My suggestions are:
Be aware of chess engines limitations.
If you use an engine try to see what rows and files and diagonals are critical to the proposed lines or strings of moves.
Any string of moves beyond 3 or 4 moves ahead just ignore as it is too complicated. I can go 4 moves ahead seldom to 5 and once in one OTB game I saw 6 moves and I was right, it was pure chance as I cannot see 6 moves ahead. What is more important than the string of moves is the pattern that develops. Where the pieces are posted and what defense needs to be eliminated to get a checkmate.
Do resolve first and foremost checkmate puzzles with one or 2 moves and when successful move to 3 puzzles with 3 moves etc. Never go to 4 moves if you cannot solve a 2 move puzzle.
Why do this? Because end game puzzles are simpler to resolve with usually a limited number of pieces on the board so the pattern becomes clearer and when you have a possible combination in the middle game the pattern learned will remain in your brain and you can eliminate the pieces that are not affecting the combination.
Great to hear from you, Paul. Thanks for the post and suggestions.
One of my chess engines does talk to me, Fritz, but it's just pat answers, quick easy responses that don't say much.
Fritz will tell me a simple plan, like White should attack kingside - but not much more than that.
I am surprised that chess engines don't correctly see the value of a sacrifice, especially in gambit openings - which I would think would be part of the engines book knowledge. But yes, engines don't understand the value of sacrifices.
Middle game analysis is definitely where engines excel, and I am going to take up a suggestion by a previous post, and that is to hide the continuations during my analysis and only record the yellow and red engine light in my annotations as ? and ??. After doing so, I will do my best to figure out why, for example, one of my moves was considered a bad move (?) or blunder (??) by the engine.
It surprises me that you rarely feel the need to calculate, or look ahead beyond five moves. I remember that your playing strength was somewhere around Class A or Expert, so I figured that to play at that level, one would need to look past five moves from time to time when playing someone of similar strength.
Absolutely great advice regarding checkmates. The best I can do right now, and that is with a puzzle book, is finding checkmate in two moves. It's hard, and that is when I know there is a checkmate to be found! In a real game, just think of all of the mates in one and two that I am missing.
I purchased John Nunn's 1001 Deadly Checkmates for my Kindle, and I absolutely love it. I am very going through each puzzle, looking at every piece and every square, working on my board visualization and calculation while also working on moving pieces in my head.
Moving pieces in my head is proving to be the most difficult! I have a real hard time, for example, visualizing a bishop on a different square than it is in the book.
Interesting thread, some good tips. Maybe computers work well for some players? I think the important idea is to make your analysis applicable to your future games. In this way, hitting analyze and glancing at the blunders is probably useless.
My favorite method when I use an engine is with equal positions where many moves are possible. I think for a long time and pick out a move, and as long as the evaluation doesn't plummet I know my idea was sound enough to play against my peers (for example if it drops .2 I don't care).
If it drops a lot, I take back the move and try again. If I can't seem to find a good move I'll have it show it's top 3-4 moves and pick one to continue. It's useful to pick moves that you really think are ugly because then you can take the other side and try to prove the computer wrong (and inevitably learn something about the position by failing to do so).
Again, no so useful in tactical situations.
0
#18
IMDeviate wrote:
Chess engines rarely see sacrifices?
Not true. They see speculative sacrifices all the time and ignore them.
If an engine sacs something on you, chances are it's already calculated out to mate or a very big advantage.
But engines do see sacs.
true, chessmaster sees sacs but only if you give it some time to think deep enough therefore if I analyze with chessmaster I let it think for at least 300 seconds per move. it takes all night for the computer to finish analyzing.
I've got blunder alert even on sound Bxf7 sac because the blunder engine isn't deep enough compared to mentor line.
I have actually started to ignore blunder alert when I play against the computer.
0
#19
IMDeviate wrote:
bobbyDK wrote:
IMDeviate wrote:
Chess engines rarely see sacrifices?
Not true. They see speculative sacrifices all the time and ignore them.
If an engine sacs something on you, chances are it's already calculated out to mate or a very big advantage.
But engines do see sacs.
true, chessmaster sees sacs but only if you give it some time to think deep enough therefore if I analyze with chessmaster I let it think for at least 300 seconds per move. it takes all night for the computer to finish analyzing.
I've got blunder alert even on sound Bxf7 sac because the blunder engine isn't deep enough compared to mentor line.
I have actually started to ignore blunder alert when I play against the computer.
Quite right, it depends on how deep and how wide the engine is set to calculate, also how the engine rejects obviously bad choices. If the reason for the sac is beyond the "horizon" then the engine might reject it. Because it doesn't see it.
Houdini doesn't like the King's Gambit, which involves a sacrifice on move two. After 1.e4 e5, I had Houdini calculate 10 variations and it never found 2.f4 as one of the good moves. After 2.f4, Houdini gives Black a slight advantage.
I enjoy playing the King's Gambit and like White's chances. Plus, this opening is a big part of chess history and perfectly fine for the amateur to play. I've heard that GM's won't play it much anymore because Black has found all the defensive resources necessary to equalize. But at my beginning level, looking for a slight advantage in the opening for White, or equalizing as Black does not apply, since the evaluation swings so many times from one side to the other.
0
#20
IMDeviate wrote:
That's another thing about engines. Engines don't play the opening well. They often are stuck repeating moves in their openings database without taking time to explore other ideas. And that openings database usually doesn't get updated during the life of the software (well ok some do) which means yesterday's take on a particular variation is the only opinion the engine has. If Nakamura or somebody else finds a new idea with promise, your engine won't stand a chance.
Now I am confused.
Houdini doesn't like 2.f4 (after 1.e4 e5) at all, giving Black a slight, maybe even a moderate advantage, which sounds absolutely terrible. If I were a brand new beginner with no book knowledge, I would never play 2.f4 again, since according to Houdini at Depth = 19, White gets nothing for the sacrificed pawn (- 0.30).
I don't see any lead in development in the engine continuation for White after 2...exf4, which just can't be true in the King's Gambit Accepted in most lines, nor does Houdini seem to think much of White's open f-file, where a rook can place pressure on f7.
I would think that Houdini would find 2...Bc5 as the absolute best move for Black to play. A really strong move. The King's Gambit Declined with 2...Bc5 is the last thing that White wants to see, and yet Houdini loves to capture that pawn. I still have Houdini thinking hard, and it likes 2...d5 (Falkbeer Counter Gambit) as Black's second best try. Now, the Falkbeer is really good, according to theory, but dang, Houdini still can't find 2...Bc5, with 2...d6 in third place and 2...Nc6 in fourth place at Depth = 20.
2...d6 is a really stupid move. Black has absolutely no reason to play such a weak move like that. Ask GM Paul van der Sterren, author of the FCO.
I'll run Houdini again with five continuations and see if it likes 2...Bc5.
Here is where things get the most confusing. The opening book that came with Fritz 12 shows 2.f4 as the second choice, after 2.Nf3, with 2.Nc3 in third, 2.Bc4 in fourth and 2.d4 in fifth. And yet, for whatever reason, Houdini is not consulting with that book. Maybe I have the settings messed up in Fritz 12?
---
Update
Houdini Depth = 21/56
After 1.e4 e5 2.f4
= (-0.23) 2...exf4
= (-0.03) 2...d5
= (-0.01) 2...d6
= (0.02) 2...Nc6
= (0.08) 2...Bc5
Please Sign Up to comment.
If you need help, please contact our Help and Support team.
During a video lesson, I heard IM Daniel Rensch say that beginners should not use computers for game analysis. What do you think?
1. Target Skill Range: Beginner (Rated 1000-1399)
I recently made the low end of beginner in Live Chess Standard, but that’s only if 1000-1399 represents my chess.com rating and not USCF. If it means USCF, then my skill level is less than that of a beginner, whatever that might be called.
3. Learn the Top 10 Reasons a Chess Player Loses
It is important for you to keep track of these common mistakes as a "reference list" as you consider your own losses…
What about reason 11? I missed a tactic that would have won material. How am I going to find these material winning tactics without using a chess engine?