Is computer analysis useful?

I don't know but whenever i play against a computer or when i do engine analysis of any position i start playing dull and lifeless chess. i know this is paradoxical as chess engines can be highly creative, probably more so than ANY human.

my question is, do you do computer analysis and how is it useful.

i know that it is virtually senseless to do computer analysis in complex positions as the move the computer suggests is near impossible for humans to see. i find that computers are best used for endgames where it is easy to see what is going on and why the computer suggest a particular move. after all, the computer is just a calculator and endgames are where calcualtion matters most. also, the computer can show their true beauty only in the endgame. in the middlegame and especially in the opening, they are virtually useless for instance i did a computer analysis of this position and the result is so beautiful. see if you can solve it! black to play and win

any thoughts?

Computer analysis can be useful, but to get the most benefit, you should analyze first without a computer, and then use the computer to look for major mistakes in the game moves or your analysis.

A computer can't give you the sort of feedback that you can get from a good coach or a strong player though -- things such as strategic mistakes or where you misjudged a position and decided on an inferior plan.

Also, don't use the chess.com computer analysis. It is ridiculously bad (assuming it hasn't changed much in the past couple of years since I last used it). Download a good free engine such as houdini 1.5 or stockfish and use that instead.

There is an old saying that you always find a misplaced item in the last place you looked. The reason is you stop looking for it after you find it. Same analogy applies to chess computers. The moment you turn it on, it shows you the best move, and thus you stop looking  at the postion, trying to figure it out. Therefore, your ability to learn to analyze is stunted. It's a shame so many authors and coaches are using computers in their works. It's an ego thing, they seem to think that if they produce flawed analysis, their reputation takes a hit. Balogny! When we play tournament games, we analyze blunders all the time as we sort out what's happening in a game. If we don't see more than our opponent, we don't win, but we learn from how they beat us. but these authors who use engines, simply encourage bad habits, and to me, that's not good coaching.

Engines can be a useful learning tool, but are most helpful to players with titles. I believe they do more harm than good to anyone below master status. I earned my NM when the best computer played 1500 strength, maybe....Many authors who use engines in their work today earned their titles before I did. Did they forget how to analyze a position? It's either laziness or an ego thing they incorporate engine analysis in their work....

great. thanks for your useful suggestions.

this question has perplexed me for some time now.

i agree with aww-rats that computers stunt your analytical ability.

however, the problem is that COMPUTERS EXIST!!!

and it is tempting to look at the analysis that they roll out at the click of a mouse!!

sometimes it produces and adrenaline rush if not a megalomania

the beautiful patterns that they produce are aesthetic as well.

however, the road to mastery is a hard one and if one wants to develop as a chess player then probably computers deter you from developing your own style.

is there a way to use computers without harming yourself. that is something i don't know however i am one of those who is tempted to click the button to see what stockfish says.

Use engines to check your own analysis for tactical errors, but only after you finished it.

aww-rats, what is the problem with using a computer AFTER one has already done very careful analysis without a computer? Of course there are good and bad ways to the use the computer even in that case, but I don't see any reason why any use at all of a computer in that case should always be more harmful than good, as you claim for anyone who isn't a master.

My video lessons course on youTube covered this just over a year ago. I said only after you have exhausted what you can see. Problem is, they are so damn tempting!

Ah, okay, that's not what you said in your previous post though, when you implied that any use of computers at all by non-titled players does more harm than good.

For somebody like me, who is well below master strength, but always analyzes as much as I can without a computer first, then they don't do more harm than good.

Yesterday I was looking at a mostly closed position with a computer.  After suggesting a number of poor moves (I'm not letting it think more than 5-10 seconds) I got frustrated and stopped.  Clearly it had no idea what to do and I didn't have the energy to spend 30 minutes and backtracking many bad lines to help it along.

Yes, sometimes you can turn them on and see the best move.  Other times I think they're full of crap.  Yes they could probably defend the position to a draw (esp. again me ;) and many times they find tactical means to support the moves I call bad, but this is not how humans play.  This kind of analysis is not helpful for practical games (and is sometimes wrong unless you give the computer lots of time and help).

Completely agree with aww-rats.  I'd add that even after you get the "final word" (through yourself, the computer, or another strong player) don't toss out the game (so to speak).  Save it and review it a year later to see if you can find any new lessons.  If it's a very dull or short game perhaps you will exhaust it quickly.  But some games I don't think you should ever really label as completely exhausted.

the computer is very useful if you use it to your advantage and learn from what it spews out.   Giving you a 25 move combination that nets you 0.1 point's worth of position is of little value to a human.   Spotting a tactical attack in the midgame that nets a piece that you missed, even looking it over yourself afterwards, is handy.   Even spewing out the name of the oddball opening variation you used is useful, now you can look it up if you were unfamiliar with that variation, see if it was good or poor.  Lot of named variations are total garbage, and once in a while, one stumbles across one. 

Should you do it yourself? Yes, but not everyone has 10 hours to look over every game afterwards.  Its a modern, busy world and for casual players, the computer gives you GM level analysis in moments.  How you use that... if it helps your game... that is up to you.

computers have certain shortcomings people dont realize.

I remember seeing a double sac once which houdini did not see. After feeding in the moves, houdini evaluated it as winning.

Computers can help on the tactical aspect but not strategy so using the, can help but I prefer another person analyzing the game

And when playing one they avoid bad trades so even if they get tied up they won't trade 2 pawns for a knight to get counterplay

dont waste time on computer analysis

Don't let the computer think for you, let it help you think.

A recent game embarassing a computer, where a human would have done well. Black played by computer4Impossible in a standard time control.

At the level I play mine at, I got similar results to Heister.  My computer tried the same moves (checking the king, playing the pawn to block bishop, playing the knight up) and all failed to white winning eventually. 

Then I kicked it up to max level, with 5 min per move and no limitations.  It soon, but not immediately,  selected Qxg5 with this line listed for its thought process:

00:00:47.6    4.41    11    34591019    Qg2xg5 Nf3xg5 Nb8-a6 Qc5-d6 Ra8-b8 Ng5xf7 Rb8-a8

--- you get out what you put in.  At the higher levels, the computer selects moves better than most of us and certainly would not lose the game to keep a queen. 

Even your computer at its highest level will make this kind of evaluation mistake somewhere.  Maybe it won't be quite as obvious as the above game, but that's why its an example ;).

Oh, absolutely.  But the mistakes it makes will take a GM to spot once you let it dig 15 + moves deep, making them useful tools for all but the top players.   If you can beat a good engine at the top levels consistently, its time to take the game pro.