Any general tips on how to analyze without the engine?

I have heard recently that going over your games only with the engine is bad for your chess improvement, because then the analysis is shallow you just see the best moves and all the inaccuracies etc.. So then you basically treat the analysis as to "get the solution" to every moment in your game and bring up your ego if you found the top/excellent move and feel like an idiot when you blunder...

but ...

I tried to analyse and annotate some of my games (30+20 Classical) without looking at the engine at all in any moment. It felt really nice, but I showed it to a 2100 he was like "bro you missed a million tactics during the game and also in the analysis. What you annotated as your opponent's blunder was actually probably the best move.. you threw here, here, here and there"... etc

So.. for us internet generation never formally trained patzers... How do you improve your own analyzing skills and not just look at the game going "hmm, looks logical.." about every move?... I did find some nice lines btw but I can assume most of them overlooked a ton of stuff, just like in the game itself...

So please give tips on self analysis! Thx 

Maybe at some higher level this sort of thing is actually good for your game. But I am also of the opinion that analysis without an engine is a waste of time.  I stopped watching hanging pawns due to the self-analysis so often being grossly wrong, and I simply don't see the point in it.  I  play mostly daily, and if I didn't find it then, there's not much chance I'll find it looking it over again.  Unless I'm willing to spend weeks figuring out how else to proceed, it ain't gonna happen.

I just looked again at a game review of one of mine, worst blundering game I've had in maybe a year.  The recommended continuations are nearly impossible for me to see.  One in particular is 14 moves deep, isn't forcing, and supposedly was best.  Well, I DID look at the first moves it recommended, but there was no way I was going down that line.  What am I supposed to learn from that?  

I think (not sure)... that the non-engine analysis is aimed at not finding the truth but sortof just continuing the train of thought of the game well after it's over and with no pressure...

so in essence you're just continuing to play the game you played for say 1 hour (if it's 30+20)

for another 2-3 hours... so you can stretch the limits of your calculation maybe?

even decide in your self-analysis some wrong things to later be corrected by a stronger player..?

I litteraly just started doing this but I have thought about this subject before

and your comment actually emphasizes what I am preplexed about: is seeing the truth of the position according to the engine really "learning"? and is it efficient? or just a memorization tool... (of both middle game and endgame patterns as well as pointless openings)

I tend to think it's like doing puzzles to learn tactics.  Some you simply don't get, lots of them actually.  But after more practice, you get better and better.  I don't think there's any memorization any more than there is doing puzzles.  But if you never see what it is you're looking for, it's unlikely you'll start finding it on your own.  JMHO of course.  

I agree with your idea of looking at it to see if maybe knowing the outcome, and knowing what the opponent played, that might help prod my mind to see the continuations differently.  That could be beneficial.  I still do it, but, no way am I going to expect to find this stuff.  And, if all we're doing is learning how to calculate better, I can do that on Chessable.  And I'm pretty confident that works better than me staring at the same position I played wrong before. 

I agree with the question because the computer tells you where you went wrong, even with tiny errors, depending on your level you may not even see those errors when you are analysing it

In my opinion when you start off you should remove the lines but keep the eval bar to try and get used to identifying positions where critical moves are available. This doesn't work quite the same for positional ideas, but it's still a good starting point.

The point is not to copy the engine, but to understand why a best move is the best move. This is not shallow at all - you need to go through greater depths than common human thinking, and henceforth develop higher intuition.

I'm over 65 years old. I grew up without Internet.

I've always analyzed WITHOUT an engine.

Nowadays if I'm analyzing a finished game, I might use a shallow engine to check for tactics but except for blunder-checking I certainly wouldn't let the engine tell me which moves to favor.

Has this hurt my progress as a chess player? Well, I'm 2350 Daily...

this is good to hear

can you share how you first learnt to analyze back in the day? like are there any techniques used to understand  complicated positions (complicated beyond your comprehension) or positions you feel you mis-evaluated during the game?... because I feel there's only so much you can improve upon your in-game-thought process

One minor tip when analyzing is to use two chess sets. One set follows the actual game moves, the second set is used to analyze lines. That way you can reset the second set back to the game continuation just by copying the position from board #1 to board #2.

If you mean : how should you approach the problem of analyzing a position or a game... are you familiar with the Static Analysis system championed back in the 1950s by GM Larry Evans?

GM Larry Evans' method of static analysis - Chess Forums - Chess.com

It's worth reading over and playing out the illustrative games.

Very interesting stuff- I read it through and I have a question about the two methods of space counting. I believe I understand the purpose of both, identifying the coordination of your position based on the difference, but in practice how would you apply this knowledge? That is, how would you utilize a space advantage where your position is focused on a few critical squares differently than you would a space advantage where your position is operating more widely?

You suggest creating threats on separate parts of the board to convert a space advantage, does this method work equally well when that advantage is focused, or is there another approach to consider in this case?

If your space advantage is focused then you have a case of over-protection, which Nimzovich identified and explored over 100 years ago. Several of his games illustrate the typical method.

Here's a game of mine where overprotection themes dominate... the critical e5 square is at one time overprotected by every single White piece, and the square is then occupied first by Knight, then by Queen, then by Rook.

An Overprotection game. - Chess Forums - Chess.com

Perfect, very illuminating, thank you.

My view of engine analysis is that it assumes perfection.  An analogy.  You arrive at a wide river and in front of you is a long row of stepping stones the size and shape of soccer balls and about five feet apart.  The engine will assume you make no mistake and will advise you to take a run at it and land foot-perfect on each one until you reach the other side.  However, It is probably more sensible, facing a human opponent, for you to run down to the bridge and cross that way instead because the chances of slipping outweigh the advantages of the shorter distance.  

What's wrong here? I had typed a lengthy answer for this topic and when I clicked 'post', this appeared. 👇

What I like to do is analyze the game without the engine and annotate my ideas and why I did something, even if that idea didn't work. If my idea fails and I know why it did, I would write that, and if I don't, that's when I would ask the engine. Sometimes your idea works very well, in which case you should also ask the engine to check whether your idea was actually good or if your opponent just overlooked a powerful counter-move, in which case you should probably not repeat your moves with the same mindset.

Remember to take what the engine says with some skepticism; not because it's inaccurate, it's very very accurate, but rather because sometimes the engine has completely inhuman ideas that clearly nobody would have ever noticed, which should not deter you too much imo. I have read a few old books which have combinations which the engine straight up refutes. Some of what we consider to be brilliant sacrifices by people like Mikhail Tal are called mistakes or even blunders by the engine, but they're still pretty solid because it would be very hard for a human to refute them. Not to mention that sometimes players aren't even going for just the best pure position and play odd moves voluntarily because they have their own ideas, which might conflict what the engine wants but are nonetheless sound. Having the engine refute a very nice move you made is like being 6 and having a 16 year old tell you Santa isn't real. Your feelings are hurt by this information but you still get gifts on Christmas anyway.

Maybe I'm a little too hung up on the words people use when they talk about this stuff  but it's just the way I am. I have heard the phrase "analyze your games without the engine" so many times, and from so many different people, that it's become almost cliche. I also find that this is largely advice from old school chess players or people who just parrot what those people say. Not that there's anything wrong with that. But I would also say that there's been a lot of evolution in the last two decades in not only the way the game is played but the way it is learned. More than maybe in the last thousand years before.

If I was going to advocate such a thing, I would say it like this. Go back and take a solid look at your completed games with a fresh and open mind. Don't use an engine and try to find all the mistakes you made in the game.

Now that makes more sense to me than a statement which has all kinds of implied meaning depending upon who is hearing it. Especially when my interpretation of "analyze without the engine" might actually be influenced by things I know for a fact have been done by other players. And I mean by grandmasters. I believe it is Gukesh who's progress basically was a life experiment, and he absolutely never used an engine until only just very recently. This on the advice of his coach. Now clearly that can work because he's pretty good. But It doesn't necessarily mean that it's the easiest way to get better.

I read through the thread about the Evans method and that kind of implies you can do that during a game, but that would be very time consuming. They do say later you don't need to actually do it, you just need to understand the concept, so that makes more sense. They imply it becomes kind of like muscle memory which is true of pretty much all things chess. I'm thinking it's something useful during your self analysis but maybe not so much during the game itself. To each his own. I would not attempt it during the game, certainly not at my current level.

I will say it seems more directed towards classical games than anything else, but I do think there are applications in rapid games- obviously not sitting there and counting everything up, but as you mentioned gathering a rough sense of the themes in the Evan's method in the position quickly. I tried to incorporate that into my rapid games and gained a noteworthy amount of elo yesterday, as well as having just started a few daily games to try and fully employ it in.

Like anything else there's going to be different opinions from everybody about its effectiveness.

Quite some time ago I thought about a graphic system that might show a weighted amount of control of squares in a game, and if it might not be a good mind exercise. I figured somebody had to have thought of it already so I searched on the internet and yes, there was a utility you could paste in an Fen and it would show the squares colored with who controlled them. Let's say green for you and red for your opponent. They would be differing intensities of color if they were hit by one, two or three pieces or so on. If you and your opponent both hit a square, they would cancel mathematically as you might expect. If you hit it twice and your opponent hits it once, it would be light green let's say.

I thought the idea had great Merit. There was a thread on here about the very topic. In essence it's a graphic representation of exactly what we're talking about above. And of course there were people who thought it was valuable and others who said it was completely worthless because it didn't take into account the tactics in the position. I sort of had to laugh, but whatever, I didn't even get involved in the discussion. I just figured yeah there's people that don't like it but I think it's pretty cool. So here we are revisiting that as one of the three main themes this GM came up with for his system back in the 1950s.

LOL