Does studying GM games really help?

Sure they are fun to work through, but the games of someone like Capablanca or Fischer aren't going to provide the amateur or class player with much instruction.

They will give you plenty of innovation and imagination to admire, but those things can't be taught.  So...other than for the simple pleasure of playing through them, why "study" GM games?

You said it yourself;

"they will give you plenty of innovation and imagination to admire"

Seeing new plans is always good. 

Moreover, there are plenty of chess channels on youtube explaining GM games. Or coaches.

Yes, seeing wonders is good but does it help you improve as a player? I can watch all the trick shots in the world in billiards, but that's not going to make me shoot pool any better. 

If having GM games explained to us was an effective way to get better at chess, it seems to me that after decades of this kind of instruction, including the past ten years of YouTube, we'd have a lot more people in the upper ratings levels. We don't...almost all chess players (in the US at least) are still pushing wood and trying not to blunder. 

Maybe someone can find more recent data, but as of 2004, only 1% all US chess players were Masters or better, and 90% were rated under 1800.   

http://archive.uschess.org/ratings/ratedist.php

The news gets grimmer as you move down the chart. Over 3/4 of all US chess players are rated under 1500, which is more or less the cut-off from beginner to intermediate player. That means a full 75% of chess players, most of whom have been playing rated games for some time, are still "beginners."  It's a safe bet that almost all of these players have parsed through a few GM games. 

Finally, a full 50%, HALF, of all US chess players are rated under 1000. This is not even beginner...it's like "learner." Granted, many of these are kids, but if you look at the non-scholastic column it's not much better...over half of all US adult players are rated under 1200! 

So I'm not sure that pouring through Petrosian is really helping anyone. 

At least if you know how a trick works, you can attempt to apply it. 

Discovering things on your own is a lot harder.

Sure, you'll fail, and fail and fail, but at least you're trying new things and discovering new positions. Which, to me, is the fun part of chess. Eventually you've made so many errors that you can punish your opponents for making them.

Studying GM games will help you recognize patterns. But, it is pointless to just go over the games. You will need to find games that have analysis by high rated players. You go over the game a few times, write your own thoughts, then read the analysis by Masters. This way you will get to know why certain moves were played in certain situations rather than trying to guess

Most don't study chess in a systematic way. I'll be honest, most of my study time this year has been reviewing master games. Master games give you a window into how good players think, play, and execute plans. Annotated master games are very helpful, and with free chess engines you can sit down and break apart a master game to almost nuts and bolts. Its a great way to learn and after spinning my metaphorical wheels for a year I am now getting traction again and my game is noticably improving again.

Well-annotated GM games are a gold mine for helping some players who already have a reasonable standard of chess knowledge and understanding to become better and better all the way up to GM level. I don't know why you'd think they're not.

If annotated GM games don't help you, nothing will (which is an unfortunately likely possibility if you've been playing for decades).

I find studying GM games are very helpful, if you do it with a friend.

Somebody who's not afraid to try out different variations the annotators may have missed. One of you may see something, "What if white scas his knight here...?" Then try it, one playing white, the other black, and see where it leads.

+1

The use of an engine fits perfect with this idea. You could see the refutation (if any) and determine if it is likely to be found over the board. 

They feed your intuition

Yes, they do, if you are FIDE 2000+.

GM games may also suggest new openings, you would like to try out.

>If annotated GM games don't help you, nothing will<

Given the statistics I rattled off, that's a bold statement. It suggests that nothing is helping anyone, because the vast majority of players in this country...stink. Over half of the adults are playing under 1200.  The next quarter is sitting under 1500. That's an awful lot of crummy, unimproved chess getting played, year after year after year. 

You guys must all be those very rare exceptions who are improving by leaps and bounds, because every other person you see at a tournament or club isn't improving at all...and half of the remainder is mired under 1500.  

I'm going to sit tight on my assertion that studying GM games isn't helping anyone improve, because there's zero data presented so far to indicate otherwise. 

>They feed your intuition<

????

That's interesting. It makes absolutely no sense of course. 

>Well-annotated GM games are a gold mine for helping some players who already have a reasonable standard of chess knowledge and understanding to become better and better all the way up to GM level. I don't know why you'd think they're not.<

Maybe because exactly no-one with a "reasonable standard of chess knowledge" is moving "all the way up to the GM level." For those incredibly few people that do, it's not from studying GM games. 

>GM games may also suggest new openings, you would like to try out.<

If you are studying GM games for opening variations, then you are likely also a GM. 

The first flaw in your argument is that you are acting like everyone studies the games. The second is that you think the ones that are, are studying the right way. Personally, I can not devote the time and energy I think is necessary to improve that way. But have done a little in the past and that is where most of my knowledge came from. To say it doesnt help is egregious as it probably is the second most said thing to do, after analyzing your own games.

i find them very very useful. That's pretty much all i use, for openings, tactics, positional chess, everything. 

Yes. It's not like every move by a GM is absolutely brilliant and mysterious. Many (if not most) moves are quite standard and can be understood with some effort and guidance. Computers can be used to resolve tactical difficulties you encounter in personal analysis. If you develop a defeatist attitude to game analysis, you stop trying to expand your comfort zone and you prematurely plateau. It's probably better to start with older games that are easier to understand. Logical Chess: Move by Move is a good game collection for the beginner. Get a board out, set up the positions, cover the remaining moves, and at each turn try to figure out a move for one side before reading on. Puzzle books can help with calculation. Review games repeatedly and you will see more each time. You're not going to understand everything overnight, but you will make progress. A large percentage of players will just browse games they see online. This is a more passive approach to study and is not the same thing.