The Paradox of improving - Please Help

I am at a bit of a loss for what to do.

I have followed the instructions to the best of my abilities: Read books, review grandmaster games, do tactics (I do this on other websites where it is free. Around 30 minutes a day), play longer games on a frequent basis, and analysis.

I always get an absolutely abysmal accuracy rating after games (my last game was 7.0 percent). So you may say "analyze your games to improve!". Ok, so I look at my games and have no idea what moves are bad if there is not an immediate capture or sway of the game. After doing nothing but shots in the dark of "analysis" I turn on the engine because there is nothing more to be done. It highlights the countless mistakes, but I cannot for the life of me figure out what is wrong with 99 percent of them.

Thus it seems like I need to already be better to analyze my games, but I can never get better because I cannot analyze my game.

I am wondering what I am supposed to do. My fear is no matter how many games I play I won't get better because I cannot do diddly squat on the analysis and have to essentially "guess" which moves are good based on the rhetoric I have heard a billion times "Don't hang your pieces, develop, etc" but even when I try to follow this to my novice mind's best ability I still make countless mistakes and my mind is lost before the pgn loads to analyze it.

Without directly looking at the game, your concerns are valid. We don't want to be making the same mistakes game after game and it is rare that people teach others how to teach themselves. 

It is like the old saying:

"Give someone a fish and they aren't hungry for one day, but teach that person to go fishing and they will never be hungry."

Look into how to analyze your own games and if you are still stumped then ask a higher rated player who might be able to explain the position to you

Luckily for us in our era, we have the computer to help us. Back in Capablanca's time, they didn't have chess computers, so they had to figure things out themselves or ask other strong players. Nowadays, almost everyone can access a 3000+ rated engine to help "explain" the positions to them! This is incredible! It is like your own GM on your phone or computer, yet most never learn how to fully utilize the chess engine or how to "interpret" what it says (as the computer doesn't explain moves in English like a human player might).

When you are in the post-game analysis, focus on moves in question. Primarily, this means focusing on "blunders" and "mistakes" and when you become really higher rated (maybe 1500 or so), then you can learn more from "inaccuracies" and by GM level, you might even be investigating flaws in moves that the quick-engine analysis didn't see anything wrong with!

Okay, so we identified a "move in question" like a blunder, so now what? Take a look at the top engine lines suggested (usually three chess variations) and try to figure out why this is better than your game move. If you think your move is better or you found something good yourself, then play it on the analysis board and see what the engine recommends against it. 

There are more complex ways to use the chess engine for learning, but the basic idea is to look at your moves the computer doesn't like and try to figure out what was better and why it was better. Doing a chess analysis is NOT about memorizing computer lines. It is more like a puzzle of trying to solve "why" the computer likes certain things and not others.

I'm going to caveat everything I say by saying that we're pretty much at the same level.  (I've been around 900 for a couple weeks.) So, I'm not speaking from a position of authority.

However, let me give you a framework for thinking about how the computer analysis works.  I think with this understanding you'll be much better equipped to tease out what's valuable in the analysis.

The computer first goes through each candidate move in a position and makes a tree of possible moves that will follow.  It goes as deeply as you've set it to go, usually 18 to 24 moves ahead, to see where those candidate moves lead.  Then, it assigns each candidate move a score based on various metrics about how good a position you'll be with 18-24 more moves of best possible play.  As you step through your game, the bar on the left shows the score of the very best move available to the player who plays next, based on the current position.  (That score's variation is also what the "moments" tab chart shows.)  The score is in units of a pawn, so a move that throws away a 3-point piece is often going to (very roughly) result in a three-point swing in score.

Moves get labeled an "inaccuracy," "mistake," or "blunder" when the score swings more than a certain amount toward your opponent.  Generally, based on what I've seen, a blunder is equivalent to making a move that throws away a minor piece like a bishop or knight.

At our level, it's the single-move blunders that get us, both making them (and losing) or catching our opponent's (and using that advantage to help win.)

If you're trying to figure out why a move is scored well or badly, you need to look at the line of play the computer used to assign that score.  Here's an example:  The first blunder in the game was your opponent, on move 8, playing Nc6.  If you look at the line the computer used to calculate that score (shown on the Analysis tab on the website under "c6 is a blunder"), you'll see it has you recapturing with your knight on d4 (playing Nxc6).  This attacks their queen and simultaneously opens up an attack on their bishop on c5 from yours on e3.  When they take your knight with the pawn (necessary to save their queen) you take their bishop and you're up a piece.  However, instead, on the next move you chose to play Qb5 to pin their knight.  The score, which briefly swung up a bunch to reflect you were about to trade knights and be up a bishop, now swings back down, and the move is marked as a "missed win."

On move 17, they blunder again (interestingly right after the move where you say "the game is lost at this point") and the score swings almost back to zero.  The reason (which I'm guessing by looking at the computer's line for black's best move) is that their playing a6 gave up an opportunity to swing their queen out and pin and capture your bishop, giving check.  Had you noticed that during the game, you could have played something like the computer's line for the blunder to even the game up.

Your move 18, then, is just marked a blunder because you didn't see their blunder and do what you had to do to save the game.

The best way to use the computer analysis is this:  If you don't understand why a move is a blunder, follow the lines the computer predicts for both your blundered move and its alternative best move.  Usually, within a few moves, the reason for the difference will be obvious.

The computer's analysis is an unbelievably powerful tool for finding where your game ran off the rails, but it takes careful work to examine both the line it predicts for your game and the line it predicts for the best move alternative to understand what's going on.

Final thought:  Like I say, at our level, most games are dominated by a few key blunders.  At 900-ish rating, blunders are a lot fewer than at, say, 400, but they still determine how games come out.  Our job as players at this level is to learn to look more carefully at the elements of our positions that we missed when we blundered before, and to try to see opponents' blunders when they happen so we can capitalize on them.

I generally don't worry too much about mistakes or inaccuracies.  While these may be ineffective or modestly harmful moves, it can be a lot harder to see why the engine is evaluating them poorly.  For these cases, having a much stronger player look at the moves might help get you an explanation that can be put into words that make sense, but in a very real way, these aren't the moves (by and large) that are winning or losing our games at 900 rating.

BTW I'm thrilled to see that my friend @KeSetoKaiba (a far, far better player than me!) has pretty much the same message to share.  Hope these thoughts help!!

Oh, also:  @PawnSlappah, you've played 18 rapid games over about three weeks AND you've won 61% of your games!  You're doing all the right things to improve, but chess is hard!  Don't get down on yourself for not magically just having an up trendline to Grand Master.  Also, feel free to add me if you'd like some practice games from someone at about the same level.

Dang there is a lot of good information between you two. I guess my tentative game plan is to use the engine when I get stuck self-evaluating and try to reverse engineer the computer line and try to make sense of it. Hopefully this will translate into become better at self-analysis. I will definitely keep at it. I suppose I am just paranoid that self improvement is an uncertainty in terms of how long it takes and if it takes place

Especially for adults, there's just so much variation in how fast people improve.  Some adults can start where we are and in four to five years work themselves up to 2000 rating, others improve for a while and then get stuck.  If you want some inspirational stories, there's a great podcast called The Perpetual Chess Podcast that has a regular series of interviews of people learning the game as adults who have made significant strides forward, and it's worth checking out.  Search in their episode list for "Adult Improver."

Awesome to see that the chess.com community has given lots of helpful posts while I was creating the game annotations  

p.s. Thanks for the kind words @Paleobotanical, you're too kind

Man, awesome analysis, @KeSetoKaiba!  Where's the like button...

" Give a man a fire and he's warm for a day, SET him on fire and he's warm for the rest of his life."

Focus on a topic and that topic only in your analysis.

For example I am studying basic mating patterns at the moment. That is fun and useful. In my analysis I focus on the question: did basic mating patterns occur, did I see them during the game and what is the name of the pattern. That’s it. This way the analysis becomes very useful in applying what I am studying.

Or if you want to improve your openings, focus on the first 5 moves of the game.

or any other topic. Focus your analysis to that topic.

Only use the engine if you doubt your own judgement. And wait together with all of us till the engine starts giving rating adjusted advise. Can’t be that hard to programm.

Similar joke with skydiving. You don't "need" a parachute to go skydiving. You need a parachute to go skydiving more than once O_o

There is no " like " button. 

If you like it, you have to quote it and say you like it!

Yeah, I was being figurative.

Not a bad game. I've seen and played worse

You should have continued Qxh7+ attack. You moved your king and gave the initiative to your opponent. He had good attacking chances and used them and you lost.

Finish what you started

Good Positional Chess, Planning & Strategy Books for Beginners and Beyond...

https://www.chess.com/blog/RussBell/introduction-to-positional-chess-planning-strategy

https://www.chess.com/blog/RussBell

I dont know how you are able to do all that on 30 minutes a day. Any one of those activities would take an hour or more for me....or maybe when you say 30 minutes a day you are just referring to tactics study? For me though even that takes much longer...... when I focus on tactics...for example, I normally work from a book as oppose to a website. I do puzzles in groups of 20. I normally do all 20 before checking the answers. I actually write down the answers and then check them at the end. The ones Im still confused about I analyze with an engine. The entire process normally takes a couple of hours. Oh another thing I do is analyze the problem from both sides of the board....meaning if the problem is white to move....I ask myself could black win if it was black to move instead? The other activities should also take a long time.....but if I were you I would focus on tactics.....you will get the most bang for your buck by studying tactics. My general advice is to slow down and make sure you are getting all you can out of your studying activities....think quality over quantity.