How to (really) get better at chess: What they aren't telling you

Hi chess.com reader, my (screen) name is Linlaoda and I want to share what I believe is a common problem facing many chess players wishing to get better at chess.

First, identify yourself. Are you anyone of the following?

a) Has read a lot of chess books about opening, strategy, endgame - but feels that their knowledge of the game is not reflected in their rating?

b) spends hours everyday on blitz, but the rating always fluctuates just within a 100 point range?

c)  have no idea what the **** i'm talking about and just want me to get to the point?

I have been playing for several years now. Here are the things I did that I think led to the most improvement (and what didn't):

1) Brush up on tactics. The only book I think ever benefitted my chess was a tactics book. Not that middlegame book about pawn structures or that opening book about how to crush your opponent in the dragon. In practical play, at least 75% of amateur games are decided by tactics - several tactics with the last one winning the game.

2) studying (grand )master games. Whether use a database or an online database, it doesn't matter. Pick a grandmaster, filter his wins, and learn to play his playing style. Probably the most efficient use of time you can spend on chess studying

*for advanced players: studying games of GM vs 300 points lower rated players might be a useful and new idea.

3) playing in OTB (over the board, "real", rated, whatever you want to call it) tournaments

You aren't going to get better at chess just staying at home without getting any practice. If you are improving, you want to be able to see that reflected in your rating. Plus, you won't learn how to concentrate for 5+ hours while sitting on your bed reading an opening book.


Why did I want to publish this? Because I felt like it. I don't get anything out of this, no publishers are getting any money, and this might just make the competition on live chess a whole lot harder. Let me know what you think. If you disagree, tell me why. If you think I'm full of s***, this is the internet, I'm not going to force you to do anything. If you end up improving, send me a challenge on chess.com (correspondence).

Oh, thank you so much for reading. Good luck fellow woodpushers.

right on

Well, I think other books have value...

But other than that, IMO your advice is exactly right :)

what do you think of blacks response to e4... Nf6.  Alekhine's defense.  I had used it in the past but don't see it played much at anymore.

Linlaoda , Great push@!

well said. thanks.

Different players study and learn in different ways, but your suggestions are certainly a good starting point for most of us.

Here's my little "donation" to this subject

http://blog.chess.com/NimzoRoy/chess-opening-principles

Good post!

Tactics are important, but you have to know how to play to steer the position towards tactical play.  Thus what openings you use are also important. Once you start playing players 1800 and above, you are not going to be winning a lot of games because they overlooked some combination.Your positional play becomes much more important.

Also studying the endgame is extremely important.  Can't put the mule before the cart.

I think you're full of s.... (sense).

I do 25 tactics per day and will also join the mentor programme, which is not cheap, but excellent.

AGREED overall but...I'd at least check out the short draws to make sure they are in fact "GM draws" maybe a few are of theoretical interest anyway? 

Good post and very sensible, although I have developed different way of thinking. 

I would say:

1. play on intuition. A good move is in line with the position, hence should be very logical and is often the first move you think off.

2. always think that your opponent will make the strongest move possible.  If not, you are lucky, if true, then are you prepared. Often you see people playing a move in the hope that the opponent will play a helping move. Don't expect your opponent that he will do that. He is your opponent after all. :-) 

3. try to see the board divided in different areas. Ones in which you can move more easily then your opponent and otherwise. Try to get your opponent busy on your side of the board. Where you are strongest.

4. look on chess.com to matches between players that are some better then you. You might learn more from those matches then the matches from GM's. For me, GM's play chess at a different level. I might think that I understand their games, but do I really? If so, why am I then not a GM?

5. learn from your own games, particularly the losses.

6. the current game is learning you how to play in the next game.

I use a combination of tactics trainer and some live games in person on my computer, taking turns. I'll set the game up in Arena and analyze the moves with a chess engine afterward to see if there were any glaring mistakes on either side that could have gone very differently. Also play against computer opponents with various weakness patterns. Working fairly well so far.

The thing with chess is that after a point, it's not about book knowledge. There are only so many variations of chess tactics. You can try brute force memorization of all possible permutations but that's a lot more effort than I think it's going to take. You need to get experience with a vast array of positions and tactics and learn to recognize the principals of the correct moves.

Tactics themselves seem to be little more than a continuation of positional development that open up due to an error on part of the opponent. Seems plausible to learn positions from learning tactics and analyzing the reasons those positions went wrong.

Ultimately, it's going to come down to neural architecture. Once you learn the way tactics and positions work, it's a matter of training your brain to correct errors in pattern recognition and reasoning. Takes time, effort and repeat exposure in order to discourage miscalculation while reinforcing the correct approach to calculation.

Any time you catch yourself looking at a move or position and thinking an alternate is 'way stronger' when in fact it's much weaker or a blunder without realizing it, you've found what you need to work on fixing. If you simply memorize the correct response without fixing those points of error, you're going to keep making similar mistakes.

Then, what about learning how to make alternate moves that are not the strongest technical move for yourself and still win with them, without your opponent making any additional mistakes? It also strikes me as a weakness to see 'easy' or 'obvious' moves such as a mate in 1 without noticing the 28 other ways to force mate with completely different approaches. Always taking the shortest path when you can tell other ones will still win is going to reduce the variety of your experience.

Woah thanks for all the responses! I'll try to respond to most of the posts in one post:

@wafflemaster: sure, definitely. For example, My System can teach one about outposts. However... while it's useful to know about outposts, it doesn't lead to direct rating points. (a knife can't cut a mushroom, a person with the knife can learn how to cut the mushroom *better)

@Estragon: I disagree that you should not filter out wins only, if only for the purpose was that this post was not for the professional player. However, this one is more complicated. If I choose a player, i.e. Garry Kasparov, and filter his wins, then I have probably 1000+ games to look at. That's enough to keep me occupied for a while, and after that I can just switch players. While it is true that you might miss some amazing draws, the number of boring draws that one gets makes it not as worthwhile as learning from the wins.

That said... advanced players may wish to learn how to steer a worse position towards drawish endgames, so maybe only then studying draws are useful. But for the common chess player still trying to reach 2000, then learning how to bash your opponent tactically (strategically, etc) is going to be most useful.

@ JamesSneller : Actually I think that you can reach an extremely high rating with just tactical prowess. Your comment " you won't be winning as many games because they overlooked some combination" is taking tactics to only be referring to middlegame puzzle like positions. However, being able to see a continuation that leads to a winning endgame is also considered tactics (calculation). Thus, studying tactics (perhaps if I just say calculation, endgame, "thinking from the end") WILL be enough to bring a player to 2000+. All GMs are tactical monsters, but not all are great positionally.

@LoekBergman: Thanks for your input! However I was not constructing a way of thinking, but instead a study plan ;)

@Sociopathy: right on, but another easy way to improve tactically is to continue learning more tactical positions. Thus, the argument for changing tactics books before tournaments.

Thanks for all the comments! Keep them coming!

This is a very interesting thread that the OP has started here. People that are posting here are bringing up some very good points.

Playing through master games is invaluable, but there's much more benefit in going through instructive annotated game collections than playing through games in a database without annotations.

There are lots of instructive annotated game collections, such as Chess: The Art of Logical Thinking, Chess Secrets: The Giants of Power Play, Understanding Chess Middlegames, Instructive Modern Chess Masterpieces, and many more. A really great way to train is to slowly go though a game while trying to guess the moves, and comparing your move to the actual move and your reasoning to the annotater's reasoning. You can't do this with unannotated games, because it's often not possible with just an engine to determine why the master's move was better.

I do agree though that tactics training is most important until you stop making lots of tactical errors, but playing through master games and learning the basics of positional play is also very important. Tactics arise naturally from a superior position, and if you allow your position to get worse and worse, it gets more and more difficult to prevent all the opponent's tactics, until at some point, (s)he has tactical threats that can't be met.

For working through master games and trying to guess the move, there is a nice program called Guess The Move, and another program called Chess Hero can also be used for training in a similar manner when configured appropriately.

Hi linlaoda,

I agree with your post. I'd like to add that endgame knowledge also is really important. Knowing to what won/drawn endings you can trade down to, adds a lot of potential "tactical" probabilities.

Most people will disagree with me, but I believe a great way to improve is to study openings. Of course NOT memorizing moves with a database, but instead following a good opening book that gives explanations of WHY you choose a certain move. I'm gonna be even more unpopular and say that I much prefer repertoire books, because the move suggested have a consistent strategy/themes.

The point is that studying for an opening repertoire you learn strategical patterns, you improve your positional understanding. You get familiar with many kind of positions that are likely to arise in your games - even on different openings! For example studying the Yugoslav attack gave me a good idea on how to attack (or defend!) a fianchetto structure. Studying the Ruy Lopez taught me the power of a mobile pawn duo in the centre, etc.

I honestly don't find much benefit in going over master games, even if annotated. The reason is that in chess you have to make choices - a lot of choices. So what's the point in seeing what a GM did in a particular position when playing in 5 other different ways would have been fine? And note that how a GM played that day could have been influenced by a ton of different factors, for example: what he expected his opponent's preparation to be, or because and engine told him that he will be slightly better after the opponent's perfect play for 10 moves.

Because of this I believe going over games of old masters (e.g. Petrosian) would be much more useful, but going over what Carsen or Kramnik is doing nowadays wouldn't give many results.

This is just for solo preparation of course.

Playing a slow game with a stronger player and then analyzing it together is surely the best way. Everytime I play OTB a 2000 friend of mine I feel I improve of a couple of ELO points!

Quoting plutonia "but instead following a good opening book that gives explanations of WHY you choose a certain move."

Could you name some titles please? 

Going through master games and comparing their move to the one chosen.  Making assessments of imbalances and weaknesses, such as a weak square, diagonal, etc.  One should keep the notation pane off and type one's thoughts into a Word document with one's calculations and reasoning.  However, some positions have no good forcing continuations and those ones are harder to calculate.