Play mostly the form of chess you want to be good at
Assuming you want to be good at slow chess played over a physical board then play mostly slow chess and do as much as your study and training as possible at a physical board (for opening study using a screen is an acceptable option).
There are two points here. Visualisation is different over the board and on a screen. If you mainly use a screen you will just not see things as well at a board and vica versa. Don’t take my word for it – the world’s leading coaches tell you this, e.g. ex world champ Anand tells his students to always solve puzzles at a board, not on the screen.
Secondly, slow chess requires both shallow and deep calculation and careful decision making whereas blitz requires virtually only shallow calculation and quick decision making. You won’t learn the deep calculation skills and imagination you need for slow chess by playing blitz. In fact quite the opposite is true; playing too much blitz can establish a pattern of shallow thinking in your brain that stifles imagination and deep thinking.
The remainder of this article assumes you want to get good at slow over the board chess.
Training
Do puzzles
Before I go any further let me say that by this I do NOT mean doing anything like puzzle rush on chess.com. All that does is make you quicker at spotting patterns you already know. It will not improve your calculation skills and will barely improve the number of patterns you know. It might be fun but In terms of improving it is not a good use of time and energy
The puzzles you do should be hard for you. In fact they should be hard enough that you fail to solve a lot of them despite trying hard. That might be dispiriting but it is essential if you really want to improve as much as possible.
This is what world renowned trainer Artur Jusopov has to say about solving puzzles:
- You should set the puzzle up on a physical board.
- You should try hard to solve the puzzle for at least 5 minutes (you will often need longer) without moving the pieces, working out all the necessary variations. If you think you have solved it, you should feel sure you have everything worked out. Write down the key variations.
- If you can’t solve it without moving the pieces you can try again for a further 10 minutes but this time you can move the pieces.
What you should NOT do is this:
- Give up before you have tried really hard.
- Guess, e.g. think to yourself something like “Oh, the answer is probably X, I expect the variations will all work out from there.”
- Do loads of easy puzzles you can solve in a few seconds.
The puzzles you solve can be:
- Positions from real games, especially middlegame and endgame positions. These might be based on small hidden tactics, spectacular combinations, the correct positional approach in a middlegame, the correct winning/drawing method in an endgame.
- Endgame studies.
- Composed puzzles (simulating a game position).
- Composed problems (e.g. mate in 2). If these do not look like a real game position they are in my opinion the least useful but I note that the Polgar sisters used them as part of their training when they were young so they perhaps have a place.
The positions to solve might not be presented as puzzles. For instance, a long time ago, I was reading a book on the middlegame which gave example after example of IM and GM tactics. I covered up the text below each diagram and treated it as a puzzle to solve.
If you get the right mix of puzzles you will improve your:
- calculation ability
- imagination
- positional sense
- planning.
Analyse your games thoroughly
Assuming you are serious about improving then you need to put some serious time and effort into this. There is a really good method you can apply to this:
- Analyse your game (at a board of course) without using an engine.
- Try to recall your thoughts during the game and try to work out what you got right and what you got wrong. Try to work stuff out in your head and then check it by moving the pieces.
- Record your conclusions as you go along either on paper or on a computer program (such as Chessbase), noting variations and evaluations. If you use a program to record your findings you will need to be switching back and forth between the board and screen – work at the board and record the conclusions on the screen. You will need to resist analysing on the screen and absolutely do not turn the engine on. In practice can be easier to work with pen and paper.
- For anyone but a complete beginner this should typically take somewhere between one and five hours for one game – don’t skimp on it, try really hard to get to the truth. You will learn more by analysing a few games thoroughly than all your games to a shallow level.
- Only when you have finished doing the above for the whole game should you turn the engine on. Use the engine to see how accurate your post-mortem was and revise your conclusions. If an engine tells you a move is good see if you can understand why it is good.
This is tough and requires real discipline. It is all too easy to consult an engine before trying to work stuff out for yourself but that way you will get only a fraction of the benefit.
Analyse grandmaster games
Get some unannotated grandmaster games and apply the above method. An hour to two hours per game is reasonable before turning on the engine.
Analyse middlegames arising from one opening variation
This is an area where you can blend study and training. Pick an opening variation you play then:
- Use opening a book/video/Chessable course that explains the middlegame ideas that result from the opening – this is study
- You might be able to find a middlegame book that does the same thing (Chess Structures by Maurico Flores Rios does this well for some openings, less well for others) – this is study
- Get a few grandmaster games with that variation – if annotated study them, if uannnoated train with them
- Go mentally exploring on your own in the middlegames arising for the opening variation– this is training. Can White do this, what if Black does that... is this move better or worse.... is this position better or worse? Work at a board without an engine for an hour or two or three... and then turn the engine on. Does it agree with your conclusions? If not, can you tell why not? Have you just mis-evaluated a set of positions or is the difference based upon some tactical point it spotted that you didn’t?
This is a good area in which to go back and forth between study and training.
Spend time with stronger players.
This can help you in a number of ways:
- If you are lucky they may analyse your games with you. An engine will be more accurate but cannot explain some stuff to you like another player – e.g an engine may see five different ways of winning an endgame and their methods may not make much sense to you, but a strong human might say something like “just cut the king off like this and then it’s easy” or “that pawn is really important, you must hang on to it.”
- Watch them analyse their games. Even if you are not participating this can be highly educational. You may pick up on positional ideas, improve your evaluation of some types of position, and whilst you are unlikely to directly improve your tactics you may be inspired by seeing the depth of their calculations or the imagination behind their piece sacrifices etc.
- Play practice games against them, preferably slower ones rather than blitz.
One way of spending quality time with a stronger player is to get a coach. Of course, this is very expensive but if you get a good one doubtless it is a very effective way of improving.
A good coach will:
- Impart their knowledge directly
- Guide your solo study
- Guide/direct your solo training.