Free online Chess server. Play Chess now in a clean interface. No registration, no ads, no plugin required. Play Chess with the computer, friends or random opponents.
Sign in
  1. Forum
  2. General Chess Discussion
  3. A Great Article on How to Improve Quickly at Chess

Hi everyone. I published this article yesterday and got ten thousand views with minimal marketing via other channels. Thought I'd share it here too since it's Lichess specific.

A very interesting and complete article. Well done!

And now, you are 1800ish again so I might formulate this as an option :
- you reach 2000 because of all you wrote in your article and then with some bad luck, you lost your rating to go back where you were previously
- you just had a good streak and reach 2000 but there is no relation with all you wrote in your article because you just over performed for some time.
Now you are back at your real level.

Blog has few errors on how computers think. Computer use alpha-beta (which gives same result as pure minimax) which consumes less computing power that mini-max. They do not look just few moves ahead. On phone stockfish looks 19 moves ahed with 2 seconds from the clock. Nor they are materialistic anymore. They can out smart 90% of chess population on positional concepts

Monte-carlo tree search in necessity when there is no evaluation function i.e it is replaces with simulations to end of game. Alpha-beta would be better just not possible. It is monte-carlo only to avoid evaluation function. Though is has evaluation function to propose moves to simulate.

advice on playing sounded pretty sound. 100% concentration is obvious but surprisingly hard

Useful article, chess is definitely a lot about mental fortitude, as with many competitive 1v1 games.

Very enjoyable to read.

It is an inspiring article.

I agree with:

"Sticking to a few openings is better than knowing a variety of openings. " The strongest players played a narrow repertoire of openings. Fischer: Ruy Lopez with white, Sicilian Najdorf and King's Indian with black, nearly always and against everybody.

"You might be surprised how early you have to start focusing. It’s often as early as moves two and three (if straying from openings you know well), you should start thinking about what type of positional game you want to achieve. Tactics will start showing up as early as move five."
Yes, play slowly in the opening, even if you know the moves. This will help reach a state of concentration.

"Always use up more clock than the opponent early on. If you get an early advantage, you won’t need much clock later to convert a win."
Yes, when two equally strong players play, the one that uses more time wins. It does need good endgame technique to convert the won position to a win in little time.

I disagree with:

"Look through your game history and count what percentage of your games is decided in the end game." This is at lower play levels only. Top grandmaster games are decided (won or saved to a draw) in the endgame.

"You don’t need to know endgames involving bishops and knights and rooks. They’ll rarely come up." Rook endings are the most frequent endgames. Knight and bishop endgames are more frequent than pawn endings. At lower play levels ignorant players tend to trade off all pieces leaving a lost pawn ending. Better players avoid trading pieces when they are at a disadvantage.

"The problem with studying end games is that there are so many different situations that whatever you study is highly unlikely to show up in game. Just capitalize on early and mid game blunders so you won’t have to ever play an end game."
This is at lower play levels. Good endgame skills and confidence that you have good endgame technique to convert a won endgame in little time allows you to spend more time in the opening and the middle game.

"Focus 100% even during your opponent’s turn and start calculating out your next best move in response to what you predict your opponent’s move will be." Never calculate on your opponent's turn. Use that time for strategy deliberations. It is pointless to predict a move and then calculate that. First he might play another move, when you have wasted your energy. Second as the move has not been played, you have to imagine it, so you have to think like 1 ply deeper which is harder and produces more errors.

"If you want to improve, you should be analyzing 100% of your own games. " You should analyse your losses only, to impregnate your mistakes. It makes you humble. Analysing wins is fun, but does not help you in any way, it makes you cocky.

"An extension to the guideline above is that it’s usually a good move to pin your opponent’s knight with your bishop if they already moved the bishop that can unpin their knight."
No, as the bishop is worth more than the knight, the opponent may gain time by kicking the bishop with pawns: ...Bg4 h3.

"Fianchetto is good as well. It is well positioned to pressure the center and also protect your king or attack the enemy king." Fianchetto is strong, but weakens the position of you king. against ...g6 and ...Bg7 attacks like h4-h5 or f4-f5 or Be3 Qd2 Bh6 are possible. The fianchetto is strong, but weakens the position of your king. Both effects cancel out.

"What about studying games from grandmasters? Is that better than or a good complement to studying your own games? Well, grandmasters don’t make many mistakes but you do. Your opponents do too. So it won’t be too helpful since you’re learning from fewer mistakes. You won’t know what grandmasters are thinking when watching their games anyways."
Some annotated grandmaster games indicate what he was thinking. You learn how to play strong moves. I agree your own loss is more helpful than a grandmaster game, but a grandmaster game is more helpful than one of your own wins.

"Chess books on theory and strategy probably aren’t too helpful either."
Endgame books and strategy books like Nimzovich' 'My System' are really helpful.

from what i hear Adderall can boost your performance a couple hundred or so

How I got to 2000 by a 1700 player ....

This topic has been archived and can no longer be replied to.