Things I wish I knew when I was a child about the game of chess.

“Hard work is a talent” Garry Kasparov

Recently I took on playing chess a bit. I used to play chess as a kid and that’s useful as I know some basics, but as a grown up I find myself approaching it differently, noticing things I wish I knew as a child.

There are not so many squares. Sure, there’s 64 of them, but (with reference to the starting position):

  • there is a white king on e1
  • and a pawn in front of it on e2
  • the pawn can move to e3 or e4
  • and then the same for black king and pawn
  • and then the same for queens

and we’ve covered a quarter of the board.

The same applies to moves: given a position, there are not that many good moves. Especially at the beginning, most pieces can’t event move, apart from pawns and knights, and not all the moves are good, hence the study of openings.

Moves fall into three categories:

  • preserve balance
  • tiny change of balance
  • decisive blow

the last one usually happens either because the opponent did a mistake, or after a chain of tiny changes of balance.

The chain of tiny changes of balance is achieved by having a plan.

Some positions have multiple choices that are relatively balanced, but sometimes the there is a single next good move.

Learning openings can be done in two different ways:

  • go through each case encyclopedically (the know it all version)
  • or know little (focus on the core ideas an principles)

or more likely something in between, but without having an idea of openings you start on a back foot.

That’s because in a given position one player usually has initiative/tempo, further restricting the choices of the other player. E.g. if the white starts with e4, for black out of the not so many moves possible at the start, also Nf6 is not such a good idea, white can start chasing it with the pawn on e5, leading to a tiny change of balance favouring white.

If you’re not aiming for world champion, instead of trying to learn “good” openings very deeply, it’s a productive to learn some “not so good” openings, that theoretically loose advantage, but know them deeper than your opponent and then profit on the opponent’s mistakes due to unfamiliarity with the obscure opening.

Playing puzzles is an effective way to improve results at beginner/intermediate level.

Computer plays silly outside the main scenario of doing the best move:

  • When the computer is restricted to play as a lower rating, it sometimes randomly choose a “not the best” move. To the computer algorithm, all the bad ideas are equally bad, but a human of a certain level will not play certain bad ideas.
  • In particular if I tried to play end-games e.g. my king vs computer’s king a pawn, the computer might not put a fight because it looks at the position, it assesses that it’s a draw if played well, therefore it happily just looses the pawn to get to the draw, instead of trying to put a fight.
  • When I played bad moves, the computer assumed a lower rating, adjusted for it lowering it’s rating, but when I sprung back, it did not adjust up accordingly.
  • I also saw online a case where the human player was doing nonsense, non-aggressive moves, the computer then reciprocated instead of continuing the attack.
  • At endish games, some computer moves are weird, especially when playing computer vs. computer, it plays such that some advantage is only visible many moves later.

There is a lot of history in chess. It’s so humbling to see how top players played computer moves two centuries ago.