Hacker News new | ask | show | jobs
by aliston 1488 days ago
I'm also in my 30s and enjoy the challenge of chess, but have never been able to get beyond the basics. Sure, I can learn one particular gambit by approaching it like an algorithm, but I've never felt like I was able to grasp the high level strategy in the way that the author describes. There just seems to be too many permutations and tricks for me to memorize them all. Did you ever reach a point where the game come together? If so, has it been from memorizing very specific lines (kings gambit, if this then that) type of stuff or is it really more like intuition?
9 comments

You should not be approaching chess as a memorization problem. Yes, as you gain skill and familiarity, your brain will pattern-match on particular aspects of positions but outside of openings memorization will not get you very far.

The things you should spend your time and effort on are tactics and positional play. Tactics will directly help with pattern-matching on situations with direct and immediate impact: taking advantage of forks, overloaded defenders, x-rays, etc. (as well as defending against those things). Positional play will help with pattern-matching on higher-level concepts: getting knights to ideal squares, understanding the implications of certain pawn structures, how to leverage bishops' strengths and weaknesses, maintaining rooks on open files, restricting your opponents' space and movement options, etc.

For the former, tactics trainers are your best bet. For the latter, I think books are probably still optimal. I'm a big fan of Jeremy Silman as an author, and Reassess Your Chess (4th Edition) is my personal favorite book for learning these concepts.

Studying openings should be done sparingly. Learning the basics and general principles of one opening system as white, and one each for 1. e4 and 1. d4 as black can be useful in helping you always get to midgame positions where you understand the basic ideas, but I wouldn't spend too much effort past that.

The psych literature on expertise draws in part from studying chess masters, and this expertise in chess mainly consists in recognizing a large number of critical positions. But this is founded in a fluency with tactics.

The parent is doing probably the most effective thing at their level by focusing on tactics, tactics, tactics. Learn the basic tactical motifs, practice finding them in puzzles, and practice employing them in games.

When you are starting with chess, don't study openings, just learn basic opening principles -- how to develop your pieces. Learning opening lines is the least productive thing you can do right now. The same positions can arise from multiple opening lines, so you are better off understanding principles and learning tactics (and how to calculate).

Using brute-force, algorithms or rules is too hard for humans, there is too much going on at the same time. Even Stockfish didn't manage that against AlphaZero :) If you do want to have a more strict repertoire, you could find a few openings that you like and that play naturally for you. The way I choose them is by simply playing them and then see the first ten results. The first 5 games with King's Indian defence were wins for me and felt very natural, the first 5 moves are slow and after that there is a game going. I tried Ben-Oni, but I am too messy of a player and just mess that up, the same with the French. Ofcourse, getting to know an opening better, with the themes that belong to it, will make you better at it long-term.

Anyway, there are so many things going on, you need to feel some intuition or patterns on what is currently important on the board. Rules are so very much based on context, like, for black, c5 is almost always good while f6 is almost always bad. But there are many games with exceptions. And if you play against people of your same level games can just be fun.

Memorizing specific lines is only important at the highest level of competition. You should however have some rough idea of what you're trying to accomplish. Hanging Pawns channel[0] has videos about a lot of openings and their main ideas. I recommend Scotch Game for white and Sicilian Defense(any variation) for black. You can also use https://listudy.org as a tool to practice memorization(though I repeat, this is not worth focusing on). The most important thing imo is to play longer games(at least 15 minutes) and analyze them afterwards. There are some people who can improve by only playing shorter games, but from my experience, that's very rare.

[0] https://www.youtube.com/c/HangingPawns/playlists

I'm also in my 30s and have also been learning, and I recognise the frustration of trying to learn through learning specific opening lines.

What helped me was to stop trying to learn specifics outside of the first 3 or 4 moves, and just playing so much that I built up memory for specific cases where it was either a huge blunder or missed win, while at the same time sticking to trying to do the same things each game.

So I'd try not to vary my responses to particular lines so that I'd get the same situations repeatedly.

If I found myself consistently being uncomfortable, I'd change something earlier about how I'd respond, such as an early Nbd7 instead of trying to more aggressively fight for the centre in the Italian.

Overall, my three pillars of learning were puzzle rush (puzzle streak equivalent on lichess), playing more (~10k games over 2 years), and every day watching one of Naroditsky's speed-run videos, where he plays rated games from rating 400 up to around 2000 and explains his thinking. (They tend to end when he starts playing against too many cheaters).

That said, I've also plateaued around 1500 rating on chess.com. That rating is high enough for me to enjoy without feeling that every game is decided by silly blunders. I'm not sure that I have the capacity to get much higher, because my rating has barely moved for a year despite playing every day and trying hard to learn.

Strategic thinking at my level is still infrequent, but I feel like I won't progress my game until I get more of it. I've reached a level where just thinking over each move in turn just looking for tactics is no longer enough. To get better I need to think about "how to improve your pieces" which is another word for strategic and positional chess.

From watching videos, I think a lot of it is intuition. Quite often during his speed-run videos when faced of a choice of two seemingly fine moves, Naroditsky's intuition will take over and he'll say, "I'll play this move, I can't explain concretely just yet why but this other move just feels worse somehow".

There's an intuition about moves which obviously is incredibly strong in a bullet-specialist GM.

So part of it is just falling into the same tricks over and over you'll come to recognise them. Other parts is there is an intuition which as you get stronger at other aspects of the game will come to fruition.

Both. Intuition tells you which moves are even possible. It's like having a voice in your head that tells you "sacrifice the knight here", since the pattern of the pieces looks like it "has that move" in it. But you don't know how. This is where the algorithmic part comes in, you brute-force the most likely ways to do it by playing your best move, then your opponent's best move and so on. Your ability to win is defined by how many moves you see and how many moves ahead you can think.

Of course, it's more fuzzy than this. Intuition can be wrong. Sometimes the messages are along the lines of "there's some good move here", or "my opponent has mode some sort of mistake", rather than something definite. But when it all comes together, it feels like you're a pattern-matching wizard machine.

Tactics is the main thing to learn in the beginning, which builds the 'intuition.' Later study a couple openings like ruy lopez or giocco piano. You can learn a couple traps too and be on the lookout for them (fried liver and scholars mate in particular).

The game 'came together' for me when I realized how easy it is to squander any advantage you've gained. So in a game, you might be rocking it and winning. Suddenly, you've done the wrong thing and now your opponent is capitalizing on it and crushing you. That situation is very frustrating in the beginning. Once you recognize it, then you can try to be the one to not step in doo-doo by being very careful with every move.

I've thought that maybe it's a good idea to play Chess 960 because memorization is out of the question from move 1. This way you learn to analyze each position individually and hopefully understand the consequences of candidate moves. I think obviously it's important to learn specific openings and problems that have already been solved as well, but 960 can help develop "chess eyes" a little bit. Don't take my word for it; I'm bad at the game, but these are just some ideas that I've had.
I would recommend two things:

- Do lots of tactics puzzles—treat it like exercise, consistency is key. 1 to 3 puzzles per day is ideal.

- Read strategy books. Don’t worry about memorizing, just get the intuitive gist of things like moving rooks to open files, knights to outposts, keeping pawns on opposite color squares of your bishops, try to pile up pieces against the enemy king, etc.