Dan Luu

There are people who try to distill those proverbs to laws that can be applied without any ambiguity. I believe those attempts fail, or only capture things that are uninteresting.

The way to progress is to develop pattern matching skill, so that you can sense what the appropriate proverb/pattern to apply in a given situation.

I'm not arguing against clarity or simplification. Just that I think that *some* of the vague talks are worthwhile.

I don't disagree with this approach in principle, but for this to work, I think you need something like you get in chess tactics puzzle books: a ton of problems that let you hone your intuition. It seems difficult for a talk to provide this.

IMO, another difference is that we have a decent idea of what works in chess and basically no idea about what works in programming. (Again IMO,) this means a justification of the idea is required in a way that's not true in chess and this is almost never provided in talks.

I’d dispute that about Go (I think it’s true of chess as well, but I know more about Go). Of course, we observe that a given player wins or loses, or is better than us. But we don’t necessarily have an adequate reasoned justification for why a move is good.

So the expert says “don’t play here, play here”, and the normal player says “it doesn’t make sense to me”, and the expert can’t really say anything convincing. They just have to say “trust me”.

Or maybe they say “if you play here, then this move and this move and this move happen, and then it’s obviously bad”, and the average player says “actually, that looks good to me”. (Or perhaps they say “ok”, but don’t know why).