Notes on Puzzles

> It’s hard in real life, too: vanishingly few people are meta-rational enough to try really hard to falsify their own ideas. Your brain really wants to find reasons to support what you believe.

I don't think he goes with "meta" deep enough. It is great for engineering problem solving mindset But it is also a good way to end up like underpaid post doc, who needs second job just to pay rent.

And this type of advices are usually coming from someone who "made it", has its own house and is practically retired. Very impractical and harmful (to some extend) for young minds.

Practical implementation for young person is not "falsifying" and trying again again. But coming with solutions that takes minimal time, is good enough and comparable to coworkers who work on the same salary. Time you save can be invested into education, family, hustle and so on.

"Real life puzzles" are too open-ended and have too many levels to really be called puzzles. A puzzle has a closed set of rules that usually gives you only one level on which to solve the problem. Many interview questions could be described as puzzles. A "real life" programming task has a bunch of different levels: what's the real problem the customer wants solved, is this the problem this customer wants solved first, do we have a bigger customer with a bigger problem you should be working on instead, can the problem be solved without programming, should the problem be solved in a different system, are there other people on the team who solve problems like this in their sleep and they'll give you the answer in five minutes if you describe it on Slack? If it does seem like you need to solve the problem, what's your level of confidence that with investment of X time you can solve the problem, for different values of X, and given this information, does it still make sense to try to solve it?

What makes puzzles relaxing and reassuring is knowing that there is a solution, and that you know all the rules. Also, you know that you'll recognize the solution when you get it. Real life rarely gives you that reassurance. With a real-life problem, you don't know if there's a solution, and even when you have one, you can't know that there wasn't another solution that would have been much better, because of possibilities you failed to consider. The only way to turn a real-life problem into a "puzzle" is to strip away the open-ended real-world context and present a subset of it that can be described in a closed form.

Many top chess players have considered themselves athletes, and if you've ever tried to calculate under pressure at a board for a few hours, I'm sure you'll agree it is an exhausting activity. Fischer engaged in athletic training when preparing for tournaments, for example, to aid in maintaining mental focus (he wasn't unique).

It would be silly to say that 2400 Elo indicates you can run a four-minute mile, or that calculating 6-ply in a closed position burns the same calories as running a block. If the claim is 2400 Elo tend to have similar vascular flow in the brain to people who engage in aerobic exercise or something of that nature -- maybe?

I play chess (poorly for the time spent on it) and I'm also a reasonably successful founder of a couple software companies. I find my struggle with chess is that I want to act intuitively, something that has served me well all my life in other avenues. But the board doesn't lie and if you don't think thoroughly you will get punished.

I have the capacity for it, I can think thoroughly in puzzles and perform much better there than my on board play but I just struggle so much with the discipline during regular games to falsify my moves. So much so that I've mostly given up on trying to improve despite really loving the game, it just grates on me. I know I could be better but I lack the discipline and I guess I just don't want to exercise that discipline in a game.

Anyways, great article.