Hacker News new | past | comments | ask | show | jobs | submit login
Intelligence and Practice in Chess Development (chessable.com)
47 points by MAXPOOL 6 months ago | hide | past | favorite | 17 comments

I played chess in my high school team, but then switched to Go at university. I felt like my mind opened up to more universal concepts of cause-and-effect and harmony with Go, as opposed to chess which is so narrow and constrained in comparison.

With Go it felt like something "switched on" in my thinking that allowed me to better evaluate life situations and to act with a better appreciation of possible consequences.

When AlphaGo showed superiority, I was dismayed that the "last frontier" of human intellect has been surpassed. Chess being easily defeated long before was one of the reasons I began to disregard it and rather consider Go.

After AlphaGo I began to appreciate that computer superiority doesn't mean an end to human play, as it is like an art. Maybe playing a game of chess or Go is like painting a picture, one with a distinct humanness to it. There is something lifeless and mechanical in GAN-produced pictures, which makes one yearn for a human touch. Although newer generatives like DALL-E are starting to encroach upon this too.

Also, a game is like a vigorous exchange of ideas with a social aspect to be enjoyed.

Anyway, I have to admit that I still have reservations against Chess, as I see it maybe like drawing with crayons on the playground. Each to their own, but I wonder when they will take up a paintbrush and canvas and upgrade to Go.

I have a similar history and opinion regarding chess and Go (maybe not quite as crass). I "liked" chess, whereas I love Go. With Go I really had to learn to picture the next steps on the board, because otherwise you couldn't make more than the easiest steps. Whereas in chess I didn't quite felt the need for that (well, I was a very weak player) but also I simply didn't have the mental capabilites. Only because I loved Go so much is why I really tried to learn to think visually (very tough for me).

The difference between Go and Chess for me boils down to the following:

Whenever I see a great chess move, I just think: Ah, yes, that's it.

Whereas when I see a great Go move, I think: Wow, that's a great move. And smile.

It can be really mindbending. At least for me. And that's also why machines being better than humans doesn't change the experience of playing and that's what counts in the end. If people have tried chess and Go and appreciate chess more, so be it. In the end both are only games and should only be played if it's fun.

After AlphaGo I began to appreciate that computer superiority doesn't mean an end to human play, as it is like an art. Maybe playing a game of chess or Go is like painting a picture, one with a distinct humanness to it. There is something lifeless and mechanical in GAN-produced pictures, which makes one yearn for a human touch.

Unfortunately, chess players at least have now been forced to acknowledge the beauty and seeming artistry of modern chess engines - which indicates the "humanness" you cite is probably a perception we will need to re-evaluate.

I'm an AGI-sceptic wrt current approaches including the Alpha series, so am not suggesting there is real intelligence there, just that we perhaps have historically assigned a near-spiritual reading to instances of human genius that (at least in the game-playing form discussed here) has now come under question.

I’d like to get into Go. Is there a good website venue for Go, like https://lichess.org for chess? Many thanks.

E: Link format

https://online-go.com/ is the main & best web based server (and open source like lichess as well)

Many thanks!

Related: When I first got into chess, I especially enjoyed Lichess's 'learn by playing' feature [1] that covered fundamental and intermediate concepts. I'd love to find something similar to help learn Go.

[1]: https://lichess.org/learn/

https://online-go.com/learn-to-play-go is perhaps similar but doesn't go very far or in depth

Thanks for this. I'll give it a try and play around with the rest of https://online-go.com too.

Haha, I kind of had the opposite history (played Go first, got more into chess later and found it more interesting)

Still love the community and experience of both games though

Is this a parody? Most of the Go players I've met are thoughtful and well-adjusted - they don't resemble this stereotype at all.

AlphaGo vs Humans is not a straight forward comparison. AlphaGo trained on a few orders of magnitude more games than any human.

This article agrees somewhat with my pet model of the relationship between intelligence and achievement. That is, (assuming you pass a certain threshold; you aren’t going to be a particle physicist with an 85 IQ) natural talent is only the dominating predictor of performance twice: when you don’t have that much practice at something and when you’re bumping up against the current limits of what can be done (e.g. when you’re late in your chess career or have seen all the gains from practice you can).

Obviously intelligence matters when you’re starting out in a field: if you can learn things more quickly and more completely than the next person, you’ll have a head start that can’t be beat with raw practice. It’s similarly self-evident that it’s necessary when you’re close to the “end” of your field: it’s /hard/ to push boundaries and do things no one has before. Mark Kac’s quote about “magical geniuses” comes to mind.

I think people spend a lot of time in the early stages of a field (I graduated with a math major and I’m still just starting out in math!) and I feel like this tends to bias people towards saying that intelligence is a dominating factor for performance over a lifetime. However, I think this neglects the fact that 99% of a given field lies somewhere between the absolute beginning and the absolute cutting edge.

This is something I’ve noticed: nearly all the high achievers I know are reasonably intelligent people who are crazy passionate. I knew people in my friend group at college who went on to get {Masters,PhDs} at {Stanford,MIT} and out of all 4 of them only one was exceptionally smart IMO (over 1/1000 rarity).

This also leads me to the conclusion that most people you see who seem crazy smart are probably a fairly normal level of smart with effort added to taste. The argument for this isn’t ideological - it’s statistical. There are simply so many more above-average people than there are exceptional people that most of the really high (but not world class) performers come from the former group.

There is https://listudy.org which offers spaced-repetition based chess training. I really wish the project received more contribution though.


Hey that's my project :)

I also wish I had more time to spend working on the site.

>The not very surprising conclusion is that both intelligence and practice are important factors in chess skill development. On their own, however, they can only explain certain aspects of development. For example, practice has its strongest effect in the beginning of expertise development, whereas intelligence’s strongest effect is at the peak and in the later stages. Together, they explain the changes across the whole life span much better than they do on their own.

I'm confused about the status of this study and the ELO graphs. The post gives some detail about the methods for measuring intelligence, but stops short of describing the sample group and calls both IQ 100 and IQ 120 groups "hypothetical". Are those graphs real data, or just descriptive of the authors' current hypothesis?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact