[ThrustVectoring]

There was an interesting study I remember reading along these lines.

They took two equivalent classes of pottery students. Group A was told that they'd be graded based on the number of pieces made, while group B was based off the quality of a single submitted piece. Following the given incentives involved, group A made a bunch of pottery, while group B tried really hard at making good pottery.

What's interesting is at the end of the class, group A's pottery was better than group B's. Making a lot of pottery without caring about the quality of any individual piece is better at making high quality pottery.

The key takeaway I got is that you're generally able to magically become better at things you do a lot of. So if you want to get good at something, just do it more, and results will generally follow.

[jholman]

There was no such study.

The story you're repeating is a parable, published in a book called "Art & Fear" (as sibling commenters have noted).

There is no evidence in that book that it's anything other than a fabrication, argumentation from "just-so story".

I happen to believe the theory, personally. But the ceramics story is not evidence at all that the theory is correct. It's just a bald-faced assertion.

[kens]

I investigated this story a while ago and concluded it was a parable, based on how it was presented in Art and Fear. However, I checked with the book's author and found that it is a true story. Strangely, the class was actually a university-level photography class, but they changed it to ceramics in the book.

[areyousure]

425 days ago you wrote that you were highly suspicious the event ever happened: https://news.ycombinator.com/item?id=11439565 (I was also active in that thread because I have researched the story.) The same happened 324 days ago https://news.ycombinator.com/item?id=12108514 but then you updated one day later that you checked with one of the authors of the book! And here you relate the information again. I'm excited to learn more about this history.

Which author did you contact? Can you relate any details about the class? Where was it (which university)? When was it? Who was the instructor?

Thanks!

[kens]

I like to get to the bottom of things :-) I emailed author David Bayles last July to ask him about the story. He didn't give a lot of details but said it was a university level photography class taught by "a very experienced and sophisticated teacher", and apart from changing the medium to ceramics the story was the literal truth. I asked for more details, but didn't hear back and didn't want to keep bothering him.

[jholman]

Thank you for your service, and for your updates. :)

A few observations, I think, flow from this:

1) Still not a study (as I note you have noted elsewhere, but I'm drawing that to the attention of anyone else who reads this), but rather an anecdote. Interesting one, though.

2) Art and Fear was published in '93, so that means chemical photography. That means that even the students who were being marked on quantity would have some lower bound for quality (in a digital photography regime, you'd just hold down the shutter for an hour and submit 3000 photos and be done for the semester). By-hand developing also has a lower bound on how much work you can do and still produce a photo. It just seems a lot less game-able than the pottery version, which makes the story seem more believable. So that makes it pretty weird to change for Bayles and Orland to change the story.

For myself, I think it's true that if you want to be good at anything worthwhile, you need to do a lot of work. But it's also true that if you want to be good, you need to struggle with perfecting your craft, rather than merely churning out whatever. I highly doubt that either is adequate alone. Of course, if you merely want to be tolerable, in many fields some people can get by on talent (however talent works).

[areyousure]

Thank you very much for your response! I encourage you to keep getting to the bottom of things.

For what it's worth, I will go on record that I still do not believe the event occurred. :)

[oh_sigh]

This makes me wonder how many times, when people say 'I read/remember a study...', they actually read the study itself, and not some second-hand analysis (or, a story which masquerades as second-hand analysis). I think 'I remember a study' should be an anti-pattern for careful conversation.

[gwern]

It happens all the time. After being burned dozens of times by urban legends and grossly misdescribed studies, I started just always reading fulltext instead of relying on summaries or citations. And even then, you'll run into misleading to the point of lying summaries of past research throughout the literature. Life is hard.

[ThrustVectoring]

Yeah, I definitely wasn't careful with my wording there. I remembered reading about a study, not "reading a study". And even then, it was less a study and more "hey, a funny thing happened in my pottery class".

[oh_sigh]

No worries! I didn't mean to call you out or anything - I've used the phrase frequently myself, and was mostly just thinking out loud for how I can communicate with more care.

[gens]

My 2 moneys.

I worked in construction a while ago. I was always slow, pedantic, and the work i did usually reflected that (usually, some things benefit from being done fast). Everybody kept telling me that i was slow so one day i said "f it" and started working faster. The quality of work went down, for obvious reasons, and never got to the quality that was before (for most of it, some work like cutting things to size stayed precise). But if you, for example, paint a hundred windows really fast, you will probably not find many windows painted nicely. (mind you that nobody notices the little imperfections in construction, like, for example, a drop of white wall paint on the white radiator pipe)

Of course doing more means learning more, but so does thinking and experimenting.

Pottery, i never tried, but i assume it is more about the "feel" then anything else. While programming benefits a lot more from learning random stuff. It's easier to write a huge mess of code that works fine then it is to make a huge pot that doesn't fall apart.

All in all, why not bout ? Hack some 100 programs quickly, and write a couple programs with lots of preparation and research.

[jdietrich]

The heart of craftsmanship is understanding tolerances. Making something practically perfect is often remarkably easy if you throw enough time and effort at it. The hard part is making something just good enough. A master craftsman knows exactly what constitutes "good enough" for any given task and knows exactly how to achieve that goal with a minimum of time and expense.

[vosper]

Do you think there was any change in quality from the first window you painted after you decided to start going fast, versus the last one? Like, did you get better at painting windows quickly, even if they weren't up to the quality of the pedantic, slow one?

[gens]

It was a long time ago and i did do more of the heavier construction then stuff like painting.

Biggest thing i learned from doing things quickly is where precision matters and where not. For example if i'm painting the ceiling it doesn't matter if i touch a wall as i will paint that wall later (unless i splat a bunch of it, ofc), but it does matter if i touch the ceiling when i'm paining the wall. Of course painting things quickly brings the risk of spraying paint around, especially when using a roller. Painting things too slowly can also be bad if the paint is highly viscous, like with the paint for outside metal surfaces (railings) as it ends up being obvious where you stopped to detail something else. Or wall paint if you cba to do it in two layers so you mix thicker paint and do it in one.

With windows specifically (and doors and such) you would put the.. paper tape (idk how to translate it) (edit: masking tape) some 2 or more cm wide on the edges so you would have a nice margin of error. With paining fast there is a bigger risk of going over it, and that can be bad (oil based window paint on a wall is really bad, wall paint on a window easily goes away with a wet rag).

In most construction it doesn't matter if you are little off, it was just me being anal.

TL;DR I learned where it matters to go a bit slower and where i can go fast without consequences.

edit: Forgot to answer the question. Yes, i did get better. Mostly because i learned where i can go fast and where i need to pay more attention. Learning to not wave brushes full of paint around like a figure skater also had a bit to do with it. I did also learn that some things come out better by default if you do them fast.

[bcook]

As someone who spent ~10 years in construction (log homes mostly), only newbies and weekend warriors seem to have this mentality. Yeah, always do quality work, but unless you are an artist, you've gotta get shit done quickly.

I've tried to paint my own house and my professional painter friends can absolutely destroy me in both speed & skill. A major problem I have is smooth caulking. It takes me forever to shittily caulk a window while a pro can make a few quick swipes and be done.

[vosper]

Learning when you can go fast and when you should go slower, with more care, seems like a valuable thing to learn (not just in construction, I'd think)

[kaybe]

The word you're looking for is masking tape.

[cosinetau]

I think this is the best answer. There is no silver bullet to learning, and we need to get away from thinking there is.

On the other hand, coming from construction too, I would assume your coworkers were teasing you, rather than really fuss over the speed of the new kid. ;)

[MikeTheGreat]

I'd be interested to know if this is true generally, or just for people learning a new skill.

I.e., Making a lot of pottery without caring about the individual quality leads to more learning (possibly because you can try lots of new stuff without fear of failure, possibly because when you find something that works you can keep practicing it, etc, etc).

Once you know what you're doing I can't imagine that blasting through something in a half-assed fashion produces better results than carefully doing a thoughtful job.

[ThrustVectoring]

Definitely the latter, sorry if the story wasn't clear about that. Spending more time on an individual piece means that within the band of results your current skill is capable of delivering, you get a better result. Repetition, on the other hand, does a much better job of improving your skill.

[ghubbard]

While a nice story, I have been unable to find any citations that would indicate that this was ever an actual study.

[TheBiv]

it looks like it was a class https://blog.codinghorror.com/quantity-always-trumps-quality...

[draw_down]

It appears in the book "Art & Fear: Observations On the Perils (and Rewards) of Artmaking" by David Bayles and Ted Orland.

PROTIP: Google will often tell you these things!

[ghubbard]

And where in that book are the references to an actual study?

PROTIP: Books are often filled with creative writing.

[draw_down]

Fine, OK. I checked it out and there are no citations in the book at all (1993 edition), so yes it's a just-so story.

[ThrustVectoring]

I may be mis-remembering the level of rigor, it may have just been a thing a professor tried because they were curious.

[bluGill]

I'm curious what would happen if they took Group A just before they started their last piece and said there was a change and the quality of this piece would count. They have now learned a lot about making pottery and one would presume just need a little more effort to perfect it.

[prolifikosis]

I'm co-founder of Prolifiko and wrote the article you kindly commented on - thank you. Whilst the story you refer to might or might not be a parable - the thrust is correct. Check out this post about prolific nobel laureates - http://prolifiko.com/limitless/

[6stringmerc]

I love how often this comment gets upvoted because it's the perfect opportunity to point out usually 1% or so of any given cohort will have inherent talent that can be recognized as "Above Average" or maybe even "Exceptional." Throw a bunch of practice in, and yup, this is inherently a reasonable consideration.

Point being, most people don't have what it takes to cut it, so might as well aim for a really high investment of time and effort, and eventually plenty of walls will arise. For me, it took somewhere in the neighborhood of "5 Walls" before I could become cross-genre competent, and I've still got Wall #6 I'm kind of reluctantly ignoring (sweep picking, FWIW).

Most people give up around Wall #3, playing through extremely painful finger tips to the point of growing useful callouses (the kind you can stick a safety pin through and not draw blood).

A lot of this discussion tends to have trouble differentiating between "Imaginative Genius" and "Craft Expertise" and basically what we call "Success" is a labor of love, passion, or lunacy somewhere in the middle.

[pvaldes]

It seems that the test was 'average' value in 1 x B against 'maximum' value in 50 x A. Different measures (and not totally fair game). There is a bigger probability of doing something well just by chance if you repeat the test more times.

[nothrabannosir]

Not exactly comparable; the 1xB group had (took) a lot more time for their pot, and consciously spent more effort on making it look good. They actively went for max, while the 50x group didn't even go for it.

[pvaldes]

Making it look good is not the time that counts really here. Extra time is useful only in some parts of the process. Cooking time would be similar in both cases for each piece and this is the critical part.

"went for max" would mean here more ornamentation and bigger pieces. Is easy to cook perfectly a small pot, but the challenge is hugely increased if you are expected to produce bigger, more fragile pieces, with more parts that can fall off, and a structure that needs to carry more weight. 'A' can afford to play safe and small in this game, 'B' can't. You would need to be 'good' to be sucessful playing the A game, but would have to be 'a master' to be successful in B group.

No only they are different levels of challenge. Is a different level of learning. Even if A can do small and simple pieces, we can't say that are better than B in this sense (nobody had evaluated B doing small pieces). The opposite is also true. There is not guarantee that A are ready for the bigger pieces at the end of the course and wouldn't fall for the same errors than B.

[77pt77]

> There was an interesting study

Source please.

[baxtr]

Where is the data for that story?