A thought-process testing opportunity

by [anonymous]1 min read22nd Apr 201328 comments

46

Exercises / Problem-SetsForecasting & PredictionPhysics
Personal Blog

If you haven't seen the video of a wet towel being wrung out in space yet, it provides a great opportunity to test some basic rationality skills.

Skill #1: Notice the opportunity. (I failed this test. I had a fuzzy, wrong idea about what would happen. I didn't notice the fuzziness of my own thinking until after I watched the video, when it was too late to apply basic rationality skills. I'll never know if I could have made a correct prediction.)

Skill #2: Enumerate possibilities.

Skill #3: Incorporate prior information. 

Skill #4: Making clear predictions.

Skill #5: Understanding why/how your prediction failed/succeeded.

Skill #6: There may be some things you predicted and some you didn't. Don't forget to notice the partial failures along with the partial success.

 

Experiment:

You are on the International Space Station. You get a towel soaking wet, then you wring it out. What happens?

Video of the experiment

 

If you haven't seen it, don't scroll into the comments.

Don't click the link until you've thought about it!

28 comments, sorted by Highlighting new comments since Today at 9:53 PM
New Comment

Got it, though if I got it wrong I would have been horribly ashamed considering that I'm a physicist working with fluid-handling systems in cases where Earth gravity is negligible.

I predicted drops would fly off as the cloth was twisted. I was completely wrong.

Me, too, though I thought of them as blobs rather than drops.

They probably would have flown off had he twisted it faster.

Yes, but then one could predict - from the abundance of wiring in the background - that he would not twist it fast. (I seen video before this thread).

Likewise, and I still think they would have if he'd rapidly twisted it as the momentum of being expelled from the towel would have overcome the surface tension.

However, if I had thought longer, I might have considered that this would have been a huge pain, and thus accurately predicted that the astronaut would do it slowly instead.

Saw the video before this post, thought to make a prediction, and was correct! :D

This seems like the sort of exercise we could pool in some data repository thread, or a lesswrong wiki article.

I encourage everyone who comes across similar videos to post them as exercises to the open thread, or a dedicated thread for such thought-process testing.

(Written before watching)

Skill #1. I would have failed this if I had not had the opportunity explicitly pointed out to me.

Skill #2.

a. Water forms a cylindrical shell around the towel.

b. Water is pushed into the parts of the towel which are least compressed, but does not exit from the towel.

c. Water flies off from towel equally in all directions perpendicular to the towel axis.

d. Water adheres to towel, in a spiral pattern following the way the towel is wrung out.

e. (Something else I haven't thought of.)

Skill #3. I'm not really sure what you mean by "incorporate prior information", and how it differs from skill #2/4. To generate the possibilities in number 2, I used my model of how wrung towels behave on earth, and my knowledge of how things behave in space, mostly from Don Pettit videos (which I assume is also the source of the linked video). But I don't really have clear physical intuitions about the situation, nor do I have enough knowledge of physics to be able to work it out.

Skill #4. Of my options in number 2, I'd go for c. 20% d. 10% a. 10% b. 5%. ( therefore e. 100-45=55%) Note: I was loath to put down numbers, because I have basically no way to calibrate those guesses, but in the spirit of the exercise, those are as close to my true expectations as I can manage. On reflecting, I do actually expect greater than even odds that none of my answers are right, so even though it feels like cheating to say " 55% something I haven't thought of", I'm sticking with it.

(Video watching time) Skill #5. I can see in a very superficial way why it works the way it does. I don't think this equates to real understanding though.

Skill #6. So first of all, it wasn't Don Pettit. So I lose points there. :p Secondly, the answer is somewhere between my prediction a. and my non-prediction e. I don't think I can chalk that down as a win. Partial success: something I hadn't thought of happened (water moving onto the hands), which I had predicted at 55%. Partial failure: something I had thought of happened (water forming a cylindrical shell around the towel), but I had only predicted it was 10% likely. Neither of these are clear cut, you could swap the labels of "success" and "failure" and not be inaccurate…

Thanks for the puzzle!

ETA: line breaks, formatting

[This comment is no longer endorsed by its author]Reply

Complete failure here: I guessed the water would fly off at different directions and perhaps from discrete points in the towel.

Prediction: The water would bead up in one big drop or bubble, because surface tension pulls water into spherical shapes in freefall. (Alternatives considered: the water flies off; the water streams away; nothing happens.)

Reality: No, not so much.

What I got wrong: I focused too much on surface tension and forgot about adhesion to other surfaces.

Goes for me too.

I failed to even think about what the water would do to the hands, and also predicted that the towel would unwring itself once it was let go (well, that could've been the fault of the towel, which is more of a piece of cloth). Otherwise, I got this one correct.

Same. I also wasn't expecting as much air to be wrung out of the cloth as there was, although that might be thanks to the moistening process he used; it would probably be hard to saturate the cloth completely that way without making a mess.

Getting his hands wet is simple enough and it's clear what I wasn't taking into account there, but I'm still trying to figure out why the water wasn't resorbed as soon as he let up the tension on the towel. That doesn't seem like something gravity could account for. Maybe you're right and it's just a quirk of the material.

It looked to me that the water that remained near the cloth was reabsorbed.

Ah, you're right. Looks like some of the water on his hands got picked up too. Guess I wasn't paying close enough attention.

Interestingly, the most influential prior, that made me change my prediction from the incorrect to correct answer was the existence of this post. Had I stumbled across the video without expectations, I would have put little effort to fix my incorrect intuition of "flies off in every direction." But reading the post, particularly the comment that I HAVE to watch the video, suggested that the actual result was interesting and perhaps in some way nonintuitive.

I also thought that the existence of this post implied some other result than the most intuitive one being the correct one, but I still failed to think of any alternative to the intuitive outcome (bubbles of water flying off in every direction).

Was that really the intuitive expectation for most people? When ringing it out in gravity, it tends to just run off the bottom. My prediction was wrong in the other direction: I expected the water to not really escape the towel very much and have only a small visible effect. (Though I was imagining a much smaller amount of water on the towel.)

I got it right. The OP could have used the new polling feature, BTW.

Some possibilities: it jets out in one direction, little droplets radiate outwards from all over, there are a bunch of miniature streams going in all directions, there are sorts of sheets of water radiating outward that split into droplets, it does any of these things at a rapid or very slow rate, the water doesn't leave at all

Reasoning: when I wring out a towel it usually all leaves in one big thing, then drips from all over. Does it all leave through one "faucet" because that's where the pressure is or because it's the lowest point? I've never paid too close attention to it, but I think it typically leaves from the bottom, which would imply that gravity's doing the work of choosing that one exit point. With the dripping off it's unclear whether the drops would just cling to the towel in the absence of gravity. It may be that the whole thing depends on the extra oomph of gravity; if when I wring out my towel it all leaves from the lowest spot (although again, I'm not sure of that,) then presumably the pressure prior to its getting there isn't enough to make it exit. All of this leans towards less water exiting than otherwise, and maybe not from all the same place, and (of course) slower than before. So my guess would be something like little droplets slowly radiating out from all over but not as much water leaving as before, unless it gets wrung really tight.

Meta reasoning: you were much more likely to post this if it was cool and/or surprising. The most intuitively cool and/or surprising things (that are still plausible according to the above thinking) would be if it all left as a sort of sheet or bubble of if none (or at least very little) left, or maybe if the sheet or bubble arrived but only after a lot of wringing. On the second thought, towels are pretty irregular surfaces, so I wouldn't expect a smooth sheet to radiate out from it. So my official guess is that very little will leave from it until it has been wrung very tightly, and then a bunch will surprisingly burst out.

~WATCHING THE VIDEO~

Hey, that was cool! I guess I was sort of right - water mostly didn't exit the towel, but I didn't predict (correctly, in my head, or explicitly, in the post above) what it would look like as it was not exiting the towel, and I was more dramatically incorrect about it all bursting out. The features I hadn't thought about were the clumpiness of the water and the way water looks like when it's about to fall from a towel but is still clinging - visibly on the surface, not hiding within the folds. I also think that if I had correctly predicted what it would look like I wouldn't have predicted the burst from the meta-reasoning, because the towel being shrunk in within the water was already cool-looking. When I predicted that something cool would happen, I suspect I should have thought through my reasoning earlier and seen at what points there was an opportunity for something cool to occur - which is a point that might be more broadly applicable. To remember the surface tension stuff I should have tried to remember mechanical features of water in general, like why droplets form in the first place and so on, rather than jumping into directly imagining how wringing out a towel looks like and reasoning from there. Broader lesson there, obviously not a new one, but perhaps one I should be better about keeping in mind: appeals to first principles are relatively more important when dealing with novel situations.

Didn't get that it would flow onto his hands, but I got that it wouldn't fly away, and that it would be reabsorbed. I was very nervous when I saw the couple of drops flying off at 1:46!

This was a really good idea for a post. Thanks, gregv!

Edit: Also, note how his knuckles stay dry.

Here's another, for programmers. Do not click this link before thinking of an answer.

First, we ask how you’d find whether a number was in an array? Most people’s first answer is to iterate through the array. [...] Then we ask for another option. Again, most people will answer that you could sort and do a binary search. [...] Next comes the big question: What factors influence which approach, between a linear scan and a sort+binary search, you should pick for an unsorted array?

Note: the hypothesis submitted by the people who proposed the experiment was correct: http://www.asc-csa.gc.ca/eng/media/news_releases/2013/0416.asp

Hm, okay, good exercise.

Water bulges out from towel as it's being wrung, reflecting the fact that the spaces between towel threads are being compressed. When it stops being wrung, the water goes back into the towel (70% probable).

If wrung sufficiently hard, the water might spray radially outward normal to the longitudinal axis of the towel. But I think surface tension should help keep it near the towel so I rate this as 20% probable.

Remaining 10% to other possibilities.

The bad news is that I have already seen this video. The good news is that I did this exercise for it the first time I saw it. My reasoning was something like. Well the situation is similar to the video on what happens to tears when someone cries in space to get similar results. It would just look like jelly attached to the towel. What I didn't expect was for the water to aggregate towards the hands.

I thought the water would form a layer all around the towel, and that happened; I did think the layer would be thinner, though. And I didn't predict the water would flow onto his hands.