Bet on whether Manifold's estimate of 1750 (currently and roughly over the last 24 hours of this market) will be closer than Veritasium's crowd-sourced estimate here:

https://manifold.markets/bens/will-manifold-be-more-accurate-than-ESUsQREOtp?play=true

@bens We also need a market for what Veritasium will reveal the [mean, median, whatever] will be from the survey!

We got a shoutout on the Manifold Markets Substack email! I am honored, but Veritasium gets credit for making the almighty standard this market operates on. If the Veritasium team is watching, you've given a community something silly to focus on in these trying times, and we appreciate it.

@bens You can modify/edit your comment so it uses the "Add Question" tool instead of a link. It would be super cool to see the odds presented here live with that edit -- it would invite people over to your market for potential arbitrage opportunities.

@Quroe or % like this:

/bens/will-manifold-be-more-accurate-than-ESUsQREOtp

What if there are zero jelly beans?

@nikki What even is a jelly bean?

@nikki If you want to wax philosophical, I'm game to debate.

@Quroe That which is a jelly bean must have in its existence a half of a jelly bean. This having been proved it is evident that the same must hold for the half of a jelly bean - namely, it too must have a half and so on until infinity. It is evident that a finite space cannot - by it's nature - hold an infinite object. Therefore, since this is impossible, the infinite cannot exist in the finite plane, and we must conclude that the jelly bean cannot exist.

@FergusArgyll Zeno's Paradox can be overcome by demonstrating that, while infinities do exist, we can count infinitely many segmented portions of jelly beans and still have finitely many jelly beans.

That, or we invoke our inner lawyer and define a jelly bean to be whatever descrete unit the Veritasium team counts in the jar when they do their tally.

I think people are forgetting that internal packing is different than packing at the edges of a jar.

For example, think about whether it would be more efficient to randomly pack jellybeans against a smooth wall or against a roughly textured interior. Again, this is not maximal packing, but random packing.

Just going off the outer wall packing that we can observe, using an ImageJ macro, one might estimate about 1800-1850 beans. But when you consider that ~2/3 of the beans are in the interior, and the packing there is about 80-90% as efficient, you get ~1641.

I averaged this value with the community's expected value, since I was assuming that they had at least some information I had not considered, such as buying their own jars or counting the beans manually in the video, or something, and got a value of 1695, which I submitted. :)

@bens you also have to remember that many jars have concavity on the bottom surface. That chops off about a dozen beans.

@bens also, if the jar was sealed under vacuum, you'd get some unexpected jellybean shrinkage, by about 3-5% per bean, which might matter if you've bought your own beans and jar.

😬 Uh oh.

@bens additionally, depending on how long the jar has been sitting for, it's possible that jellybean-consuming mites/parasites/bacteria have eaten away a void in the center of the jar, invisible to the camera. This might have resulted in a loss of a few dozen breans.

@bens you also have to consider relativistic effects

@bens I lost some trust in Veritasium after the Poynting Vector video. He might put one huge jelly bean in the middle for the views.

Ordered M2000 for 1750-1899 at 22% if anybody wants action.

I counted how many are visible on each face of the video jar. On average, about 205 beans were visible on each face.

On the jar from robm, there were around 170-175 beans visible on the one face we saw.

My methodology is not really any more complicated than that. I'm really into the >1250 and <1750 zone. Probably also <1600 but hard to say. I did enter a guess of 1453 in the online form.

If robm is lying about the count of beans in the jar I could be in big trouble.

@Eliza I'm starting to think it's a 2l jar, so 1.33x the volume of mine, or 1640.

@retr0id 's picture needs to scale my jar up about 10% to match imo, which also roughly matches that number, and still makes my jar look slightly smaller than Derek's.

@Spin had also independently linked a 1.5l jar. It looks really close, but a 2l isn't that much bigger in any linear measurement.

I must add that a face of 205 beans multiplied by a thickness of 10 beans spanning the side is 2050 beans.

@Quroe but it's not 10 beans wide, more like 7-8. And counting all visible beans on the side would estimate almost 2 layers.

@robm Perhaps there's some light refraction near the edges that we're not accounting for? If I take a sample of a few horizontal lines, I get 9 to 11 beans, which I average out to 10.

In Mark Rober we trust?

@Quroe I trust the method, but 9 is my highest count on any row

@robm 9*9*23 = 1863... Aaaaaalright maybe I can get some exposure to some lower brackets.

@robm Mark Rober's method is getting me 9*9*23=1863 so I'm still pretty confident on my guess

@Quroe oh wait jinx we sent that at the exact same time lmao

@Quroe btw @Quroe would you be willing to extend the remaining 3k on your line of credit

@Spin Frick yeah, interest arbitrage! Can you be very specific about the terms you want?

@Spin my first guess (the one I submitted) was 7*7*(7*2.2)=754

So I keep revising upward. Not a good sign.

@Quroe M3000 loan at 50% interest, same timeframe for repayment

@Spin want to get in on the insanely good 13% odds before they disappear

@Spin Are we confirming that your previous loan should actually be repaid by the one month term I thought we agreed on before? I had some confusion with the insurance policy I bought on that because of that.

@Quroe sorry yes sure we can stick to 1 month timeframe (where I can elect to pay you back earlier if I want to) so essentially I'm paying you back M7500 total on or before June 28

@Spin Final confirmation. If I extend 3k more to you, the previous loan is also due at its original date, and this new one is due one month from now. 3k to you now, 4.5k back to me in a month for this loan. Please confirm?

@Quroe perfect!

@Spin

You're live!