Will Manifold be more accurate than the Veritasium crowd-sourced estimate?

600Ṁ1222

Dec 31

56%

chance

ALL

This resolves YES if Manifold's estimate from this market is more accurate than the crowd-sourced estimate from Veritasium.

https://manifold.markets/Quroe/how-many-jelly-beans-are-in-this-ja-OUnuy9zORR

Currently, the value there is 1750, which seems like a good average over the last few days as well.

SO,

If Veritasium gives the "crowd-sourced" estimate AND the actual value of jellybeans AND Manifold's estimate of 1750 is closer, this resolves YES.
If Veritasium gives the "crowd-sourced" estimate AND the actual value of jellybeans AND Manifold's estimate of 1750 is NOT closer, this resolves NO.
If they are equidistant, rounded to the nearest integer, this resolves to 50%.
If Veritasium does NOT release the necessary information, then this resolves to N/A.

Update 2025-05-30 (PST) (AI summary of creator comment): The Manifold estimate to be used for determining resolution is 1750. This is a fixed value. It will not be updated if the estimate in the separate Manifold market on jelly bean counting (referred to in the market description) changes.

Update 2025-05-30 (PST) (AI summary of creator comment): If Veritasium presents multiple 'crowd-sourced' estimates:
- The creator will use the estimate that Veritasium most prominently highlights.
- If Veritasium presents several estimates on equal footing (without a single highlighted one), the creator will use the median of those estimates.

Manifold Business Future

Get

1,000

to start trading!

People are also trading

Is Polymarket more accurate than Manifold at p<0.05?

40% chance

Will Manifold's calibration be better at the end of 2027?

Will Manifold be more accurate a month after the pivot

13% chance

In the 2024 US election, will Manifold be within the top 25% accuracy among all large prediction markets?

33% chance

Will Manifold surpass Metaculus in DAUs by 2025?

50% chance

Is Manifold better-calibrated than financial markets on questions about future asset prices?

6% chance

Will Manifold outperform Metaculus in the 2024 ACX Contest?

18% chance

Which will be voted the most important prediction ever listed on Manifold in 2030?

Will Manifold's 2024 general election predictions outperform FiveThirtyEight's?

55% chance

If Manifold begins allowing real-money withdrawals, will its accuracy improve?

Sort by:

bought Ṁ20 NO

veritasium might p-hack to make their point more impressive

Just read the Substack email, and it helped me discover this. I enjoy that this exists.

The veritasium crowd-sourced estimate is being biased by a significant number of "69" responses, though they will probably clean up the data before they present it for the video

@spiderduckpig they might use a median or some sort of intelligent algorithm, or whatnot

https://manifold.markets/Quroe/how-many-jelly-beans-are-in-this-ja#lz2xc06fvzc

@bens

I think there is a nonzero chance that Veritasium presents more than one 'crowd sourced' estimate by analyzing the data in different ways. I think it's important this market considers that in the criteria.

@Eliza Also some clarity -- is this market using 1750 exactly, or 'whatever Manifold's estimate is on X date' (using that specific market), or something else? What if Manifold's estimate changes due to a leak, etc.

@Eliza This market is using 1750! I don't anticipate it will change dramatically for any organic reason, and it's just easier for the purposes of this market to pick a single number bc I don't want to do some sort of time-averaged bs.

@Eliza And yes, fair, it IS possible that they will present more than one estimate. If they present one that is clearly the one they are choosing to highlight, alongside several other methods, I will go with the one whose prominence they raise! If they select several on equal footing, I'll select the median of those estimates!