[request] What is the accuracy of this?
https://redd.it/1svzeqj
@TheyDidTheMath

[Request] How much cubic feet of land would fill in this area? Also what would it cost?
https://redd.it/1svxnz8
@TheyDidTheMath

[request] This feels untrue
https://redd.it/1sw0h82
@TheyDidTheMath

[Request] anybody able to validate this? What is the actual amount of energy a query from chatgpt costs vs Google from 2008?
https://redd.it/1sw2f17
@TheyDidTheMath

[Request] Are these numbers accurate?
https://redd.it/1sw6wpm
@TheyDidTheMath

[REQUEST] Is the maths correct here? Surely he hasn't spent over $100m on Golf
https://redd.it/1sw8886
@TheyDidTheMath

[Request] How many pizzas have Domino's made?
https://redd.it/1sw2mo2
@TheyDidTheMath

[Request] this is a frog during a rocket launch, based off the force of the launch, frog terminal velocity,and frog to ground impact force, did our froggy friend (mathematically) survive.
https://redd.it/1svxgsr
@TheyDidTheMath

[Request] how low was the velocity of the bullet/how hard is the snow to stop the bullet?

https://redd.it/1swf5ob
@TheyDidTheMath

[Request] How much stink is required to do that?
https://redd.it/1swhbva
@TheyDidTheMath

Self Off-site Veritasium's simulation of luck pissed me off. So I ran it myself

This is in reference to this video: being delusion is a super power. Its a good video overall, but there is a simulation inside it that is completely wrong.

It was a simulation about luck. He took 18000 thousand candidates, gave them random skill scores from 0-100. Random luck scores from 0-100. Added them in the 95:5 ratio and tried to reverse engineer how important a 5 percent luck factor is. Basically: skill is 0-95, luck from 0-5, total score 0-100. Some of his results:

The top 10 candidates had an average luck of 95%
9 out of the top 10 candidates would not have been selected had luck not been present.

It just didn't make ANY SENSE to me.

Here is a POV of his selection: There are no skill differences in his model of the world!!!

|Rank|Skill|luck|
|:-|:-|:-|
||||
|1|99.9|???|
|2|99.9|???|
|3|99.9|???|
|4|99.9|???|
|5|99.9|???|
|...|...|...|
|20|99.9|???|
|...|...|...|
|40|99.8|???|
|...|...|...|

And I figure out why:

1. he considered skill to be a uniform distribution
2. if you are selecting the top 0.2% (in this case, almost equal to the total score of of 99.8), but your max skill score contribution is 95, then OBVIOUSLY you need a luck of at least 4.8 / 5??? This is clearly a flaw with the simulation, it doesn't model anything at all!

So I fixed it: I modelled skill to be a sort of log normal distribution (almost a bell curve). I calibrated this curve based on deadlift data. (it was the best quantifiable metric of skill I landed on). And I added a performance variance due to luck.

I now simulated a deadlift competition among 18000 applicants. And took that stats of the winner and the top 10. Note that the simulation still isn't perfect

1. using deadlift weight as a proxy for "skill" still isn't perfect since real skill isn't linear. Someone could lift only 20kg more but in the grand scheme of things, that's a huge skill gap
2. I was VERY generous with luck. I assumed a 7.5% standard deviation in performance

I posted my findings on my own youtube channel (LINK TO VIDEO). But I know a lot of you guys don't have time for all that, so I am posting it here as well:

|Rank|Deadlift Skill (kg)|Win percentage|average winning luck|avg luck “needed” to beat #2|
|:-|:-|:-|:-|:-|
||||||
|1|510|58.6 %|65|35|
|2|490|31.4 %|83|63|
|3|456|5.1 %|95|92|
|4|444|2.1 %|99|99|
|5|438|0.9 %|99|99|
|6|438|0.9 %|99|99|
|7|428|0.3 %|100|100|
|8|427|0|N/A|N/A|
|9|427|0.2 %|100|100|
|10|423|0.2 %|100|100|
|10 to 18000|<423|0.3 %|100|100|

Its now much more realistic. 70 percent of the top 10 candidates deserve to be there, and the very top candidates are never left out. The above table tracks stats for winning the whole competition (being the #1), and I noticed that average luck isn't even a very good metric. A better metric is "needed" luck to beat the #2. Also, I only considered 1000 iterations, so there is a lot of noise in the data, but I decided to leave that in, since real life doesn't give you a million iterations

I hope you guys found this interesting!

https://redd.it/1swj254
@TheyDidTheMath

If someone made an indestructible baseball, how hard would a batter have to hit the ball in order for it to turn into a meteor in space? [request]
https://redd.it/1swocts
@TheyDidTheMath

[Request] What percent of what percent of what percent is he talking about?

https://redd.it/1swvjyz
@TheyDidTheMath

[Self] If you blended all 7.88 billion people on Earth into a fine goo (density of a human = 985 kg/m3, average human body mass = 62 kg), you would end up with a sphere of human goo just under 1 km
https://redd.it/1swxbh9
@TheyDidTheMath

[Request] Is this true..??
https://redd.it/1sx35k8
@TheyDidTheMath

[Request] How many times more powerful does a pair of human lungs need to be to sneeze a lungful of air fast enough to blow off at least half of Jupiter's atmosphere?
https://redd.it/1sx3tm3
@TheyDidTheMath

[Request] who paid more $ or € the 14 year old make a wish donater who traded his Disneyland tickets to feed 300 homeless or the Turkish couple who feed 4k refugees so how many people i could feed in turkey for the price of a Disneyland ticket in the us or paris
https://redd.it/1sx6rye
@TheyDidTheMath

[Request] How far over the goal line is the puck?
https://redd.it/1sx798a
@TheyDidTheMath

[Request] How wide could this alien emerging from the planet be, if that planet is Uranus?
https://redd.it/1sx481y
@TheyDidTheMath

[Request] How many tapes did this lady use?
https://redd.it/1sxbhjn
@TheyDidTheMath

[Request] how much gas would this have costed?
https://redd.it/1sxdsc3
@TheyDidTheMath