Wisdom of the Crowds competition - the answer

[gjonesy9]

36:17

watch?v=60mw8kkXFSA

Ben S. Bernanke is looking for feed back. :)

[carclubleader]

3:26 i just love how you call us a bunch of nerds and yet you are the mathmatician

[saminulh]

benford's law

[Tranquilance]

@saminulh LOL

[ThePraash]

What average did you use for getting the LOG average? :D

[winonakid1990]

i love the R output for the graphs!! although i prefer SAS for my stats analysis

[MarcusMacgregor2]

It actually worked worse than mentioned.

The mode average of 1337 should of been rejected as a joke answer.

(It's gamer geek for elite, they speak math nerd I guess)

Since this is lower than the 1680 geometric mean it would raise the mean by its removal.

The internet may not provide this "wisdom of the crowd" you speak of; you need a real crowd not a virtual one. (and six-pence)

[singingbanana]

@MarcusMacgregor2 You're over estimating the effect of 1337. There were 7 such responses out of 1000. And it wasn't an outrageous guess as it's very near the right answer.

[carclubleader]

1337... wait thats leet in well, the language of leet

[Haguatchi]

What jellybean jar has 1617 beans when you start off?

[BigCityScripts]

Another great video. Subbed. First found the channel while searching for "flexagons" after discovering them in one of Martin Gardner's books.

[pawit]

Good to know that real mathematicians use R :D

[singingbanana]

@pawit Hahaha. True.

[GuyMcpersonLol]

@pawit R?

[Materialgirl1021]

the answer is 0 after you eat it

[johnnyappleseed12345]

Hey, btw I love your video's because no one else on YouTube provides so many mathematically based videos on any type of regular basis... Anyways, I was wondering, if taking the log of all values of a distribution can help "weed out" proportional mistakes, what would taking two logarithms do? Or three, etc... Does this progression lead to any type of higher-order analysis in terms of these mistakes? Thanks!!!

[PattyObrien616]

welp..... im gonna go play black ops

[HolbytaTook]

@housemoney1 no matter what, people are going to be able to do that. The cow example had men who had butchered a cow of proportionate size probably wrote down what the weight of their cow was.

But even with the jelly bean jar guessing contests that can still happen and that will effect the wisdom of the crowd.

[osrevad]

That video was put together so well that I actually had the thought that you should take the logarithmic average before you even mentioned it. And I had NEVER EVEN HEARD of the logarithmic average!

I don't know what you guys did to guide my mind toward that inspiration, but it worked!

[eaturfeet653]

y cant you use the root mean square?

[anticorncob6]

@eaturfeet653 What do you mean?

[00DonutDude]

Lol awkward silence at the end XD

[housemoney1]

Conducting this over the web and giving people a lot of time means they can find jars that look like that one, fill them with jelly beans, and count. Would the median, mode, geometric mean be as good without those types of participants?

[wilfred8686]

"Thats amazing!"

[dahattrickboy]

I can't believe i watched a video 7:22 seconds about jelly beans. lol

[masterlink101]

This was the most interesting thing I'v seen all week. I'm loving your videos, keep it up. I thank you and all my math teachers who you remind me of for making math fun and interesting!

[jlmknight]

noooooooooooooooooooo i wanted the jelly beans!!!!!!!!!!!! :) congrats ddss6

[NotJames1]

The fact that you mentioned my googol comment, and enjoyed it, has made my day

:-)

[dirtygurty4]

i would have won if he didn't eat that 1 jelly bean...

[theindfla]

you guys are a great team!

[Artonox]

This video is one of the best yet. Do more!!!!!

"That was amazing!"

[11000010101]

of course 1337 was the most common answer ^^ i said it because it wasn´t too unlikely, and if I hadn't guessed on 1337 and then that was the correct answer, i would probably have had to kill myself ;)

[keggerous]

i was way off and therefore am a moron

[jayjayspider]

Thumbs up if you felt like you were in math class Just now. My school ended this week's Wednesday. Sure did to me. And i just love those parts you try to make it funny :D

[IcEye89]

How is double the answer and half the answer proportional to the average?

If they are just as likely to be X beneath the answer or X above the answer, that would be proportional, wouldn't it?

[BabySelvey]

haha, the mode is 1337, leet.

[ArjayBlogs]

I dnt believe that after you said "its 1 less than it was" you managed to eat 1 jelly bean then close the jar.

[hollumber]

I think instead of 10,000....u should have probably used the "over 9,000" rule

[Blackcat008]

I want a recount

[Jonsoncao]

lol most common ans is 1337

[GetMeThere1]

Nice statistics lesson--and thanks for giving a practical example of the meaning and utility of the geometric mean!

But...clearly, you counted wrong. My estimation was in the 1800's, IIRC.

[cardsfan7012]

Did you remember to count the one that you ate!?

[raginggingy17]

my 10,000 was a joke... sorry if that messed up the data

[singingbanana]

@raginggingy17 Not at all. I knew it was.

[TheMangos27]

1337- Bunch of nerds.. i was dying!

[anticorncob6]

This is wrong. Taking the average of the logs might get you the geometric mean, but the DEFINITION is to multiply them and take the nth root of the answer, where n is how many people were in. But I think you wanted to make it simple, right?

I guessed 300 and actually though there were 300 in there. Was mine considered an outlier?

[guitardudeguy00]

So much for estimating the height and the area of the base and googling the density of jellybeans lol. What's the volume of the jar by the way?

[jampk24]

Hahah "bunch of nerds"

[muffemod]

"MY FIGHT MONEY!"

[poooop83]

I have no idea how I first came across you channel but I'm quite happy I did.

[FutureInventions]

THAT WAS SO INTERESTING! I love your videos!

[JPenfield]

"1337 - Bunch of nerds." I had a pretty good laugh at that comment. Pretty leet, man.

[okmasko]

Jim, please make this clear to me. I was though in school, that geometric mean is calculated as sqrt(number1*number2) or with more that two numbers (let's say 3) as thirdroot(number1*number2*numb­er3). Is it the same or (as usual) mathematicians use one term for two things?

[singingbanana]

@okmasko Those two things are the same. There are two ways to define the geometric mean, and they give the same answer. In the video we said it was 10 to the power of the average log value. Alternatively you can say it's the nth root of the product of the data (compare this with the arithmetic mean which is the sum of the data divided by n). I prefer the first definition, as it doesn't involve nth roots!

[lizardbart1]

was that the amount before you ate one or after you ate one?

[singingbanana]

@lizardbart1 After :)

[ElMirc]

haha I can't believe it turned out to be such a magical number BY CHANCE, just because you decided to eat one :D It's like you intentionally ate it.

[concretecubes]

lol at the number being 1337 XD

[shirankao69]

I think I'll cheat next time by calculating the mean myself and giving that as an answer :P

[WhatforNameIsThat]

The thing is with having so much ways of calculating the mean is to prove a theory you should have made up your mind about the average in advance to really show more reliable proof for the theory. Now it just looks like you chose the one that is closest. I understand the reason for getting a geometric mean but that is just in hindsight.

[singingbanana]

@WhatforNameIsThat We did make the prediction in advance. Did you not get that? David didn't even know the answer until the end of the video. That was his genuine reaction.

[okmasko]

@WhatforNameIsThat take a look - genuine reaction at 6:41 :D

[rockstarduh5]

I got closer than the geometric mean! Woot!

[TyYann]

Great. That's amazing!

[Zeldakitteh]

I guess 4273 was a bit of a high guess.

[Freshman000000]

I guessed 13. :(

[Error081688]

I don't understand why the geometric mean is more useful when there are proportional mistakes. I'll do some googling, but a video on that might be nice too :)

[HappyMemoryXD]

Finally a proper math video! Awesome!

[Error081688]

Haha thanks for keeping in David's comment at the end ^^. So then, it is fair to say that there is a correlation between WotC and individual expertise, right? I wonder if you could get a much closer answer asking something the general public is more familiar with. Maybe get an strangely shaped container and ask how much water is in it measured to the nearest mL would yield a strong WotC.

Zomg now I want to test lots of different things to develop a WotC factor.

[singingbanana]

@Error081688 Yes, we thought "so the wisdom part is important" would sum it up. It is the expertise of the crowd, the better the expertise the better the result.

[TheRacemaster23]

wow i guessed 2060 and commented 2059 cause he ate one... let me finnish the vid ;D hehe

[Randja066]

5:57

he looks way too happy o_O

[ipadkeyboard13]

@ddss6 Congrats!!

[oliethefolie]

Congrats to the winner! Same I didn't win. Jelly beans are yummy!

[Tolstoievsky]

so the masses have geometric wisdom lol

[FrazNinja]

this was definitely one of the best videos xD

[guppiemagic101]

hahahaha. i guessed 1212!!!

[grande1899]

You should have excluded 1337 from the calculations lol.

[Snakke40]

He has a special feeling for humor 16-16 and then a guy who wins has a 6 in his name!

[crayonstik]

1st year university statistics finally comes handy (:

[somebodyruleE]

so what should be the error from doing this like how much would you be wrong

[mrsean22795]

I love how 1337 was the mode. We truly are a bunch of of nerds!

[ddss6]

YEAHHHHHHHH!!!!!!

[happylee02]

@ddss6 AHAH gratz!

[FHomeBrew]

@ddss6 Congratulations! Don't get sick form eating 'em all at once ;-)

[Crashkiller62]

@ddss6 xD

[RodDuncan]

Did you check to see if people's answers were getting closer to the mean as time went on? In other words, were people being influenced by the answers that had gone before? If this were happening, one might expect the standard deviation of the first 500 answers to be less than the standard deviation of the final 500 answers.

[raxcax]

"Hello everyone!" :D

[sathiyanadaraja]

@singingbanana put the mean/meadian/log in a side. i am confused with the numbers you got it from viewers.so the mode is 1337, since the mode frequency is high. that's why your bell curves extrema is located in bin width of 1500-2500. lets say if u r going to do by yourself only option left is RANDOM number in calculator or in excel. doesn't make sense, after 100 random numbers mean hit a horizontal line and there will be a confident interval. what the point of doing this? ENLIGHTEN ME

[nelsyeung]

I hated stats before, but now it seems to fun. :)

[SirWilliamCo]

haha, I learned all these concepts in stats last year, but that's a sick way of applying it in this type of scenario.

[MOSpr0ductz]

I really like David, more of him please :)

[singingbanana]

@MOSpr0ductz I agree.

[EighteenCharacters]

I SAID 1617!!!!! AND HE ATE A JELLYBEAN AT THE END!!!! *cries*

[Wallencis1]

@EighteenCharacters HAHA XD

[GamemakingDude]

I love it when you teach us about maths its more fun than university or school :D

[Drummermean2]

This is great. Even though I'm not the statistician type, a big hand to you.

:)

[animewisher]

i failed so hard...i really hate when i guess so off... i mean 8K+ what was i thinking! D: oh well, congrats to the winner!

[HappyApple330]

Histograms date me back to year 8... memories =]

[bBoyMikE13]

I had guesed about 960......lol!!!

[mihassan]

That is amazing!!!

[aandgpointlessvideos]

People like this math more than what you have to learn in school, i read a book about it's called 'Professor Stewart's Cabinet of Mathematical Curiosities - by Ian Stewart': a great book, all about the things you teach us. Keep on roll'n vids!

[eXisentZ]

so cool,this is why you are so cool!

[marcusFOREVER]

Why are you not my math teacher?

[AFVC1]

I thought the old man was Sean Connery.

[D3m190d]

He got a jar of beaheans, he got a jar of beaheans! Guess how much's inside of it!

[rororigo52]

i think theres is 1616 jelly beans...did i get it :D

oh wait...the contest is over:(

[VinylCyril]

1337... buncha nerds :D

[Helldiv3rr]

Brilliant video :)

[ghudner]

You guys are having way too much fun with math! ;)

Great to see, keep it up!

[TpendragonT]

Oh, Jelly Bean jar.. Must you follow Benford's law?

[DestroManiak]

Bill Murray???

[NEDinACTION]

this, was awesome. dam, im gonna own my math teachers face off some day.

[bugshirt4]

he called us nerds :3

[illustriouschin]

This is really clever, is this the precursor to Isaac Asimov's psychohistory idea?

[MaThWa92]

Aw I got 1723, I'm almost average :(

[MissDante]

I knew I wasn't going to win. Congrats to who did. 1616 huh? Awesome it was such a simple number, sixteen sixteen

Very interesting way of getting the mean. :]

[Srxjer]

Doh! I got 1734, 5 away from the median average. :(

[sinkamaru]

i swear i didnt put 1337... o wait yes i did

[36trooper]

1337 = leet

lol

[Asperomo]

Aww, now I need to go rob a jellybean store :(

[Cagebreaker21]

singingbanana you've helped me with some math i needed figuring out in this video. thank you very much

[oEQjet]

This was one of your best videos.

[singingbanana]

@oEQjet Thanks! Especially since there is some proper maths going on here :)

[BaDSPLeR]

Since im in quick! Id just like to say that i love 'your' math, im enjoy math alot more when you teach/tell about it.

[singingbanana]

@BaDSPLeR Aww. Thanks.

[Cagebreaker21]

@singingbanana hello, just in determining the median/geometric mean of the data. Is it possible when it's in intervals. e.g. 1000 people guessing 0-500, 500 people guessing 500-1000 etc? I can't list the data in numerical order!

[illustriouschin]

What the heck, how did I miss seeing this contest!

[TruckFilms]

SO COOL :D

[GeneralBasco]

First View....

[gohyinxian12]

second =)



[Snc6475]

second!

[MarshellMcgehee]

first