We love you so much that we could just eat you up - even if you aren't a spider 💖 You can join our web by becoming a Patreon or member on KZfaq! 🕷🕷🕷 Just visit www.patreon.com/MinuteEarth or click "JOIN". Thanks!

I love Spiders Georg, a silly joke that became a great way to teach real statistics

My most favorite fact about statistics is that; if you have 2 legs and are a human, you have an above average amount of legs!

"But she was already pouring him a brimming glass of spiders"

I presumed Spiders George is not meant to be compared with the entire population, but rather a population sample for the analysis. The researches rounded up a random group of 3500 or so people, most of whom ate 0 spiders per year, but unfortunately somehow found his way there and skewed the statistic which was then extrapolated for the general population.

its spiders georg cmon this has been solved already

Obligatory Spiders Georg refere- -...They're talking about that? It's actually what the whole video is about? ...Okay.

Even without Spiders Georg, there’s plenty of people who do eat spiders. Whether it be children doing dares, those candied spider lollipops, or exotic cuisine. Some people just like eating spiders, so a nonzero average or even a mean makes logical sense.

The average person has less than two legs.

"I'm making a power point!" "What of?" "EATING SPIDERS"

I came for Spiders Georg and was given a Scanlan cameo I was not expecting.

My favorite example of this is "most people have an above average number of arms/legs/eyes, etc.

Loved the video, but i have a slight disagreement: it's not that the mean "tends to be more useful in datasets where the data mostly falls around a particular peak". That dataset has a mean that is roughly the median, so the mean has the same meaning as the median. The median is almost always way more useful than the mean, and this is a particular example where they are about the same, so the mean piggy-backs on the median's utility. I would say instead that "the mean is only useful when the dataset is continuous, and not discrete". Example: mean speed * total time = distance traveled. As soon as discrete data points (like people, apples, etc) is present, the mean becomes less useful. The median will always be the most representative data point in a discrete data set, the mean will always have info about the outliers, and is unreliable without also looking at other parameters (like looking at mean and median and then you know about the outliers).

This video inspired me to find out something, could Spider Georg eat 24 Billion spiders a year. Assuming all he eats are tarantulas, each weighing about 85g and providing about 76.5 calories, and him eating a pretty ridiculous 7000 calories a day (Sumo wrestler level), he would eat just over 91 spiders a day. This equates to nearly 33400 spiders a year. So in reality, there would have to be an entire tribe of Spider Georgs numbering about 718000 which is a little bit over the population of Macau. TLDR; There would have to be a Macau full of Spider Georgs for this to be nutritionally feasible.

Scanlan! I knew of the median and its purpose. But nobody could explain me how to make one and make it easy. You did today.

Hello, HR, my colleague just bound and gagged me... twice! What do you mean you don't care... Wait! HR, what's your name? GEORG!!

Excellent video. And while it's true that humans accidentally eating spiders is exceedingly rare, there are a few cultures that eat spiders intentionally as food. Fried tarantulas are a regional dish in Skuon, Cambodia, for example. I suspect these regions that eat spiders are pushing the mean up more than "Spiders Georg"

As a 5’5” guy, I caught strays from this video 😭

Obviously this is false, you got it reversed: the average spider eats three people per year

Ngl just came across this channel, and I already love it 😭

Sorry my bad, the spiders just taste so good

Median and mode can be overly majoritarian and insensitive to outliers when they really do matter. The median and mode of gunshot wounds per capita in Reykjavik and Caracas are the same (zero), but the mean will tell you a very different story.

until the reveal i was gobsmacked i know he was well known, but i couldn't believe they were actually going to talk about Spider Georg

I’ve been subscribed since you’ve started this channel, and I have to say this is one of my favorite videos you’ve made

I'm so glad that news reporting on wages and house prices and such only talks about the median amounts

Spider Georg: that one nerdy kid in class who used JavaScript to break validation and entered 2.1B as age in school survey.

I have to admit that my first reaction to "Spiders Georg" is that it was a culinary speciality somewhere that was pulling up the average in a less absurd way.

Great job, Arcadi!! We've been waiting for you to be the main speaker of another vid!! Keep it up!!!

Another reason why you can't take the mean of traffic, is the values not being absolute. Hours go in a cycle, so you can get different means depending on where you separate them.

oh my god, so happy to see the evolution of quality

Who’s eating all the spiders? My cat

It’s been so long since minute earth posted, I’m so glad that they are still working new video’s for us! ❤

wasnt expecting this to actually be about spiders georg i love it

This intro and narrative style immediately got my sub. Love the content! 😂

A weighted mean is more useful than a median. For example, the actual utility of income to your life likely follows a linear trend at very low income levels before reaching some saturation point and switching to a logarithmic behavior. Similar to a greenhouse gas's concentration in the atmosphere. It would be much more relevant to people's lives to measure the mean, not of income, but rather, of log(income).

This was really really good! Loved the interactions

This video is so interesting!! Thanks for making it!

Cameron, do you need help? Wink twice for yes!

at 1:38 the text in the top right changes to add an erroneous "talked" into the sentence, but then a few seconds later, it gets removed again

What the hell... I've been subscribed to this channel from the beginning and I swear the channel hasn't been shown on my subscription feed for YEARS!! Suddenly I see it pop and I'm like "Holy shit... he's still making videos!!??"

This makes me think of something I heard long ago about how cereal is allowed so many bug parts per some cereal metric. Because there is no way to remove all the bugs from the wheat plants and they get crushed up when the wheat is harvested. So, to any Vegans get eat cereal, it's probably not vegan...

In the original joke spiders George eats 10000 spiders a day which sounds like a lot but would be statistically insignificant with an average of 3 spiders a year

Great explanation, in early grade school we learned about mean, median, and mode, but not the context in which they could representative of a data set

Can u make a video about animal and plant dispersal ? I’ve found a lot of odd range maps over the years and maybe this could clear things up. You already talked about monkeys in America but maybe make a video on general dispersal mechanism, or a compilation of a few interesting dispersals in history

Was that Ithkuil writing on the presentation stage at the beginning?

This was hilarious AND informative 😂 love you all

I saw that Bluey at the beginning, sneaky animator 😂

It was me… sowwy

😆 the spider obsession at the end 🙌

Never heard georg out loud before, and I expected differently. Very good video tho that everybody needed thanks!!!

I love how there’s a recurring joke of Cameron having eaten a spider

The mean driving hour is an unfortunate example, because our measurement of "time on the clock" sets an arbitrary REPEATING zero, and the result of averaging depends on where it is put. The same issue by the way arises with the phase of complex numbers (or, directly related, the phase of different signals). The comparison to the phase also brings a potential solution: Represent each data point as a phase corresponding to the time of the day (such that one 24h day makes a full 360° circle) times the number of cars at that time. Then average these complex number and interpret the time corresponding to the phase of that average as the average time. Geometrically, it corresponds to asking “what is the average direction a driver sees clock pointer pointing to” (assuming we had 24h clocks with pointers). The direction could be represented as a unit vector for each driver, summed up, and then renormalized to a unit circle. Sadly, it is also numerically unstable if the average happens to be close to zero, so again it works only if there is a strong bias in the data. But at least it removes the dependence on an arbitrary choice of the start of the day. I've come across this issue many times during my Job (usually related to Fourier transforms, given that the phase resulting from a Fourier transform depends on the arbitrary choice of "zero" time) and during my PhD (complex numbers in quantum mechanics). Bottom line: Any value that has some periodic characteristic makes it very hard to talk about an "average". One emerging problem from that is that visualizing “phase” can be tricky, if the complex number can become zero at any point. If it is only a single point, you can try to detect nearly-zero values and interpolate the phase between adjacent values, but that’s also just a bandair fix, that won’t work ALWAYS. __________________________________________________ Less related to the video, there is also often the question of "arithmetic" vs "geometric" mean. The first is the common amean(x) = 1/n ∑(j=1..n) xⱼ while the latter is gmean(x) = (∏(j=1..n) xⱼ)¹ᐟⁿ which also is the arithmetic mean of the logarithm of n, and is often more descriptive of quantities, that have no non-positive values. Especially given that we often apply logarithmic scales ("decibel") to such quantities.

I find it cute that Arcadi's minute earth Alter ego carries his "Bubblegum bitch" look.

ohh Scanlan 2:08 never change

2:40 Moreso for distributions on a circle like this hourly distribution, the usual definition of the mean doesn’t even work. It can be salvaged some by treating the circle as complex numbers with unit norm, so that the mean will lie inside closed unit disk, being nearer to the border if there’s more definiteness in the data, and being in the center at 0 if the data has no preference overall (though it can be periodic and the mean would end up still zero, so it still has its usual failings).

I have a question: if the mean is mostly useful using an average distribution, and to find an average distribution you already need to find the mode and/or median, what actual information does the mean provide?

Could we possibly find this man who’s eating a million spiders and make him stop embarrassing us, the aliens are watching.

me. i’m eating the spiders. i require more spiders for food.

i feel so validated by the way you pronounce spider georg. thank you.

The spider segue at the end was chef’s kiss ❤

for reference, the standard variation of the amount of spiders eaten by humans is about 268328 spiders per year

Don't forget in the mountains of Papua New Guinea they eat fried tarantula on a regular basis because protein is so rare. This is the same country that practiced funerary cannibalism and developed CJD as a result. Kuru. The Shakes. THAT ONE.

i do remember reading that fact as a child, and i am certain that it stated that humans eat in average 8 spiders throughout their whole lifespan

Before watching, I thought this would be a video about food defects, where insect parts could find their way into the end food product.

I thought this was gonna be about spiders rapidly disappearing from different ecosystems

hyper realistic minute person isn't real, they can't hurt me

I Love Your Channel & Videos

I love the scanlan shorthalt reference on the men’s size chart lol

🕷🕸🎉 SPIDERS GEORG MENTIONED 🎉🕸🕷

as soon as i saw this vid in my notifications i immediately thought about spiders georg

Check out the Datasaurus Dozen, a collection of twelve radically different-looking plots with the same summary statistics. This is why I prefer data visualization.

Much more emphasis should have been made on the mode. Even more when talking about salaries, which are way more important than imaginary spider-eaters. Neither the mean nor the median represent as accurately as the mode the situation of the *majority* of the people.

Whoever created the Spiders Georg meme should be ashamed of themself. How dare they not name such a character Renfield?!

Hilarious episode!

I legitimately worried whether spider eating might be a little bit true (1 spider a year? Tiny spiders?) Thank you for putting my mind at ease.

ooooh so this is where the georgs came from

In all honesty, seeing Spiders Georg(™) becoming a vector for knowledge at this capacity really makes me believe that the kids will be alright

If you want to go into a deep rabbit hole as to WHERE the spider swallowing myth comes from, I highly recommend Lemmino’s Eight Spiders video

Missed opportunity to have the green screen dancing spider 😭

There's a specific style of cheese from Germany called Milbenkase that's made by coating it in mites and people do in fact consume a bunch of them when eating the cheese. So it turns out Spiders Georg is actually just a bunch of Germans eating Spider Cheese.

SpiderGeorg being awake voids his inclusion ... which saves everyone 5 minutes. It was about sleep.

This is why data scientists use sigma clipping, especially for astronomical data.

Great video!

3 spiders ever year? These are rookie numbers; in my day, it was 7 spiders PER NIGHT

When yall made the eating spiders joke the years ago i nearly passed away

To be fair, knowing the FDA allowance for bug parts in food, I wouldn't be surprised if the average person consumes more than 3 spiders per year, just in little bits over time

statistics professor: "You shouldn't do this, but I'm going to teach you how to make these numbers say whatever you want"

Omg Arcadi, I love the character😭 Thanks for teaching math!

Though I must say, there are places on Earth where spider is an edible item, mostly in third world countries and there are people who just love to eat exotic food. It very likely doesn't skew the data set that much, but large tarantulas actually provides a good meal for poor people in Asia, Africa and South America.

Kudos to Arcadi for making over 200k a year btw - you've made it, man :)

Using the height example, the average height for a human is not what most of us think of as a "typical" height. It's probably less than five feet. The thing is, when we hear statistics like this, we automatically fill in the blanks for what we think the context should be. We assume, for instance, that "height" means "adult height," so we discount the billion or so children who are shorter. And even the scale for adult height has two modes: one for men and one for women. Thus, the average height is sort of a compromise and not actually the most typical.

That are things I learned at school and I am always shocked that not everybody knows how to look at data sets...

So you're telling me I've been eating all those spiders for nothing?

Arcadi is obviously the *real* spider-verse, not those show-offs in flashy costumes!

Brought to you by my high school's 'Statistics and Probability' teacher! :)

Chart: Everyone besides Cameron and Georg eats an annual average of 0 spiders Haachama and everyone else who followed suit: Are you sure about that?

Didn't think spiders Georg would be mentioned here, awesome!

"old Tumblr memes" y'all really came for me huh

I remember hearing that silly eating spiders rumour years before the internet, and i was quite young at the time, but i still had a critical enough mind to not believe any crap that was peddled at me. The big question is... .... HOW WOULD ANYBODY PROVE THAT? How do you prove how many spiders people eat each year, without having a camera recording them while they sleep every single day for a year, and having somebody monitor all those hours of footage for spider eating activity. The morale of the story is always question your sources.

What I've learned is that we should be eating more spiders.