Why does Vegas have its own value of pi?

[@LinusBoman]

Thanks for asking me to investigate this font mystery, Matt! On this one, I'd have to say I had a ball!

[@buschtoens]

Your little detective bit was sooo entertaining to watch! Thank you for one of my favorite cross-over episodes. 🙏

[@jergarmar]

Hey hey hey. Just because you solved that mystery, does NOT mean that you can abuse us with circle puns. Even if you thought it was rad.

[@chrisX1722]

pun intended?

[@KristopherBel]

Haven't gotten to this part of the video yet just came to the comments to see of anyone was talking typefaces, awesome!

[@J.C...]

So no one tried contacting Populous to ask them if they would explain how they came up with a jacked up pi number?

[@johnchessant3012]

I can't believe I hadn't heard the fact that the surface area of a cut sphere is proportional to its height until today. That's amazing!

[@standupmaths]

Thank you! I can't believe it took me by surprise as well.

[@hughobyrne2588]

I have a beach ball which has 7 colored stripes, each one-seventh of the 'height' of the ball. They each cover an equal area. I forget where I learned that, but it's really neat.

[@thedayb4tomorrow]

It's right up there with the fact that any napkin ring of the same height has the same volume regardless of the size of sphere it was cut from 🙂

[@sankang9425]

Yep. instantly thought of that one VSauce video..

[@BorisJensen]

3blue1brown has a really nice visual explanation for why that is, called "But why is a sphere' s surface area four times its shadow?".

[@capfer77018]

It's still a better Pi approximation than Matt's been able to get

[@standupmaths]

Too soon.

[@jergarmar]

Oof, man, why you gotta call out my mathematician like that?

[@martinshoosterman]

Until now...!

[@W8D_]

@@jergarmar *mattmatician

[@jonathanwallace7662]

Classic. I do always look forwards to Matt's Pi Day shenanigans.

[@kugirea]

They changed the number on the website lol

[@snabbott]

Funny story - in high school, I memorized a lot of digits of what I thought was pi from the back of a book. The problem was that the digits were arranged in groups that were in a grid of rows and columns. When I got to the end of the first group, I went the wrong direction to the next one and everything past that point was wrong. I was very sad.

[@Orbis3]

Are you familiar with the Portland Zoo station fiasco? Exact same thing happened, but the mistake was chiseled into the wall!

[@MrJellyTurtle]

@@Orbis3 I used that zoo station mistake in a riddle to propose to my partner! A bit harder to fix when it's carved into stone. en.wikipedia.org/wiki/Washington_Park_station_(TriMet)#Underground

[@Nanbread-bw7nq]

@@MrJellyTurtle what’s the riddle?

[@treelym]

As a CAD user, yes, AutoCAD has some fonts that are unique to that software. It also allows you to smash the spacing to fit.

[@MrCheeze]

I was wondering about that - if there was a very slight adjustment to the gaps between numbers in order to make it perfectly even on the right edge.

[@joelluber]

This really looks like a font design for an old school pen plotter.

[@plackt]

Yeah, as soon as he mentioned it I could see it from all my old drawings.

[@GodmanchesterGoblin]

Yes, it could have been CAD, but plenty of publishing or graphic design software allows that kind of adjustment to text layout. Any InDesign user would be familiar with this, for example.

[@richardhall5489]

If you want to see another result of typography done with CAD rather than suitable kerning done by a designer check out the giant exterior signage for The Titanic Museum in Belfast. Or as a friend called it Tit A Nic.

[@Shadowkainine]

For a stand-up mathematician you do a suspicious amount of sitting 🤔

[@persooniemand8346]

Can't stand up without sitting down first!

[@adrianbik3366]

I've sure Matt can't stand this type of comments...

[@petergerdes1094]

@@persooniemand8346Lay, crawl, stand works.

[@uplink-on-yt]

It's Parker standing (part of a large set of things named after him)

[@VAXHeadroom]

He is, however, a very upstanding guy (from what I've heard!!)

[@helloarigato]

Apparently someone at either the venue or the web design studio was told about this as it has now been updated. Interestingly the font in the new image has changed - the diagonals on the 7 and 2 are no longer curved for instance.

[@jcudejko]

That's a pretty fast turnaround for web design on this scale... cool

[@redmatrix]

Also, they added the "2" back on the columns!

[@Muhahahahaz]

The new image has exactly 99 digits… Why couldn’t they make it 100? 😅 (Similarly, the bridge columns stop at 49)

[@dannywinfield324]

Someone needs to give Max Cooper a gig in that place. My word that would be amazing. His work is very mathematical also which would be fitting

[@MarxismLilyism]

@@Muhahahahazif it’s 25 per line then it would make sense since the period is a character too

[@skiwarz]

They've since updated the website, and it now accurately depicts a continuous string of pi digits.

[@TurtleMarcus]

They even added the 2 in the bridge.

[@evasilvertant]

Ahh I guess they not only fixed the value of Pi on the website, but they changed the font as well. Right now, the image they’re using to express Pi is set in their branding typeface, Graphie. I managed to replicate their image exactly using Graphie Regular, set at 25.8 pt, with Metric spacing and 0 tracking (i.e., they did not adjust the original spacing of the font).

[@ben-the-bird]

A German magazine editor once said: "Layouters are just an entirely different kind of species." Meaning that they have little regard for the actual content and only focus on the look of things. 😅 Back in the Atari and C64 days, people could just type code, that was printed in computer magazines, into their home computers so they could play some simple games. There were infamous instances where layouters clipped parts of the printed code, because it didn't look right on the page, breaking the code for all the readers who sat down and painstakingly typed the code into their computer console. 🤣

[@timduncan6750]

UGH - I remember typing out code from COMPUTE magazine and it not working. I'd go through it and sometimes could find the issue and correct but most of the time I had to wait for the corrections to come out in the next month or month after to get it to work. I always wondered how this could happen and I'll bet you're right - it was the person doing the layout. I don't have any of these anymore but would be curious if it was always the end of a line or something...

[@albertmagician8613]

Also famous were the listings of the Forth issue of Byte. Not only were the listings not in a fixed font, they played hard and loose of spaces that are absolutely essential, making the listing all but unusable.

[@ScottiStudios]

I remember being about 10 years old copying code from magazines into an old BBC computer to make games, and always wondered why they would publish code that 50% of the time didn't work...now I know! 😂

[@germansnowman]

When I used to work in a printshop as a graphic artist and designer, I had colleagues who would typeset text _without reading it_, copying mistakes etc. I am physically unable to do that, I always need to read and understand what is going on.

[@rogerbooks4386]

,,,

[@AtomicSource11]

7:52 There's another math mistake in a graphic on their Science page! For the Geodetic Math graphic, spread out in the triangles is the repeated equation δ = 180 - ( Ω - Ω ). That makes the omegas seem irrelevant, it's just δ = 180. However, under the heading there's a more useful looking equation, Ω = (180-δ)/2. Rearranged for delta, that becomes δ = 180 - 2Ω, OR… δ = 180 - ( Ω + Ω ) They changed a plus into a minus!

[@fresh7995]

Lmao well done for realising

[@joon0]

The pi in the bridge columns is wrong too. Skips 2 before the 3rd row.

[@davidgro2000]

@@joon0 That's already in the video

[@restorer19]

@@joon0 It's just cut off with the contour of the arches. On the bottoms of the arches, you can see the contour takes precedence over the digits.

[@joon0]

​@restorer19 it's clearly not just that, as is evidenced by the first two rows being correct. They could have easily included the 2 at the begining of the 3rd row and cut off a number from the end instead.

[@Ruxinator]

The layout artist or supervisor who said "No one is going to notice such a minute detail in a wall of numbers" is really shaking their head now.

[@tehlaser]

Hey, it got them some free publicity.

[@caelum9]

assuming enough people on the internet don't care about math (or really any subject) is usually a mistake

[@Chewychaca]

Matt showed them what for!

[@mgalyean]

I don't know. It adds an Easter egg mystique that leads one to the expensive premium suites for full verification. Perhaps it worked as intended?

[@unitoonist]

You have your text block set for justified left. You need to switch to forced full to get the curning to match. The designer would have force full turned on to fill the wall properly.

[@spectralpiano3881]

I think it'll be even harder than that because of the multiple panels. I can't imagine a single digit being on the edge of two panels, so you'd have to align all numbers in each row to match the edges of the panels, so nearly impossible to recreate

[@unitoonist]

I've done many building and wall wraps. I would block the entire wall as a text panel then subtract the void areas. I'd set the fill to full. Then text drop the text in and scale it until the fill was even. Then, I'd break the panels into individual output files, at the width of the print/cut path capacity. Usually around 52"-54" (132cm-137cm).

[@misterbonzoid5623]

@@unitoonist Considering your first sentence it's interesting that you misspell 'kerning'.

[@unitoonist]

@@misterbonzoid5623 🤣🤙🏼 right!? Thanks for looking out. 🤣🤣

[@YodaWhat]

Matt, you should have cut 0.6375 in half, for 0.31875 of the circumference, then stuck a pin through the tape measure at that point, stuck the pin into the melon at any point, then marked many spots on the melon at the tape end. Thus you would have created as many reference points as needed for making a single planar cut! :) @standupmaths

[@Muhahahahaz]

My thoughts exactly… What was that wobbly nonsense? 😅

[@TroyLaurin]

It was a Parker cut

[@RobDeFino]

Or find the value of the vertical side of the right triangle he initially makes with the ground and a portion of the radius (when finding the angles to figure the portion of the circumference to measure to) , then subtract that from the radius, and then cut at that height. Then get a ruler, put a pen at the correct level, and spin the melon so a mark is made around that circumference. I feel like he went with the most complicated method of finding the "ground" lol

[@YodaWhat]

Yes, his method was . . . odd. But marking a circle by spinning the melon would also be tricky, unless the melon was _perfectly impaled_ onto some kind of vertical axle.@@RobDeFino

[@RobDeFino]

@@YodaWhat hmm yeah good point, didn’t consider that the melon would try to walk away. I would say skewer it to something but we’re now at too much effort for a small bit in a video about a written number lol

[@kibiz0r]

Matt the mathematician: "Gonna do this very precisely. Okay, 24.75 degrees..." Matt the craftsman: "Okay and now we just measure once and cut twice... Done!"

[@irrichman]

Horrific imo. If you are just going to wing it, after that measurement, at least make a single clean cut so the thing doesn't wobble. 😂

[@Muhahahahaz]

@@irrichmanomg, the wobble bothered me so much… I was expecting him to use a string or something… Wrap it around the sphere to measure the circumference, then mark 79.1% of that. Wrap it around again to find a percentage of that circumference circle If you do this 2 or 3 times with circles that all meet at the same “pole” then you can come up with a bunch of marks all at the same “height” around the bottom of the sphere, where the cap should be cut off

[@mscbijles1256]

@@Muhahahahazthat’s way too thought out for Matt 😅 Also I guess it would’ve shown the melon isn’t a really perfect sphere.

[@mrTii]

That melon bothered me. I just couldn't get over it.

[@chaos.corner]

@@Muhahahahaz Even worse, I bet he has access to a compass/dividers that holds a marker.

[@achehex]

Wasn't expecting an intersection of maths and forensic typography. Absolute treat for my nerd interests

[@mattp1337]

Forensic typography is a phrase that doesn't get used enough.

[@vigilantcosmicpenguin8721]

What a plot twist.

[@DarcyCowan]

I think Linus also gives a clue to how the web designers got those numbers and messed up. He said the wall is in panels, I suspect each panel has its own file, whoever wrote the page just got the first file and figured this is the first part of pi. So, in person confirmer should not only check if the wall has those digits on the first 4 lines but also see if the sequence length matches the width of the panel they are on.

[@Manologft]

This makes the most sense to me, I could imagine something like that happening

[@Creamcups]

Or they had an image of numbers of pi and just cropped it, not caring about accuracy because what nerd would ever notice

[@riverbraithwaite7741]

Aha, that's an excellent hypothesis. The panels (or a surface applied to them) look like they're probably laser cut. The vector files used to give instructions to the laser cutter would be super easy to convert into the graphic used on the site.

[@ano_nym]

@@Creamcups why that's not the case is explained in the video...

[@rolandsieker2286]

a) The seem to have fixed it on the website b) i played around with size of the error a bit. Thinking thermal expansion, you would have to heat up the sphere by about 10⁻¹⁷ K to change the size by that much. The energy required for that would be about 2 µJ. About 1/50 of the kinetic energy of a flying housefly. Unless i dropped an order of magnitude here or there.

[@Kenionatus]

Reminds me of a joke: A mathematician, a physicist, and an engineer are asked "what is pi?" The mathematician says "Pi is the ratio of a circumference to its diameter." The physicist says "Pi is 3.1415" The engineer says "It's about 3, but I use 4 just to be sure."

[@somethingsomethingsomethingdar]

I got to the 20th minute and started laughing out loud (literally) when I realized how eclectic this video was. The shear obsession of figuring out why they skipped numbers of pi had me smiling ear to ear

[@egodreas]

For the people actually planning to visit Vegas and confirm that the digits in question are indeed on the top left corner of that wall, here is an additional task to make the trip more worth while: The previous largest spherical building was the Stockholm Globe Arena. That building is used to represent the Sun in the largest permanent scale model of the Solar system, aptly named the Sweden Solar System. Why not figure out the proper sizes and distances for our various planets, moons and other celestial bodies, and find suitable locations for them in and around Las Vegas? If you Americans want to beat the Swedish record for largest spherical building, why not go the extra mile and break this record as well? For reference, I believe that 1 AU corresponds to just over 10 miles at this scale. Or about 55,400 feet (16,900 meters). Why don't you pick a celestial body, calculate its size and distance, and put it in a comment below?

[@Kyle-nm1kh]

USA doesn't like spending money. All our architecture is designed to be cheap. Maybe few exceptions idk lol

[@thesharpestknife]

@@Kyle-nm1kh ​ I'd say the Vegas Sphere is already a very notable exception

[@sincityminion8532]

Using your scale Tule Springs Ranch at approx 15.2 miles from the sphere puts it roughly in Mars orbit.

[@Kyle-nm1kh]

@@thesharpestknife yes, but keep in mind it's not just a building. It was built to generate money through advertisements

[@egodreas]

@@Kyle-nm1kh Well this wouldn't really have to cost anything. In Sweden, most of the 20 or so celestial bodies were stylized in nature and sponsored by local museums, schools or science institutes. There's no reason a large sphere can't be made by a bunch of volunteers, even school children, in just an afternoon. I guess the only issue might be getting the permissions necessary to display them in a public place.

[@xakaryehlynn4749]

My favorite fact, amazingly illustrated, is that you only need that first line or so (i remember up to the 323) to be able to make a sphere the size of the visible universe and you'd barely notice the difference

[@mrkitty777]

39 digits

[@slimecubeboing]

Yeah. I’ve heard 38 or 39 digits of pi are enough to calculate the circumference of the observable universe down to the width of an atom

[@KnuckleHunkybuck]

You only need the first 39 digits of pi... and a shitload of papier-mâché.

[@ironcito1101]

Theoretically, unless you have Planck length precision, you could still need more digits.

[@ThatGeorg]

At that scale of sphere, given space time curvature, I'm pretty sure the first 38 digits of pi are not the number you would need.

[@scaredyfish]

I work at a science centre and sometimes things go awry between writer and designer - I remember I had once written a panel about the electromagnetic spectrum going from radio to infrared to ultraviolet, gamma rays etc, with the classic rainbow of visible light in the middle. Somewhere between my mockup and the final design the rainbow got flipped. It was fully my fault, as I had even signed off on the proof before it went to print, because I was checking the text, but didn't think to check the direction of the rainbow.

[@erinm9445]

Oooh, what a cool job! I want that job!

[@naedolor]

As an audio guy, I read frequency charts. You have no idea how annoying is for me to read the light spectrum chart based on wavelength. You know what? I'm just gonna say it. I'm happy your graphic got flipped. There.

[@JoeSmith-cy9wj]

I've seen many of these charts. I've heard that if the spectrum were a standard piano keyboard, with the "7" colors each being an adjacent key the keyboard would reach from the earth to the sun and back eight times. Which is a linear instead of logarithmic scale.

[@davidpalmer9780]

I wrote a section of IT policy on 'Time Synchronisation' and made a reference to "Coordinated Universal Time (UTC)" with UTC mentioned throughout the policy. It came back from the printers as (CUT) with their typesetter believing I'd made a mistake and they corrected it before going to press.

[@tzisorey]

Notice how Matt gives the knife to someone more responsible and less accident-prone than himself. That's the sign of an experienced Knifey-Spooney player that knows their limitations.

[@aplusservice]

I've just started watching, but it's hard for me to imagine them making an error more egregious than abbreviating metres as "mtrs" 😂

[@AvoytDesign]

their E key was broken

[@pexfmezccle]

Because most Americans don't recognise an "m"

[@George-f5t]

The observation that the surface area of a spherical cap is proportional to the height of the spherical cap goes back to Archimedes. He wrote a treatise (in the third century BC) that included the result (On the Sphere and Cylinder), which still exists. He was so proud of the result that he requested that a sphere and cylinder be but on his tombstone. Cicero (in the first century BC) wrote about visiting Archimedes' tomb when he was a quaestor in Syracuse and seeing the sphere and cylinder there.

[@mrIceblink]

This fact can also be used to easily solve "Tarski's plank problem" (see wikipedia) in the case of a circle (eg how many equal width planks needed to cover a circular hole) The was a nice article about it in American Mathematical Monthly 2008 that I read as a student called "Three problems in search of a measure" by Jonathan King

[@thomasmclean9406]

I love the community of this channel - someone points out something and an entire video is made in response. And the interaction between Matt and his viewers is just so nice, I can't put it into words. I hope this channel lasts for years and years!

[@BirnieMac1]

100% The aspect of learning from each other is a wholesome thing you don’t see often Especially when it’s always in a constructive manner; utmost respect to Matt cultivating a community like that He’s helped me understand so many concepts better, or learn new things about ones I thought I understood well 😅

[@0hellow797]

Yeah, in CNC plotting programs I’ve used, they substitute fonts with vector lines, and in most cases they don’t adhere to the font “rules,” especially when cutting with a zero thickness curve

[@logicalfundy]

Yeah, also a lot of the "rules" in fonts assume they're going to be turned into a raster image. Certain parts of the letters are forced to align with the pixels properly to ensure they don't look blurry, hinting is done to ensure the right pixels are turned on or off even at small sizes, etc. None of that makes sense to a CNC plotter, because it doesn't work with pixels. I imagine the same is true for the slicer I use for my 3D printer.

[@mynameisben123]

Aren’t fonts already vectors? I’d have assumed they would work nicely with CNC

[@logicalfundy]

@@mynameisben123 You're right - most modern fonts are vector. *HOWEVER* - they do a lot of tricks to make themselves line up with the pixels of screens nicely. This leads to things like line widths being different on a screen than on a CNC machine. Which in turn makes identifying the exact font more difficult, especially when many fonts are very similar with only minor variations.

[@JustSomePasserby]

6:00 As soon as you pointed out that the pi's cancel and I saw the ratio becoming linear, I said "Ooh, as a consequence of the hatbox theorem." Thanks for reminding me that I haven't fully forgotten Calculus I yet.

[@barefootalien]

Taking pedantry to a whole new level and in the process discovering some cool stuff. What a video!

[@georgelionon9050]

If the melon is 20 cm in diameter the pi-error is less that a proton.. so the invisible knife cut was massively larger. Heck even if the sphere was a 1000km in size it would be still a sub-proton sized error.

[@MisterIncog]

"not to scale" sends its regards

[@dumbledorelives93]

Clearly engineers were behind this. They only respect the accuracy of pi to a relevant amount of significant figures 😅

[@matthewellisor5835]

Negative. It must have been architects or graphic designers. They only respect the aesthetics and occasionally the function (for the "experience" or "feelings"or something) ;D

[@LelouchVee]

As far as I (myself an engineer, of course) know, the most significant figures of pi anyone uses is engineering is 16 with NASA, and somewhere in the ballpark of 30 in theoretical physics. Why use more digits when less is plenty? 😅

[@Youtronics]

Nah, clearly architects. An engineer would properly round it instead of using random digits.

[@blacksarlacc91]

I never memorized more than 3.1415926 as an engineer. Even this was never useful for my purposes often we estimated pi and e with 3 to get a rough estimate and then used a calculator with however many digits it had to calculate. Quite honestly at the point where numbers went into the equation we were mostly done anyways. No need to worry about exact numbers usually.

[@appa609]

​@@LelouchVeeIf the pi shows up in an exponent then you need the accuracy.

[@danielbickford3458]

Had a thought about why they cropped pi the way they did. It could be that whoever was looking at it thought that each panel was one unit which then rolled over to the next panel instead of scrolling across every panel wrapping back to the beginning.

[@bigmuz_pilot]

Yep this is it I reckon 😅

[@brianmckeever5280]

I worked with a team of 7 graphic designers in the past. They are visual. They are not numbers, they are glyphs for an image. Surprises me not at all. I live in Vegas less than a mile from the thing and it is very interesting!

[@TheCleric42]

As a guy who’s memorized pi to 60 places AND who has set metal type by hand, this video was made for me.

[@davidjowett8195]

And for todays lesson 1. Matt knows lots of people with different interests who are as excited about their subject as Matt is about maths 2. Melons make a good representation of spherical buildings 3. Matts viewers are probably more pedantic than Matt himself 4. Matt should NEVER be trusted with sharp implements Great video again. Thank you

[@WaluigiisthekingASmith]

6:45 this is actually because the area of a sphere is the same as the area of its "label." In other words a sphere has the same area as a cylinder with radius r and height 2r. If you imagine projecting a sphere onto the surface of a cylinder, there is a loss of area due to the fact vertical distances become shorter but there is a gain of are because horizontal distances become larger and these two effects exactly cancel. This is actually the basis of the cylindrical equal area projection!

[@trueriver1950]

Fun fact The cylindrical equal area projection distorted the shape of all countries in proportion to their distance from the equator. In an attempt to remedy this the Peters equal area projection stretches the chronological projection in the North South direction to share our the distortion more fairly: Equatorial countries look thinner than they should, and polar regions still look fatter, only not so much. That's the theory: but in so doing Peters managed to get the US and most central European countries looking about the right shape. Funny how an attempt to be "fair" ends up giving US/EU the advantage. Never seen that happen before ...

[@TinBane]

As a developer who works closely with designers, they would have been like “we took a photo of it, its literally on the wall, what more do you want from us?”. I was as UCLA and there’s a building with E=MC^2, but the 2 is weird. Turns out, the architect or designer of the exterior decoration took the 2 off because it looked ugly and it was drawn on afterwards.

[@SomuaSomua]

You should see if they can let you get some squid’s on there

[@Term-0]

E²=p²c²+m²c² even the e=mc² was just a change to look better and more recognizable already

[@landfillbaby]

​​@@Term-0you mean E²=p²c²+m₀²c⁴ and that has a different use case

[@stefanschneider3681]

You beautifully answered a question that I NEVER IN MY LIFE thought I would ever have 🤣! Just so much fun how you put everything in perspective at the end, admitting how silly this all was, silly in the best way possible 👍!

[@macmcgauley3757]

CAD would also make sense because when you go to copy paste, it will only select the digits in the selection square you draw instead of selecting the whole line like you would expect in something like word

[@unnamed_channel]

Word (and other text editors) can do this too. In Word, hold the alt key when selecting.

[@JoseNovaUltra]

@@unnamed_channel what an unexpected and cool feature, thank u stranger!

[@BenJackson-on8qw]

3:42 Only in a Matt Parker video would someone get out a melon, chopping board and measuring tape simultaneously.

[@ToppyTree]

That's what they call a paper pi. Similar to a paper town for a cartographer, engineers slightly vary the value pi in their calculations as a copyright trap

[@ToppyTree]

Ah who am i kidding? we all know engineers just use pi = 3

[@Kyle-nm1kh]

I like the 3 and 1/7 route if you're going to over simplify it. It's only about .001265 off

[@BillRicker]

Or 2 π =~ 6400 mils (approximate milli Radians), which is great for small angle trig approximation

[@janmelantu7490]

@@BillRickerthat’s literally Indiana pi (3.2)

[@tonymouannes]

No one is worrying about copyright over here. Also, giving wrong information isn't acceptable. Not to mention that the value of pi isn't copyrightable anyway. It is totally legal to copy pasted on a computer (from a text, or using ocr; copying a picture or taking a screenshot to republish the visual can be copyright infringement).

[@MegaRaydiation]

They've fixed it! I went to try replicate it with the font "Graphie Book" which seems close, when I noticed they've got the right digits there now.

[@lucbloom]

The engineers that built the website and the ones that built the building will like this video. Going into detail about someone’s own work is always a treat.

[@chrispi314]

6:50 This reminded me of the Napkin ring problem and for the designer part, I can confirm that sometimes you are so much into "how to make it look good" you easily forget if you should.

[@halfsourlizard9319]

I feel like this IS a degenerate case of the napkin-ring problem ... with a zero-width hole?

[@Rulerofwax24]

You uploaded this only a few days late! This past weekend was TwitchCon Vegas, meaning a lot of internet nerds were swarming the area and I'm sure some would have gone to the 0.791 sphere for you on this mission.

[@madlep]

“Until someone goes to Vegas to confirm, it’s not 100%” - the problem there is the way pi is defined in Vegas, stays in Vegas. So we’ll still never know 😆

[@UconnPhil]

Is this like Schrödinger's truth - the truth cannot be acertained until examined?

[@KyleJMitchell]

"I went to Las Vegas, and I found the answer! Unfortunately, I then left Las Vegas and I no longer remember the answer. The place must be a numerical black hole, not just a monetary one."

[@creativecarveciteclimb5684]

@@UconnPhilSo the sphere's version of pi is both correct and wrong at the same time?

[@williammurphy8227]

Some years ago I toured the Fermi National Accelerator Laboratory (Fermilab) and the docent was describing the auditorium that the physicists designed. The outside has ribbed columns that are based on some hyperbolic function and the drawings had the dimensions with sub millimeter precision. The construction company apparently laughed and suggested that if they made it to within an 1/8th inch, that they would be happy with the result. (Of course the accelerator beam line is built with much finer precision. It is machined metal after all.) As I looked at this value of pi, I thought, “that is well beyond the precision that the building would be constructed.

[@taimunozhan]

Given that π is normal in natural base, any sequence of digits will eventually appear. This means that you could start with a row with the correct digits of π and then add whatever random digitis in the following lines and there will be a (likely gigantic) number of skips that makes it work as in that hall.

[@angelmendez-rivera351]

It has never been proven π is normal in any base, so your statement is merely a conjecture, not a fact.

[@leeprice133]

@@angelmendez-rivera351 Yep. While it can be proven non-constructively that*almost all* real numbers are normal, only a small handful of numbers are known to be, and some of those have been specifically constructed to be normal.

[@mglenadel]

About the digits of PI, as a designer, I can almost guarantee you that the full number* was once pasted in regular lines, but it was deemed to be too wide and "skinny" (too few lines) and, instead of reflowing the text, the designer just lopped it off to have the desired block of text. And, about going full-forensic on this, there is a quirk: pretty much every piece of design software implements character spacing slightly differently.

[@satyris410]

if you mapped the digits over a curve, like the digits in the archway 9:09 would that offset it enough so when it was straightened it would be out of alignment?

[@peterdefrankrijker]

He totally ignored the spacing of the decimal point. He should have tried a non-proportional font.

[@darktemp_de]

I found a way to replicate the number wrapping :) At 11:38 I paused and wanted to try it with all fonts. And then the 5th font in my list matched^^ Steps to replicate: Take "3." + 1382 digits after the point. (Or 1059 digits, if you don't want to fill the 4th row) Open mspaint (windows 10), make the canvas bigger than 2799x112 px. Add a textbox, paste the string of the first step. Select all of the text, change the font size to 12, the font to "Bahnschrift" (not sure if that's translated because my windows is set to German). Resize the textbox to a width of 2799px. (2798 also works, but shifts the last 1 to a 5th line) Important for the resize: Keep the zoom to 100% otherwise the status bar shows a different width! Also, I noticed that you can use ctrl+mousewheed inside the textbox but that is ignored after completing the textplacement.

[@R3dFlames]

Bahnschrift is indeed Bahnschrift. I can confirm it as a German who generally sets everything to English by default. But also it would be incredibly awkward and impractical if font names were translated in any way. Otherwise what would you have, fonts like... Schweizerisch Neue (as mentioned in the video of course :D) Komiker Ohne Kurier Ohne Zeiten Neu Römisch Einwirkung ...ja ja, I'm just being hyperbolic of course Nice work finding a way that replicates the wrapping btw

[@mentok9396]

I think the ending explanation of the true scale of the difference is the best part…

[@michaelwisniewski6047]

That just proves the interior designers of the Sphere worked in Paint 😂 Perhaps the architects did too...

[@ddognine]

@@michaelwisniewski6047 It's a humorous thought, for sure, but, they probably used some proprietary, bespoke, architectural software that costs more than most people make in a year.

[@FirstLast-gw5mg]

Finally, my knowledge of Pi to more than 20 decimal places has actually come in useful! ...but only to see where they went wrong in printing the value of Pi...

[@GodmanchesterGoblin]

You'll be pleased to know that you're not the only one! 🙂

[@Coloneljesus]

The most puzzling thing to me is how this happened, considering that just writing Pi in the website or design program is SO MUCH EASIER than screenshotting an cropping a design file.

[@andersjjensen]

You have no idea how little graphical artists pay attention to the things they're working with. They just make it look pretty. And things that don't want to conform to their idea of pretty WILL be made to conform. They have to tools to make that happen, and they're NOT afraid to use them!

[@Pyradox029]

Matt. Love this. And it is fascinating that any slice of the sphere is proportional to the area. But I think you are only looking at the exterior of the building for a spherical building. The interior can be a full sphere and looks way more spherical, but buried in the ground. So if take your melon and bury it in sand is it really no longer a sphere? Or is it a sphere transitioning between mediums??

[@Beesman88]

Great detective work. But the value is intentionally smaller, they will clearly open a betting game inside the sphere. Giving you odds of "Vegas PI"/4 if you land a random point outside a circle in a square. Very slowly winning in the long run - classic Vegas looong con.

[@_Matchu]

LMAAAOO

[@DavidSavinainen]

So if you bet the entire collective wealth of everyone on Earth, you'd be making a net gain of one micro-cent. What a win! (Collective wealth is of order 10^14 dollars, so with a difference of 10^-22, that would be a win of 10^-8 dollars, or 10^-6 cents)

[@Kyle-nm1kh]

​@DavidSavinainen except they charge a 1% fee each bet

[@EebstertheGreat]

@@DavidSavinainen Total spending is also on the order of $10^14 per year. (A given dollar gets spent many times each year, but most wealth is never spent.) So if every dollar in the whole world that would be spent on something else instead gets spent betting on hitting the circle, and they keep this up for a million years, then Vegas can expect to earn one cent. Seems good.

[@ed_halley]

So many fonts have different widths for various digits and the period. In the old lead type days, most typefaces would use an en-space width for all digits and the period and comma, just so ledger sheets would print properly columnar. Once you see a video clock shifting left and right for 11:11 vs 10:59, now you will see it everywhere.

[@K-o-R]

What's really annoying is that even for fonts with tabular figures, *for some reason* when it displays multiple 1s together, they are ever-so-slightly narrower. No other digit combination does that.

[@KyleDavidE]

Lots of fonts have a specific variant/style with monospace digits, but devs/designers often forget to enable it.

[@Term-0]

I have noticed that far a long time. in my opinion any number that changes should be in a monospaced font to avoid just this, as I find it rather irritating.

[@AlcottJake]

I couldn't quite get it perfect, but pressing F12 on the sphere's website indicates the font they're using throughout is 'Graphie' which is offered in the Adobe suite. When you use the 'thin' or 'light' versions of the font the number glyphs are almost an exact match for the image on the website. It makes sense that they would have a consistant style book and set of fonts they use for branding purposes

[@znacly1184]

Interesting video! A bit that you might find fun: they changed the image to be sequential. Also if I might take a shot on the font: I’m thinking “Graphie” - maybe “Graphie Book”? a sans-serif type. The 1, 2, and 6 seem to match pretty well to my untrained eye. Again, awesome video!

[@magickmarck]

I love this kind of stuff where they thought no one would notice or check! Bless you sir (and team)

[@D1g1talMess]

My hypothesis for that discrepancy between real pi and Vegas pi is that when they started making the wall with all the digits on it, they broke it up into sections that were all added to the wall separately from each other, and then someone who was making the website said, "Hey, we could borrow this first section to show the digits of pi on our website," without realizing that the numbers in that section weren't just going to the next rows down and following the correct order in this weird multi-digit-long column, but were actually continuing on in the subsequent sections of the wall which were ignored.

[@R.B.]

Since the have a picture of the wall, they might have actually started with a picture of that corner and then cropped and edited the photo to create the image used. If there are any projection artifacts from trying to photograph a flat wall, there would probably be some curvature. If not, then they might have used the source images from the CNC machine they used to etch. No real harm though. The building is still small enough that I doubt that difference would be measurable even if they used these numbers when engineering the building.

[@GreatWhiteElf]

Yeah, I came to the comments to see if anyone else came up with this conclusion. It make the most logical sense

[@sauercrowder]

This is saying the same thing as the solution discussed in the video, though.

[@werdwerdus]

Yes, I also watched the video

[@D1g1talMess]

@@sauercrowder @werdwerdus well I'm stupid, then, because I had a hard time deciphering the hypothesis that the video gave

[@juanussher5243]

For those who, like me, thought to go to compare the font of pi with the other math in the site, don't bother. They use different fonts from what I can tell. The 1 and the 2 are definitley different from the ones in the pi image. Man... I thought I was SO SO smart to catch Matt and Linus's oversight hahaha. It goes to show, pride cometh before the fall.

[@SolomonUcko]

Yeah, I tried Adobe's Graphie font in Firefox and it didn't work...

[@theobrayford4004]

Seems spot on to me. Bearing in mind they use both Graphie Light, and Book in the other webpage text.

[@orangenostril]

That seems like it reinforces the "CAD file from the wall" theory even more

[@EliteCuttlefish]

A similar thing happened to pi engraved in granite in the Portland Zoo /Washington Park train station. The best guess was a contractor was just given a dump of pi and the format or spacing confused them. Discovering and researching that was probably the most exciting part of the trip.

[@GretaKnauer1980]

Yeah, I just emailed a photo of that to Matt - hope he got it. I remember standing next to it and showing off to my friend as I recited pi from memory and was most disappointed when I got to the eleventh place and he said 'no'.

[@dudeonbike800]

This was all a bunch of fun! Funny how the solution was just waiting for you a click or two away! Can't believe I haven't seen it mentioned below in the comments (I didn't scour them.), but it seem like they'd have messed with the font KERNING to make their graphics fit. So it's not only he font style or size, but spacing between characters that can be adjusted.

[@philstav8269]

The font you were looking for re. the wrong numbers for Pi is 'Graphie Book'. It's an Adobe font and is used for the rest of the page - the reference was hiding in plain sight in the resources for the page as the font is also used in the main body text of the page. It's an exact match if you overlay the font onto the graphic.

[@coolmanjack1995]

Curses Linus the designer didnt know how to read the page source!

[@AxGryndr]

'How can they be so cavalier?' I think that basically highlights the difference between arts and engineering.

[@bouipozz]

numbers are numbers, right?? :p

[@MushookieMan]

The web designer isn't a math nerd, they were given copy and image files and directed to make these graphics and blurbs.

[@laslo67]

Yup, and I think I just confirmed this. I took the image from the website and adjusted the contrast and you can see a few extra pixels that the designer missed when removing the digits that would have otherwise been partially visible. I honestly don't remember if youtube allows image links in comments or not, but if so... i.ibb.co/DwzwM2j/badpi.png

[@petergerdes1094]

They may not be a web nerd but I guarantee you then understand how decimal numbers work and that you need to get the digits in order. They just didn't care (understandably, they are probably overworked and have a life they want to get back to and they aren't paid to make math nerds happy... indeed the owner probably appreciates this publicity).

[@killingtimeitself]

@@petergerdes1094 ultimately it probably doesnt matter either, im assuming this string compared to pi to the same level of specificity is basically identical anyway lol.

[@korg47237]

​@@petergerdes1094it's the fact that they *passively* didn't care. Who is going to their marketing website to get the accurate digits of pi? I certainly don't, there's also the factor of being super invested into something (in this case math) and if someone doesn't put in the same level of care as you, then they take offense. Hence this video 😜

[@AdmiralJota]

Someone made the image files. They're the one who's to blame.

[@MoxyMonitor]

It looks like they just fixed the pi value on the website. It was wrong earlier today but I just checked now (Oct 25, 9:00 pm Central US Time) and it looks correct. The 2's are still missing in the piers of the arches though.

[@F_L_U_X]

Maybe the website was supposed to have "..." after each string to indicate it's a really big number and you're trying to make sense of something else?

[@paiwanhan]

They changed their website after you released this video didn't they? Now the numbers on their website is in the correct sequential order.

[@jackc.5271]

I have never seen anybody so excited and drawn in by math as this guy is, he really made me enjoy learning about it for the first time in my life.

[@brianlane723]

4:30 Proper knife handling is to set it on the table and allow the next person to pick it up. You also need to hold it downward at your side while in transit and shout "Knife!" repeatedly.

[@Azeal]

I was just in Vegas for TwitchCon, and I'll just say that driving past that thing is downright intimidating. Such a feat of engineering!

[@justayoutuber1906]

What is TwitchCon? A convention for those with Turrets?

[@vex3091]

I feel like they did it on purpose, because they knew that, someone like a mathematician would look at their value of pi, and create like a 23 minute video regarding why its incorrect, in the procces giving them more recognition

[@uniman-bi1cm]

dude they fixed it! even the 2 in the bridge! they saw your video!

[@Rikerslash]

And they changed it on the page GJ! They even added the 2 at the bridge!

[@M_M_ODonnell]

I hadn't thought much about the finer details of typefaces until I had a friend who dealt with type and graphic design professionally. There's a _lot_ of detail in there; I'm still a "try it in a few fonts and see which one gives you the vibe you like" person generally, but for publishing something for broad release I'd really want to work with an expert to get everything ideal.

[@petehiggins33]

I generally use Comic Sans just to annoy the typeface nerds.

[@Domihork]

Finally a mathematician who publicly admits that the precision of pi after a few decimal points doesn't actually matter (for practical use) :D

[@toddhurst4167]

It still matters, just not very much (depending upon the practical application). =).

[@Crooked_Man]

@@toddhurst4167 it actually never matters for practical! :D

[@lightknightgames]

I like how you said "The next string does appear in Pi" as if it was a rare phenomenon :P, since every chain of (non-infinite)numbers appears an infinite number of times in Pi I understood you meant relatively close to the beginning, sensible chuckle at that one

[@WillHirschUK]

Of course anywhere in the known calculated digits of pi would be "relatively close" as they only contain about a billionth of all possible sequences of digits that length!

[@CM-lr7tf]

Actually, we don't know that for sure. It's generally assumed that π is normal, but nobody's managed to prove that yet.

[@JoseNovaUltra]

now this would be such a fun proof

[@Beertraps]

As already pointed out we do not actually know for sure if every sequence of numbers is in pi. The digits of pi could just not contain any 4 after the last digits we calculated. We just know that the digits of pi never start to just repeat.

[@angelmendez-rivera351]

@@JoseNovaUltraIt really would be, especially since never in recorded human history has it ever been proven, ever. It is still an open conjecture.

[@kovanovsky2233]

They have fixed the π on the science page!

[@claitonlovatojr]

I had to go to the page to check, and THEY FIXED IT!! Matt they clearly are watching your channel, maybe even used you videos to calculate the science for the building...

[@yeeeeeter]

I believe the font is Graphie. If you inspect the website for the font they used, that's what you'll find. And thankfully they aren't using different fonts between the portion of Pi pictured and the website body text.

[@TuberTugger]

Thank you for looking into this. I wanted to check the website code myself but they've updated things since this video released.

[@unvergebeneid]

I kinda love that someone looked at pi, thought the 2 looked kinda out of place and just deleted it 😂

[@Kyle-nm1kh]

It was just 2 much

[@edl5731]

Would make an outstanding "sun" to use in a solar system model. It would be larger than the the Sweden one.

[@IMarvinTPA]

Now we need a video on where the planets should be in relation to this size of sun.

[@joelculver1421]

Earth would be about 10 miles away by my calculations

[@superdupergrover9857]

I'm quite confident that there is some interactions between the different softwares and the copy/paste process used at play here. It could explain some of the typeface weirdness maybe? It would also explain how the clipped lines happened; there's a trick to getting a whole line without dragging the cursor completely across the screen. I'm sure most of you have it in muscle memory by now, you just start before the top line, then drag down to the next line where you want the highlighting to stop. Either a mistaken cursor drag, but I bet the selector tool was just a square instead of the normal line based one (or otherwise not what his brain expected). Being so used to: CLICK>DRAG>CRT+C>MOVE>CTRL+V to arrange all the math on that webpage, he finally made a mistake he couldn't catch. Being as Pi, even for math nerds, looks random past 10 digits or so.

[@erielle_rs]

What helped me to internalize why the surface area would relate linearly to the ratio between the diameter and the start of the cap is that there isn't variation in the displacement between points that touch the surface along the diameter when changing the orientation of the object. Compare this to a sphere which has some kind of appendage somewhere, changing the orientation of the object would change that displacement if the appendage happened to cross the diameter. Because the curvature of that surface is regular for the whole object, there'd be no additional variables to account for when considering the cap starting elsewhere along the diameter.

[@rianfelis3156]

I suspect at some point they just decided to use an image of pi that they owned the copyright for, and the one that they found was just the first panel of that room, with it continuing on a series of images set for that panel size and weirdness of non-monotype fonts. And the final inspection was just "yep looks right"

[@EebstertheGreat]

You can't copyright a typeface (at least in the U.S.), or the value of pi. So they could have just typed out a bunch of digits and used that. I assume the reason they used that vector "font" was to match the look of the exhibit, which I think is smart. I just wish they used the correct digits.

[@88porpoise]

​@@EebstertheGreatit is way more complicated than that. For example, while you cannot copyright a typeface, you can copyright the font programming. And whole you cant copyright the digita of pi, you can copyright a presentation of them. In this case, they can copyright the image created using the specific typeface and colours two write out pi. But, most likely, some graphic designer needed a graphic for pi, saw that image, and used it.

[@EebstertheGreat]

@@88porpoise You can copyright a font, but that doesn't matter, because that image is not a font. It is a numeral written in the Helvetica typeface. It didn't use any fonts at all (since as was pointed out, it used a vector image generated from a CAD interpretation of a font), but even if they used a font to generate it, that doesn't matter. A book does not have to pay royalties to the inventor of the font they used to write it. The copyright on a font is just there to protect the ability of the creator to sell that font, including things like hints. But any other person could go ahead and make their own font of the exact same typeface without violating any laws. And you certainly don't need to credit the font-creator to publish something. The copyright over a "presentation" of pi is not relevant here. The Vegas Sphere certainly could copyright their big pi installation like any other artwork, and idk, maybe they have. But that wouldn't affect whether they were allowed to just type the number on their website in Helvetica. Of course they could.

[@ETCubing]

As an American, I’ve always wondered how to visualize a meter, from now on I can confidently say to my American friends that it’s as simple as 7.5 melons

[@nerd1000ify]

It's almost the same as a yard.

[@swampertdeck]

Most people visualize meters with peoples length. A small child is 1m. A child reaches 1m between 3 and 4 years old. A very short adult is 1.5m and a very long adult is 2m. Most adults are between 1.60m and 1.90m (160cm and 190cm). A standard bed (the mattress) is 2m long (ikea only sells this length.) My bookcase (ikea billy) is 2m high and the middle shelf is at 1m. Most ceilings are between 2.5 and 3m. Hope thus helps visualizing :)

[@crabby7668]

@@swampertdeck and a London bus is approx 10 small children long , and a Wales is 100,000 x 100,000 London buses.🙂

[@PaulHester66]

Or it's 100 centimetres. Nice round figures. How many feet in a mile?

[@Jst4vdeos]

@@PaulHester66has your country been to the moon?

[@LadyTink]

Anyone who have memorized the first like 50 digits of pi, would notice that error pretty quickly, just on seeing the end and start of the lines, like I"m used to the section "46264" as being like a chunk of pi, and seeing 46 end like that hurt a little

[@mathisfun774]

Matt, the surface area of a sphere of radius r is the same as that of a cylinder of radius r and height 2r, minus the two end caps. It is a short hop to see that corresponding horizontal slices of the two objects have the same area. I, too, was blown away when I first noticed this.

[@StephenBoyd21]

I love a bit of Linus. His channel is so entertaining.

[@stevemonkey6666]

It's always good to see Matt with a good maths mistake😅

[@trummler4100]

For 3 tries in a row, I‘ve read „Vegans“ and wondered why you‘ve used the Singular „does“ in front

[@jmunt]

Did the same twice 😂

[@NoNameAtAll2]

cool, error-detection algorithm in natural language

[@appa609]

Is a Vegan someone from Vegas?

[@oberonpanopticon]

Glad I’m not alone

[@BigDaddyWes]

Vegans are well known for having their own approximation for Pi.

[@MatthewDawkinsUK]

Interestingly (maybe), the font they're actually using on the website for the text is called Graphie. But the numerals in that font don't quite match your faulty pi either. Which means the graphic designers who made the mathematical equations (and the mural) used a different font from what they used on the website. And it's also a different font again from what they used in the "sphere" logo mark. Which makes it three almost (yet frustratingly not quite) identical fonts on the same web page. As you say, so close, yet so far.

[@pi_f]

I think it's funny that I noticed the missing 2 at 8:59, completely missing the point of the missing digits to the right! I was wondering through the whole video how you didn't notice THAT one also, and then you cover it at 19:55 haha! Great video. Thank you for solving the best math mysteries!

[@dominnnio1]

Someone has already watched your video because they fixed a mistake on the website :)

[@mystak3n]

Alternate title: Matt accidentally turns a website about a giant almost-sphere into a font hunting ARG

[@Delicioushashbrowns]

It's amazing that they put so much math info on the website! Usually they try to hide that stuff as fast as possible :(

[@johnmorriss5308]

Especially since their very existence is based on their superior knowledge of the math laws of Probability...

[@johncochran8497]

Their approximation is pretty good. If you were to build something the size of Earth using it and then duplicated the result using the real value of pi, the difference between the two would be about 12 femtometers. I think that ought to be close enough.

[@peppermintmiso4341]

my occam's razor response for why there are so many maths in the website but got pi wrong is the designers and layouters were tasked by the people who worked on this to make a page detailing their process, they were given the words by the engineers, so the designers just threw a bunch of graphics to look smart with little regard for accuracy in this case pi also awesome insight on the design aspect of the mystery by linus!

[@Apes-With-Computers]

Amazing video Matt. I love your channel!