No, really, pi is wrong.

March 14, 2016 9:49 AM

The Tau Manifesto
posted by anotherpanacea (63 comments total)

I think calling it two-pi would be better. Here is my proof:

1 pie = good.
2 pie = better.
posted by nubs at 9:51 AM on March 14, 2016 [19 favorites]

You raise an interesting point. Let's discuss it. Two-pi or not two-pi, that is the question.
posted by benito.strauss at 9:53 AM on March 14, 2016 [13 favorites]

Recently mentioned in the pi-day thread by JoeZydeco
posted by filthy light thief at 9:54 AM on March 14, 2016

Good tagging.
posted by Chrysostom at 9:58 AM on March 14, 2016 [1 favorite]

PI IS A LIE!
posted by JoeZydeco at 9:58 AM on March 14, 2016 [2 favorites]

This would make things far too easy for the textbook makers. At least force them to come up with reasons why basic mathematics textbooks have to be completely replaced every year, don't hand them a change like this on a silver platter.
posted by zachlipton at 9:59 AM on March 14, 2016 [11 favorites]

So uh, happy half tau day!
posted by nzero at 10:10 AM on March 14, 2016 [6 favorites]

So I would have to divide by eight to get the area of a circle from the diameter? What do you want, blood?
posted by Huffy Puffy at 10:11 AM on March 14, 2016 [4 favorites]

Pi r round, not square.
Tau is poo.
posted by Chuffy at 10:14 AM on March 14, 2016 [6 favorites]

I have two personal arguments against tau:

1. It's much easier for me to mentally multiply by 2 than to divide by 2.
2. People who endorse tau remind me way too much of people who endorse bitcoin.
posted by phooky at 10:16 AM on March 14, 2016 [11 favorites]

Don't make my favorite pizzeria change its name to Tau/2.
posted by Foosnark at 10:17 AM on March 14, 2016 [2 favorites]

Shouldn't you post this on June 28th?
posted by Western Infidels at 10:18 AM on March 14, 2016 [17 favorites]

Pi is overrated. I always preferred cake numbers.
posted by tempestuoso at 10:22 AM on March 14, 2016

Ugh. Now I'm just going to have people in my office walking around saying "Happy Half Tau Day!!1" "There's half tau in the break room!" "How can you not like half tau? What's wrong with you?" "Finally an excuse to eat half tau on a Monday, but my thighs don't care if they're measured using pi or tau, amiright???!?"
posted by Clinging to the Wreckage at 10:29 AM on March 14, 2016

What should we call the ratio of the major axis of an ellipsoid bean to the diameter of the circular plate upon which it rests?
posted by the painkiller at 10:33 AM on March 14, 2016 [12 favorites]

On every axis I can think of, including theoretical appealingness, likely practical utility, pedagogical value, and probability of actually happening, this is exactly like trying to get the U.S. to switch to the metric system.
posted by a car full of lions at 10:34 AM on March 14, 2016 [9 favorites]

I was kind of hoping this would be a WarHammer 40K thing, but no dice.
posted by GenjiandProust at 10:35 AM on March 14, 2016 [10 favorites]

Let's all agree to celebrate Tau Day on June 28 by eating pizza *and* pie.
posted by duffell at 10:36 AM on March 14, 2016

I expect this to catch on like Dvorak.
posted by Navelgazer at 10:39 AM on March 14, 2016 [7 favorites]

Double.
posted by scruss at 10:42 AM on March 14, 2016 [3 favorites]

"Try explaining to a twelve-year-old (or to a thirty-year-old) why the angle measure for an eighth of a circle—one slice of pizza—is π/8."

There's a scene from the 1987 (?) Dragnet movie, in which Friday (Aykroyd) and Streebeck (Hanks) are at the zoo, and Friday is contemplating a group of kids who are looking at a lion whose mane has been shaved off, and he says to Streebeck, "How are those kids ever going to recover from this? What possible words exist which could let those kids know that that mane will grow back?" And Streebeck says, "Kids? It'll grow back." And the kids cheer.
posted by disconnect at 10:43 AM on March 14, 2016 [6 favorites]

Indeed, the horizontal line in each letter suggests that we interpret the “legs” as denominators, so that π has two legs in its denominator, while τ has only one.

Which suggests curly pi should be 2*pi/3? Hmmm.

I love geeky things and fun things. I wouldn't dream of trying to talk people out of obsessions which appear, to them, to be both.

But, the world already has too many named constants that are trivial multiples of other constants. (I'm looking at you, Angstrom and Jansky. You too, barred-constants, though at least there's some sort of consistency involved. Your all owe mankind a debt for the wasted memory we've devoted to your redundant definitions. You should be ashamed of yourselves.) Introducing another one is a terrible idea except as a fun goofy thing to argue about at the pub.
posted by eotvos at 10:44 AM on March 14, 2016 [4 favorites]

I'm giggling a bit that this is only coming up today because of the WRONG WRONGNESS of the US "least-significant-digit-middle" style of writing dates.
posted by indubitable at 10:44 AM on March 14, 2016 [10 favorites]

I honestly am for pi mostly out of inertia than any serious love for one notation over the other.

But some of these arguments don't actually hold water. Angles go from 0 to 2pi vs. 0 to tau? No, angles go from 0 to pi. Like, 0 to 180 degrees. Anything more than that, and you just have an angle in the opposite direction.

It's like saying "turn left 270 degrees." It's not wrong, but most people would expect you to say "turn right 90 degrees."
posted by explosion at 10:46 AM on March 14, 2016 [4 favorites]

TAU IS EXACTLY 6!
posted by the quidnunc kid at 10:53 AM on March 14, 2016 [9 favorites]

Ctrl+F "Greater Good"; 0 results. So it can't be genuine
posted by Gelatin at 10:56 AM on March 14, 2016 [3 favorites]

Would you like your pizza cut into six or eight slices?

Better make it six, I don't think I can eat eight.
posted by Confess, Fletch at 10:59 AM on March 14, 2016 [10 favorites]

Wolfram blog does an informed takedown of this:

...This shows that for formulas that have a complexity greater than 2 (most of them do) and for which the complexity is not always greater than 18, the improvement in switching from τ to π would be negative again, suggesting that we should not accept the switch. Unfortunately for supporters of τ, we do not live in a τ world....
posted by vacapinta at 11:05 AM on March 14, 2016 [7 favorites]

It's like saying "turn left 270 degrees." It's not wrong, but most people would expect you to say "turn right 90 degrees."

Turning left 270° and right 90° are only equivalent under degenerate circumstances -- either velocity or time-to-turn equalling zero. If neither is true, the resultant path is rather different.

Thus, when traffic is flying a left handed landing pattern, you don't make 270° turns to the right, you make 90° turns to the left.
posted by eriko at 11:05 AM on March 14, 2016 [3 favorites]

Right on schedule. One day after a slew of tweets and Facebook posts decrying daylight savings time, I get a bunch saying we should change from pi to tau.
Everybody wants to change the world.
posted by rocket88 at 11:06 AM on March 14, 2016 [3 favorites]

I honestly am for pi mostly out of inertia than any serious love for one notation over the other.

I'm willing to entertain the idea, but will withhold judgement until I see a comprehensive list of its strengths and weaknesses. The 'for' argument presents all sorts of supporting examples, because that what's 'for' arguments do. I will remain quite skeptical until I see a reasonably comprehensive list of examples of both tau and pi not working as efficiently as claimed, so an empirical comparison can be done. The one with the fewer problems wins, but if tau wins, that victory only means it moves to the next round. That next round would be about acceptance and adoption with questions like "a switch to tau is possible, but is it practical?"

If Feynman were around today, I could imagine him saying to the tau supporters "Okay, you've painted a pretty portrait of tau here, but I want to see the warts, the blemishes. Your presentation shows tau's 'best side' but I can't help but think you've chose this particular angle for a reason. I won't buy it until you show me the other side of tau - the uglier side. Only then I can make a decision on its merits."
posted by chambers at 11:21 AM on March 14, 2016 [2 favorites]

It's like saying "turn left 270 degrees." It's not wrong, but most people would expect you to say "turn right 90 degrees."

Wait, wait. Turning left 270 degrees is not the same thing as turning right 90 degrees, just as spinning around 360 degrees is not the same thing as standing still.
posted by baf at 11:24 AM on March 14, 2016 [4 favorites]

Vi Hart weighs in, with cherries and hand crust.
posted by hearthpig at 11:26 AM on March 14, 2016 [1 favorite]

Vacapinta, the analysis that Fortuna does in that article does conclude that "only 18% of formulas considered involve 2π, suggesting that τ, after all, would not be a better choice". However, she defines "better" formulas as "fewer terms", and gives equal weight to huge unwieldy dozen-plus term formulas as basic high school math equations. Once you get over a certain complexity of equation, frankly a computer algebra package is going to support the work so who cares.

Hartl's argument (in OP) is that number of terms is not the right measure of simplicity. There are cases where π appears in a formula by coincidence - where a factor of 1/2 cancels out the 2π. He defines better formulas as "leading to more clear insight into relationships between mathematical concepts", case in point why 1/2 τ r^2 is a better construct than πr^2 .
posted by anthill at 11:26 AM on March 14, 2016 [4 favorites]

I think calling it two-pi would be better.

Even betterer? Just call it "Shakur."
posted by ROU_Xenophobe at 11:29 AM on March 14, 2016 [3 favorites]

I have a degree in mathematics, and keep it pretty active... and I went into that article not expecting to be convinced but I was. If I were teaching math again, well, I'm not sure what I'd do to be honest. Tau is better, but pi is more common, do you want to teach people something obscure?

Perhaps I'd teach the kids tau and then partway into it, introduce pi as tau/2?
posted by lupus_yonderboy at 11:48 AM on March 14, 2016 [1 favorite]

TEACH THE CONTROVERSY!

Seriously, this made me think harder about what's going on with pi in the first place. That's the point: pi is obscure, as it's now taught. The tau/pi controversy helps you think about what a cool thing it is we're doing in the first place.
posted by anotherpanacea at 11:59 AM on March 14, 2016 [5 favorites]

Pi r round, not square.

Cornbread are square!
posted by Greg_Ace at 12:12 PM on March 14, 2016 [5 favorites]

...Or should I have added that to the "math pun" thread?
posted by Greg_Ace at 12:12 PM on March 14, 2016

Wouldn't this energy be better spent on convincing physicists to switch the signs for positive and negative charge so that a positive charge would represent the presence of electrons rather than the absence of negatively-charged electrons?
posted by straight at 12:19 PM on March 14, 2016 [9 favorites]

Vi Hart weighs in, with cherries and hand crust.

In which she illustrates how much less confusing Tau would be by accidentally writing R/D = Tau ("R is for 'R... circumference' and D is for 'Dradians.' ")
posted by straight at 12:25 PM on March 14, 2016

I used to kind of like Tau, but the pretentious, lock-step self-righteousness of angry, thin-skinned Tau-bros has really turned me off. Pi has served us just fine. Just look at all the circles everywhere - bicycle wheels that convey us happily from place to place, yogurt containers that hold our sustenance, the warming, life-giving sun (and, yes, delicious pies - both savory and sweet). Tau may look good in theory, but I don't want to throw out everything and start over with new circles. Now I'll sit back and wait for the Tau-bros to lose their shit.
posted by univac at 12:40 PM on March 14, 2016 [8 favorites]

I'll switch to tau when the US goes metric.
posted by MtDewd at 12:47 PM on March 14, 2016

Brought to you by Citizens Bent on Reminding You at Every Possible Goddamn Opportunity that the Vagina Is the Inside Part, ACTUALLY
posted by middleclasstool at 12:55 PM on March 14, 2016 [2 favorites]

Two words: Euler's identity.
posted by belarius at 12:55 PM on March 14, 2016 [3 favorites]

I made a Bob Andy Pi(e). It was delicious.
posted by Sophie1 at 12:57 PM on March 14, 2016 [1 favorite]

Two words: Euler's identity.

Oh, you mean this?

e^iτ=1

BOOM. Euler is better with tau.
posted by anotherpanacea at 1:01 PM on March 14, 2016 [14 favorites]

The Feynman point has an seven 9's in tau as opposed to a measly six in pi. If that isn't conclusive evidence of tau's superiority I don't know what conclusive evidence could possibly look like.
posted by Proofs and Refutations at 1:04 PM on March 14, 2016 [1 favorite]

Eight 9's, of course!
posted by Greg_Ace at 1:35 PM on March 14, 2016

8/it=0, classic formula for pie or pizza. I once used a pi to calculate carbs in a six inch, vs eight inch flour tortilla.
posted by Oyéah at 1:40 PM on March 14, 2016

I used to kind of like Tau, but the pretentious, lock-step self-righteousness of angry, thin-skinned Tau-bros has really turned me off. Pi has served us just fine.

Pi is Clinton. Tau is Sanders. Everything is the election. There is no escape. Abandon all hope, ye who open up a browser in the next eight months.
posted by officer_fred at 3:14 PM on March 14, 2016 [4 favorites]

Tau = nice.
Tau culture = eh.
Cost to fix all those textbooks = ugh.
posted by solarion at 4:14 PM on March 14, 2016 [1 favorite]

indubitable: "I'm giggling a bit that this is only coming up today because of the WRONG WRONGNESS of the US "least-significant-digit-middle" style of writing dates."

Here, the article was posted on 2016/03/14.
posted by Bugbread at 5:59 PM on March 14, 2016 [3 favorites]

Exp(i*tau) = 1 is a Different Identity than ex(i*pi)=-1. Among other things, you can deduce the first from the second, but not the second from the first. Additionally. The interesting part of Euler's identity is that it shows that complex powers of positive numbers can be negative, whereas real powers of positive numbers must be positive. Euler's identity conveys interesting information about the structure of complex eponentiation that the tau identity doesn't.

Anyway, I will continue to roll my eyes at tau supporters, and dread having to put up with their tedium next pi day.
posted by Elementary Penguin at 6:17 PM on March 14, 2016 [3 favorites]

This was really interesting and made me learn things that I had previously memorized.
posted by klangklangston at 6:19 PM on March 14, 2016 [1 favorite]

The interesting part about e^iτ = 1 is that it shows that the complex exponential is periodic with period τ, which in my opinion is a much more significant fact.
posted by a car full of lions at 6:41 PM on March 14, 2016 [6 favorites]

Every time you change a math textbook, some religious extremist slips anti-set-theory nonsense. Pick your battles carefully.
posted by Navelgazer at 6:53 PM on March 14, 2016 [2 favorites]

I like Tau because it seems to have the possibility of making trigonometry much more accessible.

Memorizing the "important angles" really turned me off from using radians or bothering to understand what one would could express with them.

I'm sure this happened with other people too, which is a shame. Once you get past the memorization bit, things become quite profound.

In my mind this ends up being a lot like a debate about Imperial vs SI units. One representation is familiar, and the other trivializes a whole class of problems.

Tau Culture and resistance too it are things I had never heard of. Odd. Why are people weird on the internet?
posted by ethansr at 8:28 PM on March 14, 2016 [1 favorite]

iirc the internet was invented by weirdos to communicate with other weirdos and hasn't changed much since
posted by indubitable at 5:33 AM on March 15, 2016 [2 favorites]

Maybe we could just crush them together into TauPi Day, where everyone wears a kind of greenish-yellow brown. That should still the controversy....
posted by GenjiandProust at 6:57 AM on March 15, 2016

I wonder which historical accident would be harder be undo at this point, using π instead of τ, or ascribing negative charge to electrons.
posted by dfan at 9:14 AM on March 15, 2016

The interesting part about e^iτ = 1 is that it shows that the complex exponential is periodic with period τ, which in my opinion is a much more significant fact.

But, your version of identity does not actually show that. For instance, the complex exponential could be identically equal to 1. Or it could have any period which is a sub-multiple of τ

Furthermore, it's highly significant that we can take logarithms of negative numbers, and by extension, all complex numbers. (That's at least as significant as periodicity.)

Euler's identity e^iπ = -1 contains strictly more information than the tau-based version, which does not determine the value of e^iπ - it could be +1 or -1, since you have basically squared the identity when passing from pi to tau, losing information in the process.

So the correct version of Euler's identity using tau would be e^iτ/2 = -1 . So now the "most beautiful equation in mathematics" (Feynman) has that ugly "/2" in it. (Once denominators start appearing in equations, they get a lot messier.)

In general, it's always going to be nicer to insert a multiplicative factor of 2 (as in 2π) as opposed to having to divide by 2 when we don't want it (τ/2). All of those division operators get ugly. Finally, τ is already used in many math and physics formulas (torsion, stress, etc - see https://en.wikipedia.org/wiki/Tau), and it is far too easily mistaken for "t" or "T" which are also overloaded (time, temperature, etc).
posted by crazy_yeti at 12:30 PM on March 15, 2016 [3 favorites]

Michael Hartl of the manifesto, and I are both members of the Happy Tau Day facebook group.
posted by Obscure Reference at 5:57 PM on March 15, 2016