The Truth About the Most Controversial "Number"

[@ComboClass]

This extra-long episode is my presentation about if/when/how 0.999 (repeating) equals 1. Most of this episode is mathematical demonstrations, but there is also a philosophical edge to this topic, so leave a comment letting me know your personal opinions/beliefs about this "number" (hopefully after watching this whole episode to see all of the misconceptions I cover). And/or leave a comment if you can count how many squirrels appear in this episode haha.

[@donaverboxwood]

Pardon if this is a stupid question, but in regards to infinite strings of digits in decimals, would it be fair to say they are a different kind of string than finite strings? (I mean, obviously yes, but let me explain) What I mean is, it would be completely incorrect to have a number like 0.000...(infinite 0s)...0001, where the infinite string of digits is not the last string overall, right? So there has to be a difference between what a finite string is and what an infinite string is, despite being made of the same thing (digits). I guess what I'm asking is, would it be more accurate to say that decimals can have infinite strings of digits only if the infinite component is the smallest (rightmost when written out) component? Again, sorry if this is complete nonsense I'm saying. I am by no means "good at math".

[@TaleTN]

I counted 4.999999999... squirrel appearances.

[@diribigal]

@@donaverboxwood when you're talking about real numbers in their standard decimal forms or "strings" in most other senses, yes, an infinite string like these cannot have a right endpoint. However, that doesn't mean the idea is inconceivable. If an infinite string is normally like "there's a first character and a second character, and similarly a character for every counting number", then you could certainly make up something like a super-string which has a character for every counting number, and then three extra characters which are considered to come after all of the others. This is getting very close to the mathematical idea of "ordinals".

[@MrDannyDetail]

@@donaverboxwood Maths is the study of patterns, not the study of numbers. If you mean you are not good at manually performing additions, subtractions, multiplications or divisions where the numbers are not trivially small and easy to work with then it is arithmetic you are not good at, rather than mathematics (and in any case you're probably better than you think at arithmetic). What you demonstrated in your original comment is the ability to see the range of patterns already exisiting in a mathematical system and then concevie an entirely new way of extending that system with new patterns that build on the exising system, rather than merely replacing it with a whole new system. Being able to conceive of ways of extending patterns beyond what is 'normally' done in maths classes is actually being very good at maths. Having a play with what happens if you put a finite rightmost digit (or digits) beyond a infinite string of digits on the righthand side of the decimal point could lead to all manner of interesting conclusions, to new ways of viewing existing open maths problems etc so the ability to have 'outside the box' thoughts like this about mathematical systems is what enables mathematicians to keep pushing the boundaries, finding out new things and making new theories. It's a shame that school systems in many parts of the world leave a lot of their pupils thinking that maths is just about doing hard additions, subtractions, mutilpications and divisions and similar other things like square roots and so on when really that is just arithmetic, which is merely a mathematician's basic tools for doing actual maths, and for which we have extremely good calculators and software these day anyway, whilst true mathematics almost always requires human inquisitiveness, inutittion and creativity which a machine cannot really replicate.

[@mesplin3]

Let x = 9 + 90 + 900 + 9000 +... I like that 0.999... = -1 * x.

[@GreyJaguar725]

This guy spent an infinite amount of time writing an infinite amount of "9"s after "0." for a video. Respect.

[@sirfzavers8634]

Would’ve been cool if he’d shown them all… but then the video would be infinitely long and he wouldn’t get any full views. ☹️ Edit: At least it’ll keep that fire fueled infinitely (we’ve done it boys; we’ve prevented the heat death of the universe).

[@silver6054]

@greyjaguar725 Don't think he did. He spent 1 second writing the first 9, half a second writing the next, a quarter of a second writing the next and so on (practice makes perfect). So he did the whole thing in 2 seconds, which really does earn respect.

[@sirfzavers8634]

@@silver6054 that could explain the lack of the time machine needed for our theory… 🤔

[@GreyJaguar725]

@@sirfzavers8634 I think he's referring to the sum to infinity of the geometroc series:1 +1/2 +1/4+1/8+... Which is a/(1-r) where a is the first term and r is the ratio (next term /previous term) So plugging in the numbers we get: 1/(1-[1/2])=1/(1/2)=2 so it takes 2 seconds to write infinite 9s.

[@tacobell2009]

​@@sirfzavers8634 Maybe he did, but the video is still uploading...

[@first_m2999]

Domotro has mastery over squirrels and numbers. If he would only learn to control fire, he would be unstoppable.

[@hkayakh]

And a way to prevent things from falling over

[@kamikeserpentail3778]

Squirrel Girl just needs squirrels to beat Thanos, Dr. Doom, Galactus, whomever. He's got this.

[@nitehawk86]

There's a reason why Squirrel Girl is invulnerable and all of the fire based superheroes are not.

[@MawdyDev]

He's dual classing Wizard and Druid, it might be hard for him to continue leveling if he adds Sorcerer to that list

[@peppermann]

😃🤣❤️👏👍

[@nitehawk86]

I just wanna way the camera work, on this episode is particularly fantastic. Capturing the disaster just as it happens without taking away from the lecture.

[@ComboClass]

Thanks. Shout out to my main camera guy Carlo (who’s in the credits). Although I “direct” the episodes, he has some freedom behind the camera and helps capture all the rarities :)

[@kylebowles9820]

😂 that's such a perfect way to describe this channel in general; love the chaos

[@Zombie_Chess]

This is the best video explanation I have ever seen talking about this phenomena in our arithmetic system. Thank you. You're a great teacher.

[@ComboClass]

Glad you enjoyed and it helped you learn, thanks for the compliment :)

[@strangedivine]

You’re right, he’s a great teacher. I sometimes struggled with math and it was usually because of the teacher/prof’s approach in teaching.

[@jackputnam4273]

So glad i clicked on the first combo class vid that was recommended to me. I was immediately hooked by domotro’s style and it just keeps getting better! Such an amazing channel and it deserves a lot more attention :)

[@ComboClass]

Thanks, glad you’ve been enjoying! :)

[@TheBalthassar]

In arguments like this I always like the engineers answer "It's close enough, it fits the spec."

[@willo7734]

It’s within the tolerance.

[@JimmyMatis-h9y]

I like the engineering solution to "is the glass half full or half empty?" The glass is twice as large as it needs to be. 🤓 🥰 Lol

[@mattiviljanen8109]

(Edit: a few minutes later just this was covered in the video!) As a kid when I learned about 0.999... = 1, the mind-opener thought was that there is more than one way to represent a number, e.g. 1.5 = 3/2. As for the usual 1 / 3 = 0.333... --> 0.333 * 3 = 0.999... --> 1 - 0.999... = 0.000... counter-argument, the trick is to get infinity right. In order for the 0.000... to ever end, there would need to be a final non-zero digit. But as per definition, 0.000... does _not_ have a final digit, hence it must be all zeros, and be exactly equal to zero.

[@diribigal]

I teach things like the surreals and the hyperreals and I'm very pleased with how you handled things here. My one real quibble is that around 17:12 when you defined the archimedean property, I wish you'd said/written "integer n" instead of "number n". Excellent work giving a fair and clear presentation that doesn't go too far into irrelevant detail!

[@ComboClass]

Thanks for the feedback, and I’m glad you enjoyed! :)

[@pedrogarcia8706]

the archimedian principle is not just true for integers though.

[@diribigal]

@@pedrogarcia8706 In a nonarchimedean ordered field like Robinson's Hyperreals, you can always find a "number" n. If a and b are positive, then certainly (2b/a)×a is larger than b, even if a is infinitesimal and b isn't, for instance.

[@stickmandaninacan]

Somehow the chaotic constantly interrupted style of presentation in combo class is actually really effective at keeping the attention of my adhd brain, it feels soothing 🧠 cute squirrel

[@Fire_Axus]

your feelings are irrational

[@samueldeandrade8535]

​@@Fire_Axus haha. More than irrational, transcendental.

[@rmdodsonbills]

I believe it's been proven that the infinite series 1/2 + 1/4 + 1/8 + 1/16 + 1/32 + 1/64 ... (and so on) equals 1. Has to be equal to 1. In binary that's represented by 0.11111111.... (and so on). Seems like a similar logic would work for 0.9999999999... (and so on).

[@MuffinsAPlenty]

"I believe it's been proven that the infinite series 1/2 + 1/4 + 1/8 + 1/16 + 1/32 + 1/64 ... (and so on) equals 1. *_Has to_* be equal to 1." (emphasis added) It _has to_ be equal to 1 in the same sense that Domotro talked about in the video. It doesn't really _have to._ There's no _a priori_ reason that 1/2 + 1/4 + 1/8 + 1/16 + 1/32 + 1/64 ... should have any value at all. However, if we impose upon ourselves the restrictions that it _should_ have a value, and that value should be consistent with certain arithmetic properties working in a reasonable way, then we have no other option but to recognize 1/2 + 1/4 + 1/8 + 1/16 + 1/32 + 1/64 ... as being equal to 1. However, this conclusion relies on self-impositions, not on some universal truth or "nature" or anything like that.

[@howdy832]

In Knuth's base 1+i, any gaussian integer is represented as a + b(1+i) +c(1+I)²… wher the coefficients are 0, ±1, or ±i. Each integer has 4 representations, where leading non-zero coefficient is each option. E.g. 1 is either 1, -i +(1+i), i -(1+i), or -1+(1+i)² -(1+i)³

[@howdy832]

You can do this with eisenstein integers too: the digits become 0, 1, w=-½+sqrt3/2, & z=w², allowing multiples of ±1, while the base is b=1-z. Now you can write 1 as 1, b+z, or w - w*b

[@samueldeandrade8535]

Nice ... very nice.

[@briangronberg6507]

I’m thrilled I found your channel! This was a really solid presentation and I appreciated the reference to the p-adics and the small taste of the idea that there exist number systems/algebras that may not satisfy commutativity or even associativity like the quaternions or octonions.

[@Ninja20704]

While I never really doubted it, the most convincing argument to me is that theres no number you could fit strictly between 0.99… and 1. Most people can see that intuitively, but there are also rigorous ways to show that. Usually i just tend to say if we don’t accept it, we cannot accpet any fraction with a non-terminating decimal, like 1/7 or 1/11 as well. I appreciate you adressing topics like this, maybe do more of them.

[@BlackBull.]

0.99...95

[@pepebriguglio6125]

Per my intuition, I agree because an infinite string of decimal 9's WILL get literally infinitely close to the number 1, and only 1 can be 'infinitely close to' 1. BUT technically, I don't find it difficult to find an infinite amount of real numbers between 0.999... and 1. Of course I must be overlooking something. But here it is: 0.999... = S(9/(10^n)) for n=1->inf. But this is an infinite sum with ordered place holders (n=1, n=2, etc.). So let's construct an infinite sum, which approaches 1, say 11/10, times faster, which would be: S(99/(100^n)), n=1->inf. Normally we would say that this is just an alternative way to describe 0.999..., because the decimal places would then simply be occupied pair-wise, instead of one by one. But still it stands to reason, that for every value of n, the number grows by 11/10 more than in the case of S(9/(10^n)). Another way to look at it, could be in base100. Here we have 0.99;99;99;..., which again would approach 1 by a factor of 11/10 faster than 0.999... in base10 would. So I suppose the question is, whether 'faster than' implies 'bigger than', when it comes to infinite sums.

[@pyropulseIXXI]

There is a number, and it is 0.99... + ε, where ε is an infinitesimal. You can literally prove 0.99... = 1 with infinitesimals, so idk why he said introducing them messes things up; it doesn't. The hyperreals are an ordered field with all the same properties as the reals, so associativity and commutativity holds. The reals are a subfield of the hyperreals, just as the rationals are a subfield of the reals. What you meant to say is that there is no *_real number_* that fits in-between 0.99... and 1. Also, your intuition says that 0.9... = 1 is obvious; or at least my intuition does. It is so far beyond obvious, but saying "it is hard to prove, therefore the intuition is wrong!" is pure absurdity. Proving 1 + 1 = 2 rigorously is also quite hard, for the non-math initiated, but we don't say our intuition of 1 + 1 = 2 is wrong because of a challenging proof

[@martind2520]

@@pyropulseIXXI No, you are incorrect. 0.999... + ε is not a number between 0.999... and 1. 0.999... _is_ 1, they are the _same number_ so there can be no number between them. 0.999... + ε is equal to 1 + ε, which is a number slightly higher than 1.

[@martind2520]

@@BlackBull. You number ends, it ends at ...95. The number 0.999... _doesn't end_ and so is larger than your number.

[@EpicMathTime]

You can get as close as you want to 1 with _finitely_ many 9s, and 0.999... is greater than all of those numbers. The common objection that continuing to add 9's will "never reach" 1 does not make sense as an objection, because such a process never reaches 0.999... either.

[@agentofforce3467]

Its probably not possible to add an infinite amount of numbers.

[@MuffinsAPlenty]

Perfectly explanation for a common misconception, like always.

[@isaacbruner65]

@@agentofforce3467 it's absolutely possible. Just look at any convergent infinite series, for example.

[@otonanoC]

He fed squirrels, and burned a guitar. After the smoke cleared, we concluded 0.9999.. = 1.0

[@eiman2498]

I love watching this channel. You always upload interesting content that never fails to enlighten others (including me) !

[@olly8453]

Subscribed! You have given this the most thorough, intuitive explanation I've ever seen. You are exceptional as a teacher. Also, I love the chaos of your approach and the set lol It's a great schtick that keeps the vids entertaining. If you keep going you're going to hit 100,000 subs and beyond in no time. Keep up the great work.

[@WillToWinvlog]

When I was a kid we used to assume it was wrong but we used to troll each other saying if 1/3 = 333... then 3/3 = 999... I guess our intuitions were correct!

[@arcturuslight_]

I remember back in school opening an algebra textbook on a small print "conventions" section, where one of the points is "for the purposes of this book we will define 0.9...=1"

[@matematicke_morce]

18:43 Here we see Domotro and the squirrel, a failed version of Achilles and the tortoise where the animal actually runs off to infinity

[@busomite]

I’d seen all the explanations but the one about points on a number line, that’s one I hadn’t considered before. Somehow that lands well with me, it says they need to be the same point. Very cool, thanks!

[@JonBrase]

The binary version of this gets really interesting. Two's complement is used to represent signed integers in computers, but some early machines used one's complement. But if you allow an infinite number of digits on either side of the radix point, two's complement and one's complement are equivalent.

[@CassandraComar]

these are the 2-adic numbers. some of the other p-adics can even represent i (ie solutions to x^2 + 1 = 0).

[@JonBrase]

@@CassandraComar Not exactly. The 2-adics don't include digits to the right of the radix point.

[@CassandraComar]

@@JonBrase the 2-adic integers don't but the 2-adic rationals do. they represent fractions with power of 2 denominators.

[@JonBrase]

@@CassandraComar I'm still a bit nervous about saying that what I'm talking about is too closely related to the p-adics, because there's some weird topological stuff going on with the p-adics that I don't understand and I'm not sure if it's intrinsic to all digit sequences extending infinitely to the left in a positional number system, or if it's just a useful topology to define on top of such digit sequences for the type of problems the p-adics have been used as a tool for. I think there may be multiple concepts in that space that are related to the p-adics in terms of their representations in a positional number system, but quite distinct in their deeper structure.

[@hughobyrne2588]

For years after I learned about 1s complement and 2s complement, I had this nagging feeling that the extra '+1' step of 2s complement was... hiding something. It took me a long time, but I came to the same conclusion as you did, including all the digits after the 'point' makes it all harmonious.

[@SirWilliamKidney]

I went from liking this channel to loving this channel @18:31 haha

[@TankorSmash]

This was both greatly educational and yet greatly uncomfortable. Looking forward to more!

[@ehxolotl4194]

17:10 As written, "Archimedian property" should be spelt "Archimedean property", and it would be inaccurate, pick x=0, y=1 and we have no number n such that nx>y. If x, y are restricted to the positive reals (and n to a positive integer), this would work.

[@ComboClass]

You're right. I misspelled it, and also didn't clarify the positive/integer restrictions, so I added a clarification to the description.

[@consciouscode8150]

After watching, I think I would consider it a notational quirk which emerges from the imprecision of what is meant by overbar, ellipses, etc. Rather it's probably better to think of real numbers represented using base-10 notation constructively, such that 1 approximately equals 0.9, 0.99, 0.999, etc but this notation alone can't ever equal 1. As soon as you say some variation of "and so on" however, what you've effectively done is taken the limit of the pattern - so it becomes almost obvious that it would be exactly equal to 1, because notationally it's essentially the same as an explicit limit. But maybe that doesn't feel so obvious because we think of the overbar or ellipses as being part of the base-10 notation itself rather than an implicit operation.

[@kazedcat]

All real numbers are defined by an infinite sequence of rational numbers. So when the domain is in the set of real numbers then it is by definition a limit.

[@CatherineKimport]

The argument that finally got me *comfortable* with the idea that 0.99999.... = 1 was one about how there isn't anything special about base ten. So, like, assume that 0.99999.... was some number infintessimally smaller than one. Then, shouldn't hexadecimal 0.FFFFF...... ALSO be some number infintessimally smaller than one? Would it be the SAME number as decimal 0.99999....? That seems weird, because 0.9 and 0.F are not the same, nor 0.99 and 0.FF, nor 0.999 and 0.FFF, and so on. So if 0.99999... and 0.FFFFF... represented DIFFERENT numbers, then that would mean that every base had a unique set of numbers it could possibly represent, and had a whole bunch of gaps about numbers that it COULDN'T represent. But if 0.99999.... and 0.FFFFF... both secretly equal 1, then those gaps go away. And the latter just felt less uncomfortable than the former.

[@mrmistmonster]

I uhh didn't prove but demonstrated this to myself with Zeno's paradox shenanigans a month ago. If Achilles starts 90 meters behind the Hare and moves at 10 m/s while the Hare moves 1 m/s. If you go through it you get to Achilles passing the Hare at 9.99999999 etc meters past the Hare's starting point. But if you just solve the equation you'll get 10 meters.

[@Kopiovastaava]

I saw a squirrel, some blue tetrominoes, burning stuff and there might have been some numbers somewhere along the way. Lovely stuff as always.

[@nnnnick]

some day i will understand why this man has so many clocks

[@Programmable_Rook]

Assuming a clock ticks exactly once every second, a clock that ticks normally has slight variation from other clocks, meaning it could be wrong at every point of the day. A clock that doesn’t tick is exactly right twice a day. So if you have many clocks that don’t tick, they will be exactly right more often than a normal clock. His clocks are more likely to tell you the exact time of day than a normal one. The real puzzle is why he doesn’t make them tick backward, then they’d be exactly right 4 times a day.

[@MarloTheBlueberry]

How is a backward.clock right four times a day?

[@AaronHollander314]

He hates to be late

[@StevenLubick]

You still have time. 😀😀😀😀

[@MarloTheBlueberry]

Bro.........@@StevenLubick

[@ThisCanBePronounced]

9:42: Wow. While I admit I may have caught a glance, I wasn't looking at the screen but I instantly RECOGNIZED the sound of what had just fallen over. It's been 20 years since we threw that thing away after finally being too damaged.

[@maxerboi20]

Combo class be comboling my brain

[@JerusalemStrayCat]

I am reminded of the bijective base notation system - I don't remember whether it was covered on the channel yet. The idea is that instead of having numerals from 0 to b-1 (for base b), there would be numerals from 1 to b. This prevents quirks like 0.999...=1, but also cannot represent 0, among other drawbacks.

[@orterves]

You're right - I think something that maybe isn't taught enough in school are the constraints of the maths people are taught. There is confusion about the answers to questions like this because people don't realise that the answer depends on the rules of the system they are working with. I think that this also applies to many disagreements in life, people argue about some question not realising the question doesn't even apply given the constraints of the topic. Perhaps in general we should spend more time figuring out where we really are before arguing about where we want to get to.

[@justsomeguy5628]

This video is great. Btw, floating point arithmetic (of any arbitrarily large but value) is an example of a non-archemedian system, as floatingpoint +0 is the smallest number.

[@strangedivine]

Math was not my best subject in school, especially post-secondary math, but damn you make it fascinating!

[@glarynth]

Been waiting for this one. Well done!

[@pvzpokra8602]

why does this guy speak in 0.75x speed

[@samueldeandrade8535]

Sorry to tell you,he doesn't. It is you. you hear in 0.75 speed.

[@Ethan13371]

Since his speech contains 25% more info per word than normal, he slows it down for us plebeians

[@JimmyMatis-h9y]

Because we're all in too big of a hurry

[@RichardBuckman]

I realized that what bothered me about it when I first learned this is it seemed like there could be a way to make it so these could be considered different. Now I realized the concept I was sniffing was the infinitesimals and hyper real numbers that can be used to define nonstandard analysis.

[@martind2520]

Except that in the hyper-reals 0.999... is _still_ exactly equal to 1.

[@RichardBuckman]

@@martind2520 In spirit though, it was neat to find out that the idea I had a long time ago before I knew much math had some merit to it even if it didn’t apply directly.

[@johnlabonte-ch5ul]

​@@martind2520I'd like to see that proof. I could learn a lot.

[@Chris_5318]

@@johnlabonte-ch5ul Karen, you have proven that you are incapable of learning any math. You don't even know that if a number can be written as p/q where p and q are natural numbers, that it is a rational number. You don't even know how to write 0.999.... In the surreals and the hyperreals, 0.999... = 1.

[@kqawiyy]

I used to deny that the two were equal, but now I see just how wrong I was in *many* different avenues. Thx

[@gonegahgah]

You were right the first time.

[@AlexanderScott66]

No, you were right. What people leave out is two fold. One, the infinite sum series specifies that it's the limit. Two, limits do not mean the function(in tis case 9/10^n)has to equal anything, rather, it simply approaches it, getting closer and closer. It explains why there's no number between(although, if we were to talk about just integers with no decimals for illustration purposes, there is no number between 1 and 2, so is 1 equal to 2? No, because being as close as possible doesn't mean it's exact), it explains why it mentions limit in the sum of an infinite series, it explains why both sides think the way they do. But no. People have to argue with baseless facts, like saying 0.333... or 0.999... is even defined at all, despite an infinite sequence inherently being unable to be defined, which is where Wiki nerds get it wrong. Why there's no number in between? There is. You'll use a number line and say plot it, but what about plotting based on precision? Plot 0.9, zoom in, then 0.99, then zoom in again and 0.999, so on and so on: 0.9999, 5 9s, 6 9s, 60 9s, 100 9s, 9 novemdecillion 9s. Tell me when you mathematically can't zoom in and plot again. TLDR people forget the beautiful thing called limits and how they work.

[@IsntPhoenix]

0:34 then why is 4>3 not equal to 4 ≥ 3

[@TerranIV]

This reminds me of Fourier trigonometric series that basically shows that you can make any shape out of an infinite number of smaller and smaller cosine and sine waves. Its like everything is made of an infinite series of waves, but we just mostly interact with things that have a harmonic form.

[@ryewaldman2214]

23:42 "Any number that has a terminating decimal expansion ... will have another form [with infinite digit string]" I think you mean any number other than 0. I could be wrong, but i cannot see how to represent 0 with a digit string terminating in infinite 9s. I think to generate a second decimal representation of a terminating decimal number, you have to consider which size of zero your number is on, so you can take the least digit down (toward zero) by one before appending the infinite string of 9s. In zero, you cannot take the least digit down toward zero by one. It's the same type of singularity that occurs with a compass at a magnetic pole, you are already at "zero" so any step you take can only go in the wrong direction.

[@ComboClass]

True I should have said non-zero there. I’ll add a clarification to the description

[@erwinmulder1338]

Can you repeat part of that? I got distracted by a squirrel.

[@dananichols349]

Just a thought... If I travel at 0.999(repeating) the speed of light, would I be traveling at the speed of light???

[@gabrielgabi543]

Haha still need infinite energy

[@BenHebert-no4qp]

I believe base i has an infinite amount of decimal representations for numbers, as the values for each digit position repeat every four positions. For example, i^8 = i^4 = i^0 = i^-4, therefore 10000000 = 10000 = 1 = 0.0001, which would all represent the number 1. Despite having infinite representations for real and imaginary integers, base i has no representations for non-integer numbers.

[@cuomostan]

So what would happen if I multiplied .99999… by n, where n is any real number? Would that be equal to n?

[@cmilkau]

Did you know 1/99 = 0.01010101..., 1/999 = 0.001001001...? Stumbled over this (in fact the general geometric series limit) on my own in middle school and turned it into a popular little school calculator program that could recover arbitrary fractions from their infinite decimal representation.

[@mrosskne]

so, the frac > dec button that every calculator already has?

[@MuffinsAPlenty]

It's very nice to be able to figure out a pattern like that at your own, especially at such a young age!

[@TerranIV]

This is an amazing video! So educational and entertaining. That squirrel is hilarious! :)

[@adamswierczynski]

I tried to explain that numbers have infinite names for the same identity (due to fractions behaving as you explained) in 7th grade honors math and the class laughed at me. Even the teacher treated me like I was crazy.

[@mokey345]

Similarly in my 7th grade honors class, my teacher convinced most of my class that “of” in word problems means “divided by” instead of “times”

[@chimpochimpay]

This is the most definitive discussion on this subject as far as I'm concerned. Bravo!

[@ilikemitchhedberg]

Are you surreal right meow?

[@ComboClass]

I often live in semi-surreality

[@cinnamoncat8950]

In the first half I was somewhat hating the video cause it just kept repeating the same "proofs" everyone would use to say 1=0.999... but right after halfway the explanation of epsilon and the infinitesimals helped me realize how to describe the fundamental disagreement I have with this argument. The fundamental thing I disagree on is that I think that 1/3 does NOT equal 0.333..and π does NOT equal 3.1415... I think they are APPROXIMATIONS for values that do not work in our system. They are approximations that functionally have no difference compared to the actual number in the real world but one that exists mathematically, which is why the epsilon now helps me know the difference. It is an infinitesimally small difference but just like actual infinity, it is something that can not be represented through numbers in this system. Thanks for coming to my Ted talk. Also I appreciate the video for exploring deeper than most :)

[@AaronALAI]

It's the same as 1 because you would need to add 0.000 repeating with a "1" at the end which is infinity small; for 0.999 repeating to equal 1.

[@wiggles7976]

How can a real number have a digit after an infinite amount of digits to the right past the decimal point? The number 0.123 has a 1 in the 10^-1 place, a 2 in the 10^-2 place, and a 3 in the 10^-3 place. In you number, 0.000...1, you say 1 is in the 10^n place. What is n?

[@marvinmallette6795]

@@wiggles7976 "n" is an unsolvable. Because you can't convert 1/3 into Base 10, you also cannot get the final component of 0.999... to reverse the operation. 0.333... is not a finite number from which to perform inerrant calculations upon. All subsequent calculations are based on an unfinished calculations and are therefore incorrect. By graphing the "limit" of 0.999... it makes it obvious in the abstract, but Aaron's statement is also an observation in the abstract. He understands the problem. 0.999... is incorrect, but the margin of error is infinitely small to the point of meaninglessness.

[@wiggles7976]

@@marvinmallette6795 It seems like you and Aaron accept that 0.999... is equal to 1. You do accept that 0.999... is exactly equal to 1?

[@AaronALAI]

@@wiggles7976 Yes, 0.999 repeating is exactly equal to 1, "0.999" by itself is not equal to 1 because you could add 0.001. If I add 0.0000 repeating with a 1 at the end to any number, the sum does not change because 0.0000...1 is infinitely small. I think ""n" is an unsolvable" is the correct response, but consider this to your original question, "you say 1 is in the 10^n place. What is n?" What if n were inf then it would be 10^-inf * 1 which is 0

[@wiggles7976]

@@AaronALAI OK, you are right about repeating decimals; 0.999... = 1. However, this idea of putting a 1 after infinitely many 0s does not make sense for real numbers. I don't know if some exotic number set could be defined using ordinals instead of integers for the powers of 10 that each get scaled by some digit from 0 to 9. In the real numbers however, integers are used for the exponents. When we have 10^n, n is an integer, not an ordinal or something else. Infinity is not an integer. Thus, it does not make sense to talk about the digit in the "infinitieths place" of a real number. What you are writing as "0.000...1" is just a haphazard way of describing something exactly equal to 0.

[@Elrog3]

Your presentation style is great. I prefer to include infinitesimals as numbers and though you may not, I'm glad you acknowledged that it is whatever we define it to be rather than stated that .999... = 1. You don't need to throw out the assumptions we are using for everyday math. You just need to treat the equals sign we are using as if it has an asterisk where things are not truly equal, but instead, are within a given range of each other. Then, you can continue on using our same symbolic rules we are used to and simultaneously not say it means .999... is literally 1. Do you want the world to switch to base 6? I think we should go for 12 or 30. We already learn multiplication tables up to 12x12 in school anyway, at least in the US. 10 is not great because 3 is a better prime number to evenly divide than 5 because it gives us more frequent terminating divisions. But with 6, numbers would start to require more digits to write out. I know memorizing isn't fun, but you only need to do it once. And even multiplications up to 30x30 are well within the amount of information people can retain. People that rarely do math may not know them that well, but for the people who do math a lot, it would be a benefit. I think we should make the tool specialized for the people who do the job. They do the job, so their needs should come first. We don't need to take specialized systems within every discipline of knowledge known to mankind and dumb them down for the lay person. And we don't need to here either.

[@Elrog3]

For the argument that goes: x=0.999... 10x=9.999... the .999... in the first expression shouldn't be considered the same as the .999... in the second expression. It will have 1 less 9 in it. People say often say some infinities are the same size, but that's a conflation between terms. What they mean is they are the same cardinality. I also reject the notion that linear bijection reflects what 'size' is.

[@caspermadlener4191]

I find it really refreshing how you acknowledge that it is not possible to give a satisfying proof of this. Axioms aren't as important as their direct consequences, those shaped the axioms in the first place.

[@mrosskne]

There's nothing to prove. A decimal expansion is just another way of writing the same number.

[@gilililili]

Me: wants to sleep The shark biologist from Jaws in a backyard of clocks slowly losing sanity: I don't think so

[@ilikemitchhedberg]

0.99999...st!

[@ComboClass]

I sometimes nickname it “zero point ninefinity” haha, but I didn’t say that nickname in this episode because I thought it might add confusion

[@ilikemitchhedberg]

@@ComboClass thank you for the lovely lecture!

[@tyruskarmesin5418]

0.99999st!

[@ilikemitchhedberg]

@@tyruskarmesin5418 yeah, that's what I should have said

[@danieldover3745]

I had already been convinced that 0.9999... was 1, but the explanation that helped me really understand what was really going on, and why my initial repulsion to it was also correct, was the concept of a limit. At no specific, definable point does 0.999... equal 1, it just approximates 1 and approaches the limit of 1 if the sequence is taken to infinity.

[@Chris-5318]

Your faulty reasoning is revealed by your, " At no specific, definable point does 0.999... equal 1" and your "approaches the limit of 1 if the sequence is taken to infinity". What you don't realise is that 0.999... is constant/unchanging/fixed/static and so cannot approach anything. It doesn't approach a limit, it's value IS a limit. You are confusing the series 0.999... (= 0.9 + 0.09 + 0.009 + ...) with the sequence 0.9, 0.99, 0.999, .... It's that sequence that approaches 1 as you step through. Here's the thing, it also approaches [the value of] 0.999.... The n th term of that sequence is 0.999...9 (n 9s), and that is easily seen to be 1 - 1/10^n. In fact, the value of 0.999... := lim n->oo 0.999...9 (n 9s) = lim n->oo 1 - 1/10^n = 1. The " := " means is equal by definition. The last equality follows from the definition of limit. I suggest that you look up "geometric series". The Wiki is especially relevant.

[@gumenski]

Since when was this controversial? We're stuck back in the 1700's again?

[@ComboClass]

Maybe you define controversial differently, but if you look at the comments of any video like this, you will see that people still have a wide variety of different opinions on this question

[@MuffinsAPlenty]

Yeah, I can understand both sides of this. The equality 0.999... = 1 is absolutely not controversial among experts in mathematics, but it is controversial among the general public, and any online discussion of 0.999... will reveal that. However, at the same time, I don't know how many people would be defending a video which claims that anthropogenic climate change is controversial, even if there is a sizeable portion of the general public which denies it, since there is no controversy among the experts. To be fair, the equality 0.999... = 1 won't have as direct of an impact on most people's lives as climate change will, but I do think the comparison gives me pause to completely agree with Domotro here.

[@gumenski]

I didn't know mathematics was opinion-based. Would you consider flat-earth vs the regular known globe earth model to also be a controversy since there are many unintelligent people rooting for us living under a dome that god made?@@ComboClass

[@ThePiiX]

Not gonna argue both sides. I was fermly on the "different" side at the start. because of my computer science mind (where you never test equalities for floating points numbers) But now I'm on the "Equal" side. And could defend it pretty well. Nice video. Still got chills over all the destroyed material though..

[@science_gang]

you are late on the topic! science gang has covered this months ago. yk, we love your content! well done ❤

[@asseroy]

It feels rlly nice being that early, I rlly rlly appreciate your content 💛

[@ComboClass]

And I appreciate you for appreciating/commenting! :)

[@PlebRoyale]

Amazing episode. Thank you Domotro. You may appreciate a poem I once wrote: "I have seen where the one mad God lives Far from here, yet, just a heirs breadth away I saw him whisper into his own ear 'The world is not made,' He said 'It is Mad' A bit of a weird fellow."

[@Bombito_]

Finally, a video that matches how people tend to see me when i tell them that there's a number between 3 and 4 which has no decimal representation because its a new kind of value when we consider that numbers are pixels, meaning there are numbers that are right at the sides of each one

[@Chris-5318]

Your comment is pure nonsense.

[@padraiggluck2980]

What is the numerical difference between these two numbers?

[@Neelo5000]

0.000...

[@userpd4ze3kt4u]

Is there a number system that orders all the numbers between zero and one and would it have any practical use? Zero point zero recurring one would be the smallest number after zero, so it would be called one. Zero point nine recurring would be the second biggest number after one so it would be called infinity minus one. Half could just be infinity over two.

[@oddlyspecificmath]

I've had terrible luck lately developing anything discussion-worthy, so I'm just going to kludge out the method I prefer. _There's no need to fully fill a division slot_ ... i.e., 4 / 2 is 2...sure, but that "fills" the slot. You can say instead that 4/2 is 1 remainder 2. Then 2/2 is 0.9 remainder 0.2, and so on...so you get 1.999999999.... with a remainder of 0.00000000...2 until you decide the "limit" has been reached and you fully "fill" the last division, then (under conversion from carryless arithmetic slots to based-digit slots) finally carry and end up with 2.000000.... as your answer. While I did develop this on my own, I found extensive references / wasn't first so suspect "delayed completion" isn't crazy.

[@andrewdenne6943]

So would it be accurate to say that 1=0.9• when = means accurate to within an infinitesimal of the exact answer but not when = means the exact answer but the reason we use the first is that in the second 1/3 doesn't equal .3• and any requiring decimals don't equal thier associated numbers so we just ignore the infintesimal difference as it makes other important math things work

[@martind2520]

You are incorrect. 0.333... is _exactly_ equal to 1/3 and 0.999... is _exactly equal to 1. Proof: The sequence 0.3, 0.33, 0.333, ... is always less than the sequence 1/3, 1/3, 1/3, ... The sequence 0.4, 0.34, 0.334, ... is always more the sequence 1/3, 1/3, 1/3, ... The limit of both decimal sequences is 0.333... The limit of the fraction sequence is 1/3. Therefore, by the sandwich theorem, the limits of all three sequences are equal. Hence 0.333... = 1/3 exactly.

[@andrewdenne6943]

@@martind2520 for the doing it sequencualy the difference between 1 and .9 is .1 the difference between 1 and .99 is .01 this sequence applys the whole way down so with 1 and .9(followed by n-1 9's) the difference is 1/10^n if n is infinite (as is the case for .9 requiring) we get 1/10^infinty which cause of how infinite works is 1/infinite which is a infinitesimal so 1=.9•+1/infinity if we divide all of it by 3 we get 1/3=.3•+1/3infinity which is also an infinitesimal I think sandwich thereory makes the same assumptions of accuracy to an infinitesimal off on

[@martind2520]

@@andrewdenne6943 So your take on this is that the sandwhich theorem, a well know and rigorously proven theorem in mathematics is "a bit off"? What exactly is your definition of 0.999...?

[@andrewdenne6943]

@@martind2520 sorry the bit addressing sandwich thereory is a bit misleading what I mean is that since the two points at the end of your use of the sandwich thereory don't quite met but have an infinitesimal width of space between them because if we take the upper bound sequence and minus your lower bound we get that same equation from before (.3-.4=.1, .33-.34=.01, .333-.334=.001 which produces the same equation of 1/10^n where n is the number of decimal places which when you have infinite decimal places like you would if you brought those bounds to their limits you'll have 1/10^infinity or an infinitesimal space between them this kind of thing isn't accounted for in the sandwich theorem normally cause they are operating with the method which ignores infinitesimals however if we are thinking about infinitesimal we have a slight gap which 1/3 goes through between.3• and .3•4 My definition of.999... or .9(repeating) is a zero a decimal point and then an infinite list of zeros

[@martind2520]

@@andrewdenne6943 Look at the way I used the sandwhich theorem. I did not falsely equate a sequence (or the "end" of a sequence) with its limit. I took sequences where the limit (not the sequence) were 0.333... and I took a different sequence where the limit was 1/3. I then used the sandwhich theorem (which is well proven and correct) to show that all the limits (not the sequences) were equal. Just because limits and sequences are different things, doesn't mean that the limits are somehow a more nebulous concept. Limits are still well defined specific numbers. The limit of two of the sequences was 0.333..., the limit of the other sequence was 1/3. Those limits are equal (which, remember, are still well defined and specific numbers), so 0.333... is equal to 1/3. Including infinitesimals will not change any of that, as none of my sequences has infinitesimals in, nor did any of the limits.

[@IsntPhoenix]

Also how would you write 1 - 0.¯9

[@smaza2]

welp you conviced me pretty quickly and to be honest your explanations are very intuitive. the thing that got it for me was limits. thank you as always for your phenomenal vids

[@beolach]

I'm a Pythagorean cultist. I don't believe in irrational numbers - not in any concrete real world way (irrationals only exist in a fictional albeit consistent & useful imaginary abstract world). But even in my philosophy 0.999... == 0.(9) == 1. It is rational & it is the unit one, from which all other real (read: rational (Pythagorean, remember)) numbers are defined. Excellent job explaining why here. This was something I struggled with when younger, going through a few different phases of my understanding of this difficult concept. I remember at one point feeling frustrated, and thinking it basically didn't really matter - at the time I was still wanting to trust my instinct & felt like 0.(9) != 1, but I was willing to (grudgingly - I was also in a pretty contrarian phase at the time) concede that the difference was immaterial. But it does actually matter, and I think you also did a good job explaining why it matters in this video. Ceci n'est pas une pipe, the map is not the territory, and symbols are not what they symbolize - but it is still very important to have a correct understanding of the symbolism; any flaws in understanding the fundamental symbolism can lead to flaws in the overall understanding. And it is perfectly acceptable to have multiple different symbols representing the same concept.

[@maynardtrendle820]

The squirrel running up and down " Yggdrasil" was awesome!😂🐿️

[@qwertyuuytrewq825]

Wow! Very funny and interesting. Great style )

[@АлёшаИнкогнитов]

3:27 How to become a believer in a god of math. "something off" and paper of infinity immediately drop off.

[@avibank]

Excellent camera work

[@ComboClass]

Shout out to my main camera guy Carlo! And a few other friends who helped me film some of the title cards, who are named in the credits/description :)

[@AlaaBanna]

Perfect video, thanks. Although I thought you'd be the perfect person to mention this, but here's how I look at it, The issue with those infinate 3s of (1/3) or infinate 9s .. etc, NOT a real one, but language related one (the base system we're using). For example, if we're using base-3 to calcualte 1/3 we'll simply have 0.1 (with a terminating 1, not repeating), And there in base-3: 0.1x3 (actually it's 10 in base-3, because only 0,1,2 are allowed) = 1.0 simple and clear. Of course, even in base-3 (or all other base systems as mentioned in the video), we'll find infinately repeating digits for other ratios (Let alone irrational numbers), but ALL rational numbers can have a terminating representation in some other system that fits them. So, finding the proper base for representing the scenario you'll find that 0.999... WILL actually actually mean 1.

[@Chris-5318]

Bases have nothing to do with languages. If b is a natural number, then 0.bbb... (base b+1) = 1. There is nothing special about base 10.

[@AlaaBanna]

@@Chris-5318 You are correct and that's what I'm saying. I just meant specifically for the 1=0.999 conflict could simply be resolved in the base-3 system as it will clearly be 1 and no-one would even think otherwise, that's why I mentioned it's a language (or the base) used. Similarly, ALL 0.bbb (related to rational numbers or ratios) can have a clear exact answer if another base, of course base-10 is not special here.

[@Chris-5318]

@@AlaaBanna I have no idea what your reply (or your original post) is supposed to be achieving. What conflict are you referring to, and what is resolved by using base 3? FWIW 0.222... (base 3) = 1. There is nothing special about base 3, or any base.

[@AlaaBanna]

@@Chris-5318OK let me put it in another way :), - Main conflict the OP (the video we're commenting) is: "Is 1 = 0.99999", right? - He offers many proofs or alternative points of views, to show us how it's normal we can find other representations of any number, and that 1, can be expressed as 0.9999.. - I did the same, offering another point of view, to indicate that if we looked at same numbers, from base-3 system, we'll find that conflict is resolved by itself, that because 1/3 will be represented as 0.1 in base-3 and not (0.333 in base-10), 2/3 will be represented as 0.2 base-3, and 3/3 will be 1 (without these 0.3333 * 2 = 0.66666 and *3 = 0.99999), that with base-3 for this problem specifically, we will not even have an issue. And that is not an special case with base-3 of course, nor something special with base-10 (agreeing with you here), but some bases are better for specific numbers to help us avoid inaccuracies of other bases. I hope my point is clear this time 🥲

[@Chris-5318]

​@@AlaaBanna You: "- Main conflict the OP (the video we're commenting) is: "Is 1 = 0.99999", right?" Me: First it's 0.999... = 1, not 0.99999 = 1, and second there is no conflict. What is it supposed to be conflicting with? I'll ignore the fact that you are not using the correct ... notation for now. You: "- He offers many proofs or alternative points of views, to show us how it's normal we can find other representations of any number, and that 1, can be expressed as 0.9999.." Me: So what? You: "- I did the same, offering another point of view, ..." Me: No you didn't. You: "... to indicate that if we looked at same numbers, from base-3 system, we'll find that conflict is resolved by itself, that because 1/3 will be represented as 0.1 in base-3 ..." Me: You resolved nothing and I have no idea what you think needs to be resolved. 0.1 (base 3) = 0.0222... (base 3) and 1 = 0.222... (base 3). You: "... and not (0.333 in base-10), 2/3 will be represented as 0.2 base-3, and 3/3 will be 1 (without these 0.3333 * 2 = 0.66666 and *3 = 0.99999), that with base-3 for this problem specifically, we will not even have an issue." Me: 1/3 = 0.333... (base 10) = 0.1 (base 3) = 0.0222... (base 3). 2/3 = 0.666... (base 10) = 0.2 (base 3) = 0.1222... (base 3). 1 = 0.999... (base 10) = 0.222... (base 3). You: "And that is not an special case with base-3 of course, nor something special with base-10 (agreeing with you here), but some bases are better for specific numbers to help us avoid inaccuracies of other bases. Me: The only inaccuracies I see are when you write, e.g., 0.99999 instead of 0.999.... 0.bbb... (base b+1) = 1 precisely in every natural number base - there is no inaccuracy. Every real number can be represented in every natural been with perfect precision. No base is more accurate than another. You: "I hope my point is clear this time" Me: I have no idea what this "point" is that you think that you are making.

[@arofhoof]

The 1/3 argument is only valid on base10? The paradox come from the lack of "resolution" of the base used. An "equal to" sign shouldn't be use but another symbol "approximate to"

[@Chris-5318]

It is valid in every natural number base. 0.333... = 1/3 exactly: it is not an approximation. Proof: 10 * 0.333... = 3.333... => 9 * 0.333... + 0.333... = 3 + 0.333... => 9 * 0.333... = 3 => 0.333... = 3/9 = 1/3

[@arofhoof]

@@Chris-5318 Base 12: one third is 1/4 -> paradox gone.

[@Chris-5318]

@@arofhoof 1/4 = 0.3 (base 12) = 0.2BBB... (base 12) -> "paradox" back. Also 1 (a whole number) = 0.BBB... (base 12)

[@arofhoof]

@@Chris-5318 0.3 is not equal to 0.2BBB..

[@Chris-5318]

@@arofhoof Of course 0.3 (base 12) = 0.2BBB... (base 12). Be clear, B is the eleventh digit (for base 12). Proof: working in base 12 throughout: 10 * 0.2BBB... = 2.BBB... = 2 + 0.BBB... 100 * 0.2BBB... = 2B.BBB... = 2B + 0.BBB... subtracting => (100 - 10)*0.2BBB... = 2B - 2 => B0*0.2BBB... = 29 => 0.2BBB... = 29/B0 = 0.3 (To be clear 29 (base 12) / B0 (base 12) = 33 (base 10) / 132 (base 10) = 1/4 and even you know that's 0.3 (base 12). If that is too hard for you to follow, working in base 10, unless stated otherwise), first note that 0.2BBB... (base 12) = 2/12 + 11/12^2 + 11/12^3 + 11/12^4 + ... Using the geometric series formula for he terms starting at 11/12^2 we have the sum is 1/6 + (11/12^2) / (1 - 1/12) = 1/6 + 11/132 = 1/6 + 1/12 = 3/12 = 1/4 = 0.3 (base 12). For an independent calculation, enter "2/12 + (11/12^2 + 11/12^3 + 11/12^4 + ...)" into Wolframalpha. It'll tell you that's 1/4. Also look up the geometric series Wiki.

[@jonathanschenck8154]

Plank length could = epsilon right? Right!?

[@Kyanzes]

Is nonary 0.88888... the same as decimal 0.99999... ?

[@Chris-5318]

Yes. More generally, if b is a natural number, then 0.bbb... (base b+1) = 1.

[@universallanguageproject]

Seems like the ship of Theseus paradox. If we lose a chunk of our skin through an injury, are we still ourselves? We would be more different for from the infinitesimal difference of .9 recurring and 1. I love your point about the numbers we create as being representational. Love your work 🙂

[@Chris-5318]

1 - 0.999... = 0. There is no difference, even if infinitesimals are allowed.

[@tomgooch1422]

Excellent solution!!! (The fire, I mean.) Dr. John Gustafson, of Gustafson's law, has written an excellent book, The End of Error, dealing superbly with this maddening ambiguity. It will end the era of heating computer rooms with wasted compute cycles seeking false precision when adopted.

[@martymoo]

After seeing use of the vinculum symbol instead of dots or ellipsis, I just wanted to point out, as Wikipedia says "At present, there is no single universally accepted notation or phrasing for repeating decimals." and "There are several notational conventions for representing repeating decimals. None of them are accepted universally."

[@ComboClass]

That's true. The vinculum is most common where I live so that's what I use to make it understandable to most people, but I have no specific dedication to it and I'm fine with any method that's understandable

[@PhredMacmurray]

Do you collaborate with Spielberg on these?

[@Tubluer]

That was a really good video, but how does it relate to Magic the Gathering?

[@percy9228]

loved the explanation and enthusiasm! ty

[@Faroshkas]

What about base 1? I was thinking a while back about that, and if a base 1 could exist, all of it's decimal representations would just be .000000000000... and not describe anything in particular.

[@chipichipichapachapaWHY]

Tally marks

[@SteinGauslaaStrindhaug]

28:23 two plus two could equal bleem if Professor Ersheim was right... ;P (See "The Secret Number" a short story by Igor Teper and a short film directed by Colin Levy, both are available free online. It's really absurd but really good)

[@maxwchase]

As far as avoiding glitches goes, I think p-adic numbers with representations that terminate on the right avoid them... as well as the ability to represent most irrational numbers, so.

[@Chris_5318]

Don't you regard ...999... = -1 as being a "glitch" in the same ways 0.999... = 1 is?

[@jaybingham3711]

Squirrel ain't up for being played as fool. He rolled out the sniff test. The bite test. And the lick test. "Noice. But ffs the same can't be said for that lab coat!"

[@zionfultz8495]

You can find an identical representation for any number by adding the 0.9 repeating representation of 1 and subtracting the normal single digit representation of 1.

[@Chris-5318]

That's only true for terminating decimals. You can't do it for 0.121212... but you can do it for 0.25 = 0.24999...

[@Point5_]

This video made me go through all 5 stages of grief

[@phyarth8082]

Simple computer software has built-in functions of approximation of numbers: Floor and Ceiling functions, rounding function and truncation function of numbers. Controversial matter when simple computer software has 4 dedicated functions to approximate number. Plus in physics we have capital Greek letter ∆ - difference of terms, in computer science we have δ value very small but finite it used to do numerical integration (summation loop) operation on digital computer. And in mathematics we have  (epsilon) infinite small value that is basis of calculus which nobody proven that that it exist or define value of this small value, Zeno paradox is space continuous and uniform. Exist computer software that can perform symbolical integration but it is done not numerically but using logic, Wolfram alpha can make symbolic integration and give answer in letters x, y ..., etc. But that is not real numerical calculation. Mathematics calculus and epsilon is still abstract value hanging on trust that calculus always match observable Nature.

[@coltenh581]

This video has given me much respect for the word “infinitesimal” Thank you Master Domo

[@Schody_lol]

0:20 wouldn’t that be 0.(9)?

[@mikumikudice]

personally, I don't consider this a glitch. it's a consequence of what math is. a representation of the nature; anything that represents something but isn't really that thing, will end up with multiple ways to represent it. many ways you can represent a car (a drawing, a word, etc). I think it's beautiful that one of the most complex things we ever invented is not a science, but a language that represents the really and events on it

[@landsgevaer]

On the Q that was posed, a system with a non-integer base would allow numbers to be written in more than two ways (unless you insist on allowing only a canonical form, but that also deals with decimal 0.999..). Like, in base phi, the number one equals 1.0000000.. or 0.1100000.. or 0.1011000.. or 0.1010101.. and many more. (I suspect *uncountably infinitely* many more even!)

[@HopUpOutDaBed]

I think it's still countable, just map them to the natural numbers following the pattern 1-> 1 2->0.11 3-> .1011 n -> n-1 representation but change the last 1 to an 011 of course you still have the infinite string of .01010101... forever but that's still just ONE number you can easily map by mapping 0 to it (or whatever trick you want to account for it.) Where are the uncountable representations coming from?

[@landsgevaer]

@@HopUpOutDaBed I was thinking about non-canonical forms that feature multiple pairs of 1s or series of more than two 1s in a row, but those indeed never seem to add to 1.000.. exactly. Have to give it some thought whether the number one is special in that sense, or even all integers are special. My reasoning was that when you are forming the digit expansion and you have a remainder that is just a tiny bit over the value of the next digit, then you can always choose to continue either ..100.. or ..011.. (which have the same value base phi). Another way to put it is that you can replace any ..100.. in an expansion with a ..011.. These latter forms are not canonical, but they are expressable (just like 0.999.. is not a canonical form for 1.000..). I surmised that for a typical 'random' number, such opportunities will occur arbitrarily often, i.e. there will be arbitrary many places where the expansion would have a 1 followed by two 0s, so you get a countably infinite number of places where you can choose between two representations, and 2^countably_inf equals uncountably_inf. On second thought, that might only hold for 'random' numbers (not exactly sure what the requirement is, but the numbers with arbitrary many ..100.. in base phi should be dense on the real line, I bet). So maybe I should rephrase it more narrowly that *with the exception of only a countably infinite number of reals* (including the integers as these exceptions), base phi allows uncountably infinitely many different expansions of a number. Or, alternatively, the numbers that can be written in uncountably infinitely many different ways in base phi are dense on R.

[@R.B.]

I feel like on the left of the decimal 9-bar.9-bar works as an equivalent of Infinity, for the same reason 0.9-bar is equal to 1. For the left side of the decimal, that is the only number which seems to have any meaning, but I'd argue it follows the same number system. The problem I see with this is you should define that as dividing by one less than the base, so 1-bar.1-bar + 8-bar.8-bar = 9-bar.9-bar, but defining 1-bar or 8-bar isn't definable in the same way which 0.9-bar is definable as 1/9 + 8/9... So it is more difficult to define. One of the other strange properties that this reveals is that 9-bar.9-bar is infinitesimaly smaller than Infinity just like 0.9-bar is infinitesimaly smaller than 1, which suggests that 9-bar.9-bar might be argued to be 2*Epsilon smaller than Infinity, but in the same way I think it actually reinforces the notion that 9-bar.9-bar is Infinity and 0.9-bar is 1, because Epsilon would be equal to 2*Epsilon for the same reason Infinity is equal to 2*Infinity.

[@martind2520]

No. Infinite 9s on both sides actually ends up being equal to 0, not infinity. (It is also a mathematical mess of a concept, but still ends up as 0.) And _no_ 0.999... is _not_ "infinitesimally smaller" than 1. They are _exactly_ equal. There are a multitude of proofs to this fact.

[@R.B.]

@@martind2520 you'll have to explain to me how an infinite number of nines on the left side of the decimal point is equal to 0. As for infinite nines to the right, the infinitesimally close to 1, but not quite reaching it was a point of view discussed in the video. I don't question that it is, myself, but I was trying to relate the concept to 0.9-bar as was discussed.

[@endo9902]

I like the presentation style.. kinda cool