Statnik 11: Birthdays + Interview with Dewinged

Published: March 06, 2018

stat_banner
Hello fellow once-a-year-birthday-havers and welcome to a post that will answer how old people on this website are (or at least those that willingly posted their ages on a sputnik list). So, I surveyed you all to give me more ideas for Statnik, and many of you suggested things I had already done. No worries, the statnik/macman76 cannon is long and it bends toward being sometimes overly detailed and boring and, thus, easy to forget. One thing many of you did not suggest was letting me analyze data you had already collected.

In stepped our hero, Dewinged. He stalked my request list, waiting for the right time to spring on me… that he had asked the sputnik userbase what their birthdays were. So, I scraped the data from his two lists and will definitively answer once and for all what the average age for those 129 users is. And what the median is. And the range. And the standard deviation.

(Data cleaning note: Dewinged seemed unsure of some of the birthdays but I used whatever was listed, I threw out one entry because it involved googling a date, and added the 15th for the day of another since it only listed month and year).

measure value
mean 24.75 years
median 23.59 years
range 33.86 years
sd 6.12 years

The average (sum total divided by count) user in that sample is ~24.75 years, median (value that, after ordering the values splits the data in half) is ~23.5 years, range (difference between highest and lowest age) is ~33.8 years, and standard deviation is ~6.1 years. But what does the distribution look like?age_hist

Age-wise, we are a spitting image of the world, probably.


The Birthday Problem

One fallacy that you probably don’t know you believe, that by reading this I will dispel from your mind before it festers and grows and forces you to dissolve your father’s business, is that the probability of a group of people having the same birthday (meaning month and day being the same, not year) is higher than you would believe intuitively. If 30 people are in a room and you had to guess the probability, many of you would think the answer is 30/356 because there are 30 people and 365 days. You would be wrong, and you would have known that was wrong if you had checked how your top was spinning and remembered what happened to Marion Cotillard.

Imagine a room with many people. If you ask 1 person their birthday, it’s the first possible birthday, and there is a 365/365 chance (excluding leap years) that it doesn’t match someone else’s since no one else’s birthday has been counted. The next person has a 364/365 chance of NOT having that same birthday or a 0.27% probability that there IS a match. The third person now has up to two possible birthday matches and the probability that any of the three DON’T match is 364/365 times 363/365 or a 0.82% probability that there IS a match. The fourth person makes it a 364/365 times 363/365 times 362/365 that NO ONE matches or a 1.63% chance that there IS a match. This goes on and on, and even at a measly 23 people, there’s just over a 50% probability of at least 1 pair of people having the same birthday. At 30 it is a 70.6% probability. At a 129, it’s effectively 100% probability. This is known as the birthday paradox. (For code to check it out for yourself, someone did an example in R, and I found this which is better explainer than I can give). If you’re still confused, no worries, I don’t fully understand it either.

Not only did we have 1 pair of users with the same birthday, we actually had 20 birthdays with at least one pair and 45 users total in one of those 20 birthdays (November 11 had 4!). Only two of the birthdays were June 9th (6/9), one was April 20th (4/20), and one was September 11th (9/11); so you guys were either giving primarily uninspiring fake dates for your birthdays or the ones you gave were mostly the real thing.

Is 20 unique birthdays with at least one pair a lot for 129 birthdays? Not really.

Histogram of the number of unique birthday pairs in 100000 simulations, vertical line is 20 pairs, the amount in the Sputnik birthday dataset

Histogram of the number of unique birthday pairs in 100000 simulations. Vertical line is 20 pairs, the amount in the Sputnik birthday dataset

After 100,000 simulations, the average amount of unique birthday pairs that would exist with 129 birthdays is ~18 pairs. Twenty birthdays is only in the ~70th percentile of the simulation. Even in all 100,000 simulations, at least one pair of matching birthdays existed. (I could probably have found the average amount of unique birthday pairs with calculus or math or something, but I don’t even know how I would go about doing that, so I simulated it instead.)

So, next time you are in a party and want to impress people, remember to ask everybody for their birthdays, don’t tell them why, then get the attention of all these possibly-drunk people, and explain to them with full-on fraction multiplication why it’s possible that one pair of them has the same birthday. Or get hammered. I don’t know.


Interview

As promised in my Stat Tank list, I granted Dewinged a mini-interview for suggesting I do this specific post. 

Alright, Dewinged, let’s start this interview. First off, no pressure, but any grammatical/spelling mistake you make will be italicizedbolded, and I’ll add a [sic] to them. I gotta [sic] keep this strictly professional [sic].

Haha no pressure, how is it going Mac. I have roughly 30 min. with constant interruptions from my coworkers. Then I have lunch break. Soon after I’ll shoot myself, and then I’ll be back to keep answering your questions. How’s the weather?

It’s colder than it usually is, but if people knew where I lived, they would probably tell me that I couldn’t complain. So, this interview is gonna appear below my post where I analyze the birthday data, but in real time of this interview I haven’t done that yet. So, we’ll play a game right now. What do you think will be the mean age (sum of ages divided by count of ages), median age (the age that, after sorting, splits the data in half), and the range of the ages (highest minus lowest age)? We’ll go Price is Right rules (closest without going over) and I’ll go first since you are my guest. I’ll say 24.5 years for the mean, 23.5 years for the median, and 27 years for the range.

You know, this brings me back to that young dewi frustrated in high school with his constant failure in anything related to numbers, so I am terrified, to say the least, about the mathematical vortex this interview could POTENTIALLY become. But I’m gonna differ dear Mac. My wild guess is that the mean is gonna be 22, the median 21 and (not countin the fact that KILL is 145 years old and Chortles 11) the range is gonna be… umm… 29.

It took citing the ages of KILL and Chortles to finally do it, but with your mathematical breakthrough, it looks like 145-11 is 29. Haha. BTW, I was ready to shut down this interview if you put 1 for an answer or added .1 to any of my answers (those people on the Price is Right know who they are and what they’ve done). No, I think we are through with the math portion of this interview. I was looking through your lists and skimming some of your reviews and I noticed how detailed some of them are. You also blend a lot of styles of lists (i.e. the sputnik birthdays is a user participation list, you have a weekly roundup of digs and recs, and a non-music-centric list on documentaries). One currency that users’ have on sputnik is the level of obsession/dedication they have to something, like a particular genre of music or statistical analyses (that’s me). Did that draw you to sputnik, or has that aspect made it a little bit more enjoyable?

I think everyone in Sputnik has a strength, and that is a beautiful thing. I know I can count on Relinquished, Pots, Lucid or Sniff to come up with amazing electronic music lists, for example, which is a genre I hardly know of. It wasn’t what drew me to the site but it certainly makes it enjoyable. I try to make my lists helpful and I love to have people participating sometimes. The weekly thing was only my thing when I started it in January but now I have recs from TwigTW, Papa Uni, brokencycle and Sniff going on every week and it’s great. I think users prefer lists where they can participate, like rec competitions and stuff like that, it’s basically the spirit of Sput.

I’ve never done a list/statnik on it because it is an ill defined prospect, but there are definitely users that are more influential than others. You’ve probably answered this elsewhere, but what does your username mean, and how did you find sputnik?

I make music from time to time and “Dewinged” was a tune I wrote a few years back. I can’t remember how it came up but it was something about getting your wings cut, in the sense of chasing your dreams, etc, (cheeseeeee…!) and sent back to the cruel reality. About the site, I stumbled upon Sput through Mongi’s review of Cult of Luna & Julie Christmas’ “Mariner”. I liked the light hearted and sometimes hilarious feel of the conversations of some other threads I checked after so I made an account right away. I also remember magicuba taught me not to write in my own shoutbox.

Had you ever joined a music forum type website before?

Nope, never. And I usually get bored of all things new rather quickly, but Sput has something that keeps you coming back to it, right? It might be the Brand New ranked lists.

Certainly. I never make ranked lists, but I might have to make an exception someday for Temple of the Dog or Heaven and Hell.
Let’s turn the turn-tables, as they say. Shoot me some questions.

Ok Mac, I have a couple of questions for ya. I always wanted to know, why your avatar is a football player with a moosehead? And unrelated, if I ever hit a casino in the near future, would your stat power come in handy so we could ransack the place in a legal manner? Maybe then we could lend mx some cash to fix the question marks bug.

So. I am a big sports fan (specifically american football, basketball, baseball, and I’ve started to get into association football). Deadspin, the sports blog, used to do photoshop contests not unlike what you see on reddit. Back in 2012, when I started on sputnik, they had a contest for the arrest picture of Rolando McClain, linebacker at the time for Oakland Raiders. It’s a funny picture, he was smiling pretty hysterically while being arrested. Meanwhile, the Philadelphia 76ers were having a contest to determine their next mascot. That smiling moose was one of the mascot options, which makes no sense for a team that’s iconography is styled after the American flag and other US founding iconography. Someone in that photoshop contest put the moose in place of Rolando McClain’s head and it was one the best things I had ever seen to that point. That was my original avatar… Skip forward to 2016, near my sputversary. I asked people what they were gonna give me, and SandwichBubble found or made, I’m not really sure, an image of that moose on Rolando McClain’s body now with him on the Dallas Cowboys.

Casinos are designed to make you lose over time. Odds and payout are set such that you would lose money if you gambled at most things you can find at a casino. Now, games of skill, like poker, where you can be smarter than other people about the value of your and possibly other people’s hands, are probably your only avenue for consistently making serious money as long as you put in the work to learn the game. Aside, NPR did a great episode about a group of Russian criminals that discovered that a particular model of digital slot machine had a detectable and deterministic order to the draws. They sent people to casinos with an app that, if they entered the sequence of results, would tell them when to bet high (because it was gonna pay out) and when to bet the minimum (because it was not gonna pay out). Anyway, the FBI caught some of the people. So, unless you wanna do some serious time, I gotta say we are pretty screwed regarding that bug (link to the episode btw).

Well, that basically sums up the fact that the Macman is far more interesting than myself haha. I have one more question if I may. What would be the Sputnik statistic calculation that would make mac nope out of it? Either because of its ill nature, or because the amount of data would bury you dead.

I don’t know, probably going to graduate school, getting a more demanding job, or starting a family would probably get me to decrease my time on sputnik. It’s little known, but I was once just a pleb user, nothing special about me, and I then I just stopped posting on the site for like 3 years, only briefly stopping by to discover some new music. I came back after that hiatus, so I probably would do the same even if I left for some time.

To wrap this up, there is an interview podcast that I listen to that ends every interview with the same question; so I’ll steal it just like I do everything else: what was the last great thing you listened to/saw/or read?

The new album from Anna Von Hausswolff has been released today. I am only two songs in as we speak and I am effectvely blown away so i would highly recommend everyone to get on that one. There is also a new Spanish TV show on Netflix called Paper House that is really good too. Maybe that’s what brought up the casino idea in my head, since it’s about a heist. What about you, mac?

Yesterday, I rewatched Steven Soderberg’s Traffic which was very good. It was very much of its time, a multi-storyline movie (a la Pulp Fiction, Magnolia, Amores Perros, for example) that’s real gritty. There are some shots in that movie that are really grainy, bright, and suffocatingly hot. I also rewatched the Untouchables which was mostly terrible except for the scene where Sean Connery’s character was killed. There’s really great bit of misdirection to that scene. As for music, I’m not a music nerd like everyone else so I’m gonna give a boring answer and just say that the song “Fear and Trembling” by ETID is pretty heavy in my gym songs rotation right now.

Alright, pleasure. My [sic]’s are gonna get a real good workout. Also, my birthday predictions are gonna win me a damn boat with… a brand new jet ski. Your showcase is, I’ll say, a new car but it’s like a Hyundai Elantra, so have fun paying taxes on that prize if you win it.

It’s been a pleasure Mac, looking forward to reading those stats. And forgive my [sic]s, please! I’m non native speaker after all! 😞


So, “actual retail price” was a mean of 24.75 years, median of 23.59 years, and a range of 33.86 years (all as of March 5th, 2018 which is when I did these calculations). I was almost on the freakin’ screws with guesses of 24.5, 23.5, and (meh) 27 while Dewinged guessed 22, 21, and 29, respectively. I think this makes this a double showcase which is a damn shame because I did not want that Elantra.

Pop / Top 40 / General
follow us on Twitter      Contact      Privacy Policy      Terms of Service
Copyright © BANDMINE // All Right Reserved
Return to top