Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Birthday Paradox Experiment (2018)

pudding.cool

Post reply on HN
Print view View original

61–70 of 74 posts

Re: The Birthday Paradox Experiment (2018)

#61

Post by kqr »

bambax wrote: post #17
bsimpson wrote: post #3

It's interesting how uniform the birthday distribution at the end looks. I'd expect more seasonality (e.g. more babies conceived in the cold, dark months).

Yes, it's also extremely unlikely. The actual distribution in developed countries is not uniform: there is a spike at the end of September (because many more people make babies at or around New Year's Eve) and a considerable drop on Dec. 25th (because people will avoid that date and provoque the birth some days before in case it might happen). Also, on the site there is a huge spike on Nov. 15 which, incidentally, is…

> The actual distribution in developed countries is not uniform

This is a nice story I've heard many times, but is it actually true? Like what are your statistical sources here?

Re: The Birthday Paradox Experiment (2018)

#62

Post by GuB-42 »

tosser0001 wrote: post #24

An analog of the birthday paradox that gets me all the time is what I think of as The Locker Room Paradox. This is where when I go into the locker room after working out and the guy who comes in behind me ends in the locker right next to mine. So there’s two of us in a big empty room awkwardly jostling away.

I think it is closer to the reason why it is surprisingly difficult to throw a rock through a wire fence even when the rock is much smaller than the holes in the fence. We tend to underestimate the area of interaction between the rock and the fence.

If you take a locker in the middle, there will be 8 lockers right next to yours, which may represent a sizable fraction of the total number. Combine that people are not random and that they tend to forget about the times where it doesn't happen and it may seem like it happens all the time even when it is uncommon on average.

Re: The Birthday Paradox Experiment (2018)

#63

Post by Balgair »

lukol wrote: post #45

While the math is clear, I'm a bit annoyed by the label "paradox" as the whole setup is too simplistic and reductionistic. The actual chance of being in the same room with someone who shares your birthday needs to include other factors like your socioeconomic background, the cultural environment you are in, your present location, and certain historical facts. Without having done the math, I'm fairly certain that a me…

Wait, how is the math clear? There was no math presented on the site, I didn't see a single formula

Re: The Birthday Paradox Experiment (2018)

#64

Post by Balgair »

Okay, that was about as clear as mud to me. Maybe I just need more coffee, but nothing about it led me to understand why there is a 50% chance with 23 people. Can someone else explain this?

The odds that someone else shares just your birthday out of 23 people sounds crazy still. It should be 182.5 to get to 50%, right?

Re: The Birthday Paradox Experiment (2018)

#65

Post by lukol »

seritools wrote: post #55
lukol wrote: post #45

While the math is clear, I'm a bit annoyed by the label "paradox" as the whole setup is too simplistic and reductionistic. The actual chance of being in the same room with someone who shares your birthday needs to include other factors like your socioeconomic background, the cultural environment you are in, your present location, and certain historical facts. Without having done the math, I'm fairly certain that a me…

> While the math is clear, I'm a bit annoyed by the label "paradox" as the whole setup is too simplistic and reductionistic. https://en.wikipedia.org/wiki/Paradox#Veridical_paradox (also https://www.youtube.com/watch?v=ppX7Qjbe6BM for a 40min video discussing the weird usages of the word "paradox")

Thanks, I was not aware of this classification. I primarily tend to think of paradoxes as "self-contradictory statements" but you just expanded my definition.

Re: The Birthday Paradox Experiment (2018)

#66

Post by lukol »

Balgair wrote: post #63
lukol wrote: post #45

While the math is clear, I'm a bit annoyed by the label "paradox" as the whole setup is too simplistic and reductionistic. The actual chance of being in the same room with someone who shares your birthday needs to include other factors like your socioeconomic background, the cultural environment you are in, your present location, and certain historical facts. Without having done the math, I'm fairly certain that a me…

Wait, how is the math clear? There was no math presented on the site, I didn't see a single formula

Took a brief look at the Wikipedia article: https://en.wikipedia.org/wiki/Birthday_problem

Re: The Birthday Paradox Experiment (2018)

#67

Post by xydone »

Eddy_Viscosity2 wrote: post #49
xydone wrote: post #44

Earlier quoted context omitted.

Ohh, I see! At my gym, locker keys are given to you by the front desk and you put in something as deposit (such as your gym card or whatever you wish) and on your way out you give the key and you get your deposit back.

But why? Seems like it would be inconvenient to gym-users to add these extra steps to getting in and out of the gym. Especially if you have to wait behind other people just to get or return a key. What is the benefit of such a system?

It's a relatively small gym so I assume they don't have the resources to improve the system and quite frankly I too would prefer if they spent the money on more equipment

Re: The Birthday Paradox Experiment (2018)

#68

Post by lukol »

wodenokoto wrote: post #46
lukol wrote: post #45

While the math is clear, I'm a bit annoyed by the label "paradox" as the whole setup is too simplistic and reductionistic. The actual chance of being in the same room with someone who shares your birthday needs to include other factors like your socioeconomic background, the cultural environment you are in, your present location, and certain historical facts. Without having done the math, I'm fairly certain that a me…

Wait, what? Are you saying that income influences what time of the year you are born? The paradox doesn’t talk about meeting your birthday sibling, but meeting two people who are birthday siblings. So by the time the 12 year old has met ~20 people, there is a 50/50 chance that amongst those 20, there’s a birthday pair.

> So by the time the 12 year old has met ~20 people, there is a 50/50 chance that amongst those 20, there’s a birthday pair.

If you randomly choose the ~20 people from the global population then yes, this will be the case (especially after a number of rounds). And yes, I'm aware this is also the definition of the paradox.

But if you choose the people from your vicinity (i.e the people you are actually likely to meet), the chances will vary based on your individual parameters (which defines the number and quality of the sample size).

Re: The Birthday Paradox Experiment (2018)

#69

Post by aroberge »

Balgair wrote: post #64

Okay, that was about as clear as mud to me. Maybe I just need more coffee, but nothing about it led me to understand why there is a 50% chance with 23 people. Can someone else explain this? The odds that someone else shares just your birthday out of 23 people sounds crazy still. It should be 182.5 to get to 50%, right?

It is not that someone else shares YOUR birthday: it is that two people (among those 23) will have the same birthday.

Re: The Birthday Paradox Experiment (2018)

#70

Post by dustincoates »

xydone wrote: post #44
dustincoates wrote: post #40

Earlier quoted context omitted.

I've never been to a gym where you're assigned a locker for the day (or given a key). Either you have one permanently assigned (rare) or you go in and find one that isn't occupied.

Ohh, I see! At my gym, locker keys are given to you by the front desk and you put in something as deposit (such as your gym card or whatever you wish) and on your way out you give the key and you get your deposit back.

That's interesting. At every gym I've been to, you either bring your own lock or they have locks where you can set a temporary code.
Post reply on HN