Why do odds of flipping a coin or rolling a dice change with progression?

Non-Running Other

Moderation Moderation Information Moderation Information & Rules

Page 3 of 3

1 2 3

Next Last

8 years ago 02/09/2017 10:11am EST

re: Idiot

Idiot wrote:
rf wrote:
Bayesian statistics has just as rigorous mathematical foundations as those for frequentist stats (whether or not either is really math is a different question). Working in a non idealized setting updating odds after iterations using Bayes' rule or some other method is perfectly reasonable and most likely more accurate than insisting trials fit some idealized distribution.
But this scenario the poster laid with 100 straight crooked flips has nothing to do with statistics. It's an environment where someone controls the outcome and decides what that outcome is based on a static input. Its 3 card Monty. An outside observer could bet on what the coin flipper will manipulate the outcome to be based on if it's a long term game or a one time flip, but the person betting on the flip doesn't realize he has imperfect information and the flipper has both perfect information and total control. It's a behavioral question.

Someone gets it ... Bayesian priors

https://en.wikipedia.org/wiki/Checking_whether_a_coin_is_fair

8 years ago 02/09/2017 12:27pm EST

re: 73 and still doing distance

73 and still doing distance wrote:
ehhhh wrote:
Although some of the responses to this seem intelligent I don't think they answered your questions. I am also not sure if you are serious with the question or if you are trolling.
A binary outcome does not imply even likelihood of occurrence of the two possible outcomes. Just because at the moment of birth the child could be either a boy or a girl doesn't mean the likelihood is 50/50.
For example, if there is an implied birth rate of 52% female vs male, then at the moment of birth, you have a 52% chance of having a girl and a 48% chance of having a boy.
Ehhhh, what you say is accurate mathematically, but it's not accurate demographically, and it's not the answer to the question what about this? asked. I answered his question.
The poster wanted to understand why, if there are more females in the world, the probability of producing a girl shouldn't be higher than .5. I answered his question using demography.
My answer is accurate. When you're asking a question about relative abundance in a population of organisms with life spans and differential mortality, then you cannot formulate an answer without considering when in the population life span you're looking for your answer. Note that I specified "cohort," which is required in order to make the problem tractable in this context, where we aren't professional demographers.
Your answer involved an imagined set of probabilities at the time of birth, probabilities that don't match what we actually find in the world among human beings.

I didn't intend to imply you were incorrect or not answering ANY part of his question. You appeared to be responding more to his "There are more females in the world" part of his comment - I was referring to his initial line of roughly "There are 2 outcomes - boy or girl - so why are their more females than males?".

I was more approaching the subject that just because a scenario is binary in result does not imply equal probability of each result. I read his question as based on that premise - if I can have a boy or a girl at birth, how come there are more females. You answered the back half, that after birth, (I'm putting it in a very macro statement) average life spans can vary between sexes, therefore likelihood of male/female at birth varies from likelihood of male/female by age buckets. I am not disagreeing with this, you are spot on. I am just focusing on how I interpreted his question - based on the premise that male/female outcome at birth is 50/50. I just wanted to point out that just because there are two outcomes of an event does not imply 50/50 odds of each outcome (regardless of how the ratio shifts as time passes - perhaps similarly to the discussion of how a coin can wear down after so many flips, perhaps changing the likelihood of certain outcomes).

I apologize if they way I worded it came across as a shot at your response - was not intended to do so. I interpreted the poster's question as based on an incorrect premise that a binary outcome implies equal distribution of occurrence of each outcome, whereas it seems you interpreted it as "how can something occur at X rate but end up with Y rate?"

8 years ago 02/09/2017 2:17pm EST

re: Far Out West

Far Out West wrote:
What about birthdays? My friend and her sister have the same birthday and people always say "what are the odds of that"? I always respond 1/365. Am I right?

The odds of two specific people sharing a birthday is 1/365 (assuming there are periods of the year where having a child is more or less likely).

The odds of any two people in a group sharing a birthday is much lower. With a group of 23 the odds of two people having a birthday is 50/50.

8 years ago 02/09/2017 2:38pm EST

re: Far Out West

Far Out West wrote:
What about birthdays? My friend and her sister have the same birthday and people always say "what are the odds of that"? I always respond 1/365. Am I right?

A detailed explanation can be found here:

http://bfy.tw/9zxt

8 years ago 02/09/2017 2:52pm EST

re: easy weeks

8 years ago 02/10/2017 3:37pm EST

re: ehhhh

No need to apologize, but thanks. I wasn't trying to start a fight either.

Page 3 of 3

1 2 3

Next Last

What People Are Talking About On LetsRun

No top threads at the moment. Check back soon.

Reply Replying to

Username

Password

Leave the password field blank to post anonymously.

Post Preview

By posting you acknowledge that you have read and abide by our Terms and Conditions.

Remember me on this device.

Why do odds of flipping a coin or rolling a dice change with progression?

This thread has already been deleted.

You have been subscribed.