- #1

- 13

- 0

You are using an out of date browser. It may not display this or other websites correctly.

You should upgrade or use an alternative browser.

You should upgrade or use an alternative browser.

- Thread starter Tungamirai
- Start date

- #1

- 13

- 0

- #2

Gold Member

- 808

- 111

- #3

Science Advisor

- 4,816

- 134

You should look at probability theory to understand what constraints a space has to have for it to be a probability space.

If you flip a coin that has two outcomes (heads and tails) infinitely then then the population distribution will be represented by the relative frequencies of the coin toss and the parameter will be a function of said distribution. The infinite part is what matters here as it defines the population distribution and hence the population parameter.

It does not contain any conditional information however - just the zeroth order distribution for a coin toss stochastic process with a parameter p.

- #4

Mentor

- 4,714

- 3,673

Are you really asking about the effect of population size( number of coin flips)?

- #5

- 94

- 10

- #6

Science Advisor

Homework Helper

Gold Member

- 7,896

- 3,532

Is there a "simulation argument" that could help us understand specifically what your question is?

There is one issue that needs to be clarified: There is an important difference between the odds of a fair coin giving a particular exact sequence of heads and tails versus the believability, given particular sequence result, that the coin really was fair.

1) For a fair coin, the sequence of all heads is no less likely than any other exact sequence of heads and tails.

2) There are aspects of the all-heads sequence that are not at all like a random process. So as heads keeps coming up, it becomes certain that the coin is not fair.

3) Other exact sequences may prove that a coin is not fair. An infinitely long sequence of exactly alternating heads and tails is not from a fair coin.

4) On the other hand, there are randomly mixed heads/tails sequences that have exactly half heads that may be from a fair coin. As the coin tosses go on to infinity, the odds of exactly half being heads goes to zero, but 50/50 remains the most likely value. All heads (or all tails) are always the least likely.

- #7

Science Advisor

Gold Member

- 3,297

- 682

I didn't understand your question before. You are asking what the result of infinitely many flips of a coin will be?

If that is what you meant, then The Strong Law of Large Numbers says that the sample average will converge almost surely to the expectation. So if the coin is fair, an infinite sequence will have an average of 1/2 almost surely - with probability 1. Is that what you were asking?

If that is what you meant, then The Strong Law of Large Numbers says that the sample average will converge almost surely to the expectation. So if the coin is fair, an infinite sequence will have an average of 1/2 almost surely - with probability 1. Is that what you were asking?

Last edited:

Share:

- Replies
- 20

- Views
- 766

- Replies
- 41

- Views
- 2K

- Replies
- 29

- Views
- 1K

- Replies
- 4

- Views
- 570

- Replies
- 47

- Views
- 4K

- Replies
- 6

- Views
- 691

- Replies
- 13

- Views
- 150

- Replies
- 22

- Views
- 815

- Replies
- 5

- Views
- 651

- Replies
- 14

- Views
- 2K