Welcome to the mathematics section
of the Wikipedia reference desk.

Select a section:

WP:RD/MA

Want a faster answer?

Main page: Help searching Wikipedia

How can I get my question answered?

Select the section of the desk that best fits the general topic of your question (see the navigation column to the right).
Post your question to only one section, providing a short header that gives the topic of your question.
Type '~~~~' (that is, four tilde characters) at the end – this signs and dates your contribution so we know who wrote what and when.
Don't post personal contact information – it will be removed. Any answers will be provided here.
Please be as specific as possible, and include all relevant context – the usefulness of answers may depend on the context.
Note:
- We don't answer (and may remove) questions that require medical diagnosis or legal advice.
- We don't answer requests for opinions, predictions or debate.
- We don't do your homework for you, though we'll help you past the stuck point.
- We don't conduct original research or provide a free source of ideas, but we'll help you find information you need.

Ready? Ask a new question!

How do I answer a question?

Main page: Wikipedia:Reference desk/Guidelines

The best answers address the question directly, and back up facts with wikilinks and links to sources. Do not edit others' comments and do not give any medical or legal advice.

June 29

Confirmation of probability calculation

Am I correct in thinking that the probability of no overlap between four random selections of 5 items from 63 is (58C5)(53C5)(48C5)/[(63C5)^3]? 2A00:23C6:AA08:E500:98F5:73C0:9C7E:A44F (talk) 14:13, 29 June 2020 (UTC)[reply]

Experimentally confirmed. On a run of one million trials, 648260 had no overlap, about ~~0.15σ~~ 0.05σ off from the expected value from the formula, 64837.9. The obvious generalization also appears to hold experimentally. --Lambiam 15:07, 29 June 2020 (UTC)[reply]

That generalization is given by the formula

\textstyle \prod _{k=0}^{s{-}1}{\binom {n{-}ki}{i}}\left/{\binom {n}{i}}^{s}\right.

, in which

s

stands for the number of selections,

i

for the number of items in each selection, and

n

for the number of elements to choose from. A moment of combinatorial reflection has convinced me that this is correct. --Lambiam 15:30, 29 June 2020 (UTC)[reply]

Thanks. This arose from four ex-colleagues each choosing 5 favourites from a set of 63, and my wondering how likely at least one common choice would be. The first approach was to treat it as a "birthday" problem (63 possible days and 20 persons), then I realised the error of that, in that no coincidence could arise in each quintet. A simulation followed, giving the same 6.48% as you report, then finally I convinced myself that the formula I asked about was right. I think you did ten million trials, though.→2A00:23C6:AA08:E500:ED00:AF95:D6C2:CE64 (talk) 16:47, 29 June 2020 (UTC)[reply]

No, just one million; I made a mistake in going from the fraction printed out to the numerator of that fraction. --Lambiam 22:07, 29 June 2020 (UTC)[reply]

Another way of writing the result is

{\frac {(n-i)!^{s}}{n!^{s-1}(n-si)!}}.

More generally if there are m_i items in the ith selection, then the probability of no overlap is

{\frac {\prod (n-m_{i})!}{n!^{s-1}(n-\sum m_{i})!}}.

--RDBury (talk) 19:18, 30 June 2020 (UTC)[reply]

July 5

Maximum likelihood

Maximum likelihood are not sufficient statistics . Is true?