Talk:Poisson distribution

This edit request has been answered. Set the |answered= parameter to no to reactivate your request.

The "Weibull and stable count" subsection is entirely original research and should be removed. It is currently uncited. The only references I could find to the concept of the "stable count distribution" outside Wikipedia were a series of SSRN preprints by the independent researcher Stephen H.T. Lihn (not peer reviewed, and as far I can tell never subsequently published or cited except as self-citations in other preprints). The username associated with these edits suggests this is the same author. Other edits by the same user either introduced similar material uncited on other pages, or included it with citations to the SSRN preprints. Even if this had been published work, there are thousands of articles that generalise the Poisson distribution in some way, and clearly not all belong in a high-level overview article about the Poisson distribution! But this one is clearly in breach of WP:NOR. It may be worth someone taking time to check whether any other original research has been introduced into the article in the past. Petam576 (talk) 12:16, 1 October 2025 (UTC)

@Petam576 Thanks for noticing this. I will remove the section and as for the article "stable count distribution" to be deleted. Best, Malparti (talk) 12:36, 1 October 2025 (UTC)

Thank you, I have tried to follow the AFD instructions and there is now Wikipedia:Articles for deletion/Stable count distribution! Petam576 (talk) 13:04, 1 October 2025 (UTC)

Already done AlphaBetaGamma (Talk/report any mistakes here) 13:45, 1 October 2025 (UTC)

This edit request has been answered. Set the |answered= parameter to no to reactivate your request.

Change the stated mode to the floor of lambda, instead of the ceiling of lambda take away one then a comma with the floor of lambda. This is because the median is given in terms of the floor, so this would be standardisation across this page. BenBq (talk) 16:35, 18 January 2026 (UTC)

 Not done: it's not clear what changes you want made. Please detail the specific changes in a "change X to Y" format and provide a reliable source if appropriate. GearsDatapack (talk) 19:58, 18 January 2026 (UTC)

Hi @BenBq,

The reason the mode is stated as ${\displaystyle \lceil \lambda \rceil -1,\lfloor \lambda \rfloor }$ is that when ${\displaystyle \lambda }$ is an integer, the values ${\displaystyle \lceil \lambda \rceil -1}$ and ${\displaystyle \lfloor \lambda \rfloor }$ differ and are equally likely under the Poisson distribution with parameter ${\displaystyle \lambda }$. Thus, any of these two values can be called "a mode".

I agree that this is a rather cryptic way to state things; but at least it is "technically correct"; whereas the change you proposed would yield something "not quite correct when ${\displaystyle \lambda }$ is an integer.

Cheers, Malparti (talk) 20:21, 18 January 2026 (UTC)

Hello!

The image of chewing gum on a sidewalk being used as an example of an approximately Poisson distributed process is not actually Poisson distributed. I saw the image and was skeptical (and also so amused that someone saw it and their first thought was to calculate the statistical distribution), so I plotted a histogram of the counts (number of pieces of gum per tile) and fit with a Poisson distribution as a qualitative cross-check and found that it is indeed not Poisson distributed. I am happy to share the plots and code, though I'm sure anyone can do it themselves--totally fair that nobody thought to check it before!

I suggest removing the image as it is an incorrect example. Happy to help find an example to replace it! SchnorzchildMetric (talk) 04:03, 10 March 2026 (UTC)