Talk:Erdős number

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
WikiProject Mathematics (Rated B-class, Mid-priority)
WikiProject Mathematics
This article is within the scope of WikiProject Mathematics, a collaborative effort to improve the coverage of Mathematics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Mathematics rating:
B Class
Mid Priority
 Field:  General
One of the 500 most frequently viewed mathematics articles.

Very old assorted topics[edit]

Removed cfdnotice, cfd has completed. --Kbdank71 16:08, 9 May 2008 (UTC)

I have heard of people who published multiple papers with Erdos having fractional Erdos numbers. For instance if you published 5 papers directly with Erdos you have a number of 1/5. Has anyone else heard of this usage? —Preceding unsigned comment added by (talk) 02:55, 1 September 2007 (UTC) Ihave a erdos number through mckay! Why is there a link to "Umlaut"? In fact the 4th character in Erdos' name is not the ö (o-umlaut, ö) mentioned on the Umlaut page.

ö has two dots above the o, while Erdos has a "long Umlaut", which looks more or less like two acute accents close to each other, but the Umlaut page does not mention it. Aleph4 12:17, 8 Dec 2003 (UTC)

There is an Erdos Number of 5 on sale at ebay: . This is too ephemeral to put into the article yet, I think, but maybe once the auction is over it might be worth a mention. --Zero 23:04, 21 Apr 2004 (UTC)

The link is now dead and should probably be removed from the article

Since the location of the Paul Erdos article no longer has an umlaut (see Talk:Paul Erdos for discussion), I think that for consistency this article should live at Erdos number. --Saforrest 23:03, Apr 14, 2005 (UTC)

Urhixidur 01:54, 2005 Apr 15 (UTC)

Not just mathematicians[edit]

There seems to be some disagreement about the scope and applicability of Erdos numbers. I removed the explicit references to "mathematicians" in the definition, then someone restored it, and now we're back to the more generic "authors". I'd just like to point out that many physicists and computer scientists have Erdos numbers. For example, our own article on the physicist Brian Greene mentions the fact that he has both an Erdos number and a Bacon number. I would imagine that academics in many other fields have Erdos numbers as well. In fact, the whole point of the concept is to illustrate the small world phenomenon, so artificially restricting it to the even smaller world of mathematics defeats the purpose. --MarkSweep 19:22, 23 May 2005 (UTC)

I noticed the following sentence in the article: A small number of people are connected to both Erdős and Bacon and thus have a finite Erdős-Bacon number. Wouldn't that small number be either 0 or else everybody in the union of both graphs? (which would in fact then be the same graph, with a different specified root)--Ramsey2006 16:01, 20 January 2007 (UTC)
Never mind...different criteria for the edges. --Ramsey2006 16:03, 20 January 2007 (UTC)
The concept is not restricted to mathematicians, but it is restricted to "mathematical papers", as the definition says. So Brian Greene may well have an Erdos number, if he coauthored a mathematical paper. As to it being "arbitrary", of course it is, all definitions are arbitrary. Erdos number could have been defined to include say any kind of published material, not just mathematics, but it wasn't. The article merely reflects that original choice. In my experience, the definition given in the article correctly describes how the term is currently understood and used. If you think that the article incorrectly defines Erdos number, I would have to see some reliable source for this notion, before I could agree with it. Paul August 19:58, May 23, 2005 (UTC) "There is an edge between vertices u and v if u and v have published at least one mathematics article together. (There is no reason to restrict this to the field of mathematics, of course.)" This is further clarified: "Our criterion for inclusion of an edge between vertices u and v is some research collaboration between them resulting in a published work. Any number of additional co-authors is permitted. " Also see: The definition used by the Erdos Number Project, although not "official" in an academic or bureaucratic sense, is well-established and does seem to include physicists and biologists. A note on the end indicates that a peace manifesto coauthored by Einstein, and reproduced in the New York Times, assigned Einstein a 2 and the other to-be-Nobelist collaborators a 3.

Where does one draw the line, though? Ian Stewart, co-author of The Science of Discworld, has am Erdos number of 3. If it's any published collaboration, does this mean Terry Pratchett has an Erdos number of 4? Daibhid C (talk) 11:35, 9 October 2008 (UTC)

There seem to be some reasons for limiting the connections to maths papers, but there don't seem to be any sources or citations for some authority using those reasons to reach that conclusion. I.e., there is nothing verifiable saying an Erdos number is based only on maths articles. I don't think we here on a wikipedia talk page get to decide what the definition is. (talk) 19:46, 17 July 2009 (UTC)

Wikipedia has not decided whether Erdos numbers must involve math papers. The article on Erdős numbers presently says "as measured by authorship of mathematical papers." (Not all other parts of the article are consistent with that requirement.) The article on Erdős–Bacon numbers explicitly allows non-mathematical papers, such as neurology papers. I always thought only math papers were intended to count, and I don't think it makes sense in terms of the original idea of connection to the math community to count non-math papers. (This does not mean the author has to be a mathematician. The paper should be mathematical.) Do we want consistency? And by what definition? Zaslav (talk) 01:00, 25 August 2016 (UTC)

Do Wikipedia articles count?[edit]

Does any Wikipedia article count as a mathematical paper? --Army1987 22:01, 11 September 2005 (UTC)

Definitely not. Wikipedia is encyclopedic; it contains only things that have already been thought.
—Preceding unsigned comment added by (talkcontribs)
While I think I agree with you re the conclusion, your arguments are not adequate. A maths paper may be encyclopedic, and it may be a survey paper containing only things that have already been thought. Indeed, were Erdos to have edited a wikipedia article with someone as a form of maths scholarship (maybe re the history of Erdos numbers), I think you'd have to look to practice in the scholarly field to determine whether the wikipedia article counted for erdos number purposes. (talk) 22:12, 17 July 2009 (UTC)
Rats --h2g2bob 21:04, 6 November 2006 (UTC)
Ok, this makes me wonder: which Wikipedia articles has Erdős edited? 23:27, 17 January 2007 (UTC)
He hasn't, so far as I know. But McKay has edited this talk page, so I guess that my adding this edit bumps me up from 3 to 2 ;-p --Ramsey2006 08:02, 20 January 2007 (UTC)
thanks for the idea,i now have a Erdős number of two.Ha! Acadamia — Preceding unsigned comment added by (talk) 00:48, 30 August 2011 (UTC)

Wikipedia articles don't have individual authors who get credited for their publication, per WP:OWN, so they don't lead to co-authorship. Sorry. —David Eppstein (talk) 03:20, 30 August 2011 (UTC)

Not exactly an Erdős number...[edit]

Although I don't have an actual Erdős number, and probably never will, I have studied small amounts of computer science and mathematics under someone who took a combinatorics course (in computer science) under Ralph G. Stanton (who has Erdős number 2). CanadaGirl 12:07, 7 April 2006 (UTC)

Courses don't count, I'm afraid. If they did, I'd have an Erdős number by virtue of this guy. --Saforrest

Average of 5?[edit]

From the article: "...the average is less than 5, and almost everyone with a finite Erdős number has a number less than 8." Of course, the average is not less than 5, since many people have an infinite Erdős number. Among those with finite, what is the source for the average being less than 5? Perhaps it should be median? I am inclined to delete the sentence if there is not a source. (Cj67 21:11, 25 June 2006 (UTC))

CJ, some people mean "finite Erdos Number" when they say "Erdos Number". So the line can be made unambiguous by writing "...the average finite Erdos Number is less than 5...". Since I myself wouuld prefer that high-school students not consider infinity as a number (but as part of a process), I prefer to say "Spencer's Erdos Number is 1 and Pete does not have an Erdos Number" instead of "...Pete's EN is infinity". Some things called infinity, such as Aleph-Null and Omega, can indeed be considered as numbers in certain number fields, but that's an advanced topic and people studying those topics don't need to be warned about the ambiguity. Pete St.John (talk) 17:46, 17 December 2007 (UTC)
  • The Erdos number project analysed the data from Math Reviews. I added a few words to the article. --Aleph4 22:14, 25 June 2006 (UTC)

Gauss-Minkowski - just ain't gonna work[edit]

I'm sorry, but this just doesn't cut it:

However, according to MathSciNet Carl Friedrich Gauss (born 1777) has Erdős numbers 4 as follows: Gauss – Hermann Minkowski – Albert Einstein – Ernst Gabor Straus – Erdős. The connection between Gauss and Minkowski is a collection of essays containing separate works of both authors.

Gauss died in 1855. Minkowski was born in 1864. So, there is no way they could in any real sense of "co-author" co-author anything. Having Gauss & Minkowski's works collected in the same volume does not count in any way as a collaboration between the two, but merely a decision of later editors..... So I will delete this bit. --SJK 11:14, 20 July 2006 (UTC)

Agreed. McKay 13:08, 21 July 2006 (UTC)

Infinity is not a number. If a person doesn't have a chain of coauthors linking them to Erdős, you can say their Erdős number is infinite or say that it's undefined. Erdős number and finite Erdős number are synonyms. --Awis 06:15, 8 August 2006 (UTC)

I've just realized that my Erdös number is 3. I have published with Gilles Brassard. Hugo Dufort 03:43, 25 October 2006 (UTC)

Concise and meaningful language[edit]

This article keeps using the term 'finite' to describe Erdős numbers. I think this is a bit ridiculous, even more so to say that if a person doesn't have a chain of co-authors linking them to Erdős that their Erdős number is infinite! What is the point in such language when it doesn't serve to convey a concise meaning? Far better to say in more concise and explanatory fashion that they either have an Erdős number or they don't.

The article also talks at one stage about the earliest person to have a *positive* finite Erdős number. This is even more ridiculous since by definition there is no such thing as a negative Erdős number. I have to wonder about the motivation to embellish the language with such pointless words.


I agree with you about the "positive". As for "infinite", the problem is that more than one definition of "Erdős number" is going around and both definitions are mixed up in the article. In one version, some people have an Erdős number and some don't. In the other version, everyone has an Erdős number but some of them are "infinite" (a phrase never defined precisely, but the whole thing is informal). My recollection of the way combinatorialists talked about it when Erdős was still alive is that the second version was more popular. It is also the definition used in the quasi-official Erdős number project [1]. So I propose writing the whole article in that fashion, with a passing mention that some people prefer "not existing" over "infinite". McKay (Erdős number 1) 03:35, 13 November 2006 (UTC)
Here "positive" must mean "non-zero". The earliest person to have a non-negative Erdos number is, by definition, Paul Erdos. —Preceding unsigned comment added by (talk) 11:32, 17 February 2008 (UTC)
I argue that infinite Erdős numbers are almost mandatory. Suppose, arguendo, that instead we took the "you either have one or not" position. Then how could we tidily express the conditions under which people would have them? "You have an Erdős number if... umm... you are he, or a co-author, or a co-author of a co-author... umm... well, you have one provided... umm, provided you're one of the people who has one." Yes, of course one could define it by recourse to graph theoretic jargon (e.g., "You have one if you and Erdős are in the same component of that graph on the human species in which adjacency corresponds to having co-authored in the scientific literature") or to recursion (e.g., "Erdős does, as does every non-Erdős who's a co-author with anybody who does"), but yuck and yuck. How much simpler to say "Every human has an Erdős number." Now that is concise. And meaningful.—PaulTanenbaum (talk) 03:46, 19 June 2009 (UTC)


Would someone who is familiar with IPA please add the pronunciation according to it? Most people aren't familiar with Hungarian and will mispronounce the name. -Emiellaiendiay 07:42, 17 November 2006 (UTC)

Could someone please add a stress mark to the IPA? Is it ˈɛrdøːʃ or ɛrˈdøːʃ? Stevvers 21:27, 8 April 2007 (UTC)
Still missing after 7 months. Surely someone must know. And is there an anglicised equivalent (UK ˈɛədɜːʃ/US ˈɛrdɜːʃ, perhaps)? — 15:35, 11 November 2007 (UTC)
I corrected it. In Hungarian the stress is always on the first syllable.A Hungarian -- (talk) 12:38, 26 February 2009 (UTC)


can someone explain why erdos was chosen, and why it is considered so important? I'm struggling to see how having a high Erdos number is in any way "good" or worthy of any note whatsoever. Same with the Kevin Bacon thing. May I suggest a football one for Liam Brady?

). Got a 61 in his course though (not due to lack of effort). -

  • You should read about Paul Erdos; a fascinating man. He was highly idiosyncratic, very much respected as a genius in the field; and he travelled widely, meeting huge numbers of fellow mathematcians. That's why he was chosen.
  • A high Erdos Number isn't good. Low numbers are prestigious; 1 means you coauthored with the Great Man himself, 2 means you coauthored with someone 1, and so on. The lower your number, the closer you are (in this sense) to a group of very productive and regarded mathematicians. Of course you might also be close to them in other ways, and not have any number at all. Having a low number is like attending a prestigious school, it doesn't mean you are cool, it is just one piece of evidence.
  • Yes, the same thing is done with other personalities, e.g. appearing in a movie with Kevin Bacon. Pete St.John 19:50, 2 November 2007 (UTC)
  • I was told that Erdos is the guy this is done with because he worked on the field of random graph theory, which is related to the idea of the Erdos #. -- 08:41, 4 December 2007 (UTC)
Erdos did a lot of work in Graph Theory generally; the Erdos-Spencer Probailistic Method is conspicuous but not particularly pertinent to Erdos Numbers. Erdos coauthored with a huge number of people (500 odd). If you consider the graph G with E set defined by (x,y) in E when x coauthors with y, then Erdos is the highest degree node in the largest connected component. That makes him a natural "root" node to designate. Pete St.John 19:14, 4 December 2007 (UTC)
And 0 means you are the Great Man himself and thus we now know that there is an afterlife! Teeheehee. -- (talk) 11:17, 25 July 2009 (UTC)

Include Gowers Numbers?[edit]

number of co-authors[edit]

The Erdos Project lists Erdos as having 511 co-authors (i.e. people with a number of 1) but this article only lists 509. I don't know anything about this, but can somebody who does either explain that or fix it (would assume we'd need to fix the number of 2s,3s, but maybe thats in the project website again) Spurgistan 19:59, 29 April 2007 (UTC)

a great many of the papers involved were never in electronic form (unlike today, when almost any paper would be composed originally MSWord, at least, if not TeX or whatever). So cross-linking all the bibliographies is a huge manual job. Pete St.John 19:54, 2 November 2007 (UTC)

Deletion of the Category[edit]

The category for Erdos Numbers, e.g. "Carliz (category persons with Erdos Number 2)" has been deleted. There have been several debates about this, but the most recent one led to deletion. It's actually a very interesting social phenomenon; I invite mathematicians with any concern for pedagogy and general public perceptions to look at some of the reasons given for the deletion. The admin who deleted it has begun to compile some of the reasons given at a talk page, and I have added some rebuttals, but I can tell you it's frustrating. It just astonishes me that so many innumerate people would even care about Erdos Numbers. Pete St.John 19:41, 2 November 2007 (UTC)

Being the editor who nominated these categories for deletion in this third instance, I'd like to comment that I personally do appreciate the pedagogy argument. I just thought that Wikipedia's present category scheme is unsuitable for this activity. A lot of useful connections could be categorized if we had a more flexible technical solution than at present. __meco 11:31, 3 November 2007 (UTC)
The desirability of a better mechanism doesn't justify destroying the exisiting one. Please program a better one, or propose a better existing one, and then vote to switch to the better system, instead of just destroying the existing one. Pete St.John 17:41, 5 November 2007 (UTC)
I think's a great pity that a few of the editors who wanted this category kept have not also retained their civility. Describing the delete !voters as "innumerate people" is a quite unnecessary personal attack. --BrownHairedGirl (talk) • (contribs) 01:07, 4 November 2007 (UTC)
I don't mean to "attack" any individuals. I've described the group as "innumerate", yes, because the logic presented isn't coherent to me. As a whole, they seem to be non-mathematicians with no identifiable interest in the topic, and I have no clue, honestly, what they want. Pete St.John 17:41, 5 November 2007 (UTC)
It may be taht there is a problem with using the category system for this. If there is I must admit that I do not understand what the problem is. It seems to me for example much better to use categories than lists due to the problem of maintenance. At the moment the Edos numbers are being deleted from pages which is a shame as people have put in the effort to calculate (at least an upper bound for) them. When I added the Erdos number to a biography I placed an example of a sequence of authors giving taht number as a comment (sometimes in the text as a comment as wel as edit summary). This leads me to propose that for Mathematicians at least with Edos number up to 4 we make a template that will give an upper bound on the Erdos number with an example of a chai of the right length (surnames of authors will do) in small print. This way if some other mechanism becomes appropriate for forming lists or doing searches it can just be added to the template. How does taht sound? Billlion 17:01, 4 November 2007 (UTC)
The suggestion of an infobox has been made in the discussion on the CfD talk page. I would suggest that there is no reason to stop at 4 or 6 or any number. If a math bio article is about one of the five vertices with Erdos number 13 (supposedly the diameter of the graph at this time), or if the vertex is not connected to the Erdos component, that fact just as interesting, and worthy of mention in an infobox. I'm not realy sure that a chain is really necessary, except in cases where the standard automated tools at MathSciNet don't provide one, or where the standard tools give a number higher than the actual verifiable bound. --Ramsey2006 17:18, 4 November 2007 (UTC)

Distribution over time[edit]

Someone just deleted the statement ... As a result of this drift, the mean Erdős number of living people must increase over time... as unsourced and "illogical". The statement does not say monotone increasing; changing it to "must tend to increase..." would be an improvement. But a little simple logic (which this is) in common language should be OK for explaining what's going on, to laymen. In an article about prime numbers I would show why 3 is prime, not cite a specific reference for that particular piece of fact. So I would prefer improving the explication, to merely deleting it. Pete St.John 16:36, 3 December 2007 (UTC)

  • Perhaps "tends to increase"? While you are at it, you might also want to do something about the section title ("Effect of Erdős' death on the Erdős number") which doesn't really fit the contents. Roger Hui 18:53, 3 December 2007 (UTC)
Related to the section "Effect of Erdős' death on the Erdős number," it should be noted that a handful of people continue to get Erdős numbers of 1. According to the Erdős Number Project, the latest was a paper in 2007. Sillcat (talk) 11:50, 17 February 2008 (UTC)

One of these things is not like the other.[edit]

Article is self-contradictory, viz:

"Effect of Erdős' death on the Erdős number

Given that Erdős died in 1996 and no works of his remain to be published, it is no longer possible for a person to be newly assigned an Erdős number of 1."; and

"eBay auctions

%< snip >%

This is noteworthy because with the exception of a few co-written articles to be published posthumously, 2 is the lowest number that can now be achieved." —Preceding unsigned comment added by (talk) 03:07, 26 December 2007 (UTC)

Gentle, humorous tribute to a great man, or profound measurement of collaboration?[edit]

On the Paul Erdős page regarding Erdős numbers, it says, "Because of his prolific output, friends created the Erdős number as a humorous tribute..." That is a gentle and succinct introduction to what is, in effect, an in-joke between mathematicians who knew of the man. In contrast, the Erdős number page drops no hint for the outsider that this number is not meant to be taken too seriously. I never met the man, but I doubt Paul Erdős would have encouraged it to be considered a true measure of collaboration even though he himself was a fine example of someone who spent his life in collaboration.

The idea of the Erdős number reminds me of the Mornington Crescent (game) which must never be explained lest the joke be lost, or that other pretend game, cricket, whose rules contain no rhyme nor reason (just joking about the cricket). Sadly, if the article is meant to enlighten, the joke needs to be revealed, unless the intention is to keep outsiders in the dark. Surely that isn't the point?

Could the opening paragraph be changed to include a reference to its humorous nature to prevent strangers to the subject from being mislead? —Preceding unsigned comment added by (talk) 21:00, 16 September 2008 (UTC)

You have a point - "created as a humorous tribute" is a great way to put it, too. Let's stick it in the lede and see what happens. - DavidWBrooks (talk) 22:16, 16 September 2008 (UTC)

John von Neumann[edit]

Given that von Neumann, Stanislaw Ulam, and Robert D. Richtmyer wrote the paper which defined the now commonplace method known by the name Monte Carlo (Statistical methods in neutron diffusion , Los Alamos Scientific Laboratory report LAMS–551, 1947), should not von Neumann be listed with an Erdös number of 2? William R. Buckley (talk) 03:03, 15 June 2009 (UTC)

xkcd mention[edit]

Is a webcomic strip mentioning Erdős really notable enough to justify an addition to this article? Evanturner (talk) 06:14, 19 June 2009 (UTC)

Anal note: "Notability" is a term that's applied to article subjects and never to article contents. Sorry to nitpick, but we have enough trouble with one arbitrary cut-off point! If people get the idea that statements need multiple independent blah blah we might as well pack up and go home.
Does the line improve the article? Hm, I'm cool with it. The article is about an interesting idea that's caught on. Mentioning things that people have done with the idea is appropriate. If the article accumulates enough stuff to harm the readers' understanding of the more important bits, then we should trim. --Kizor 07:30, 19 June 2009 (UTC)
I think both WP:TRIVIA and this earlier xkcd are both very relevant. Based on that I think the whole "Cultural anecdotes" subsection should go. —David Eppstein (talk) 13:56, 19 June 2009 (UTC)
As currently existing (one item about Hank Aaron and a trivial bit about one unversity math department), yes it should go. But in general, not necessarily - "popular culture" subsections are useful if they cast light on whether a concept has percolated through society. They are particularly useful for a relatively obscure topic like this one if they show that it is far less obscure than might be assumed. The difficulty is keeping them from turning into tedious lists of "this was mentioned in the Simpsons!" I think in this case the xkcd mention *would* be relevant, because that strip has bled into mainstream. - DavidWBrooks (talk) 16:47, 19 June 2009 (UTC)
Where does this policy end, though? Unfortunately, or fortunately, the author has adopted a shtick which is a penchant for math and physics. Each time he doodles something does the associated math or physics topic on Wikipedia get to mention his latest triumph? IMO, at best his feats should get mention in the XKCD article, and no place else. Quaeler (talk) 17:08, 19 June 2009 (UTC)
No, not each time he doodles but if one of the doodles happens to be useful for reflecting new light on a particular article, then yes. The decision will have to be made each time- this isn't all or nothing.
It's like with the Simpsons; they have mentioned umpty-gazillion different events or items over the years, and umpty-gazillion-minus-10 (or so) do not deserve to be in wikipedia because they add nothing to the article in question. But for those other 10 or so, the mention might cause the casual reader to realize that the topic has more impact than would otherwise be obvious. - DavidWBrooks (talk) 17:26, 19 June 2009 (UTC)
Hmm.. ok, so in this specific case, it's my opinion that the work should not be mentioned the article. If, for some reason, it turns into a wide-spread meme beyond the regular cult-of-XKCD types (ie. it (and by 'it', i mean this specific doodle) gets mentioned in a popular main stream media - like Newsweek, CNN, ...) then i could see it being mentioned in the article; however, the logic of ("XKCD's like totally popular" && "XKCD's latest burp alludes to the Erdős number") doesn't meet the bar. Quaeler (talk) 17:43, 19 June 2009 (UTC)
When an article is about a succesful cultural anecdote, I feel that the cultural anecdotes it has caused are neither trivial nor irrelevant. Avoiding mentions of popular culture is counterproductive when we write about popular culture. --Kizor 17:52, 19 June 2009 (UTC)
This argument is mixing 'culture'. If i write an article about a cultural anecdote which is 'successful' among tribes across the Amazon basin, but unheard of in America's rust-belt, a mention of that anecdote by something low on the radar in the rust-belt is, seemingly by definition, trivial. Reframed: when we write about an article about the Erdős number, we are nowhere close to writing "about popular culture", which seems to make your final sentence not applicable to the case at hand. Quaeler (talk) 18:19, 19 June 2009 (UTC)
I think it's worth mentioning Jredwards (talk) 18:23, 19 June 2009 (UTC)
XKCD mentions things like this all the time. There is no significance to this specific reference. Mintrick (talk) 19:38, 19 June 2009 (UTC)
Not to beat this totally to death, but sampling en Wikipedia as a whole citing a specific XKCD doodle in an article is actually done in 73 different articles. Quaeler (talk) 19:58, 19 June 2009 (UTC)
I agree with not including it. Consider: practically every single XKCD strip will reference some concept, which is bound to have a Wikipedia article. Are we going to add a link to XKCD for every single strip? XKCD is notable to have its own article, but nowhere near notable that referencing a topic, whether it's Erdos number, or wood, means that it should be mentioned in all those articles. If anywhere, the only appropropriate place would be the XKCD article - but I doubt mentions of an individual strip would be considered notable there, so they're certainly not notable here. Mdwh (talk) 21:30, 19 June 2009 (UTC)
Fair enough; I guess xkcd isn't really mainstream enough to warrant mention; if "Zits" mentioned Erdos number, that'll be a different matter! (However, please note that we could, as keen insightful wikipedia editors, mention it in this article without that meaning it has to be mentioned in other articles.) - DavidWBrooks (talk) 22:04, 19 June 2009 (UTC)
No, that wouldn't be important either. Mintrick (talk) 22:07, 19 June 2009 (UTC)
I agree with Mintrick. It's not simply about how mainstream XKCD is, but I think also that it's simply a passing reference in a single strip (i.e., we're not just looking at the notability of XKCD or Zits as a whole - we're looking at the notability of one single strip - and since that individual specific strip probably gets nothing in the way of third party references, it's not notable). Perhaps if there was an entire film built around the concept of an Erdos number, where the film in itself was notable with its own specific article, that might be worth mentioning here. But if there was just a passing mention in a film, we probably shouldn't mention it. Mdwh (talk) 23:39, 19 June 2009 (UTC)
AEEEEIIIIGH *cough* *cough* sorry. It's just that, right at the start of this section, I went into how notability applies to article topics and article topics only, and anything else whould be apocalyptic to our ability to write an encyclopedia. What reason is there to require third-party references for simple, individual, clear, uncontroversial sentences? If there isn't any, please don't... --Kizor 18:14, 20 June 2009 (UTC)
If you have some sort of phobia against calling it notability, how about WP:UNDUE. We should not give undue weight to trivialities, and in this case I think any weight would be undue. —David Eppstein (talk) 18:19, 20 June 2009 (UTC)
I agree that content within an article doesn't need to be notable, however, mentioning the erdos number in a random comic strip is WP:TRIVIA. The only grounds I could see for inclusion would be on the grounds of notability. If it's not notable, I'm not sure on what grounds we should be listing every mention of the erdos number (or mentioning in Wikipedia every single concept that is ever covered by any webcomic)? (And third party references should really still be be provided, not for notability, but for Verifiability and to avoid Original Research.) Mdwh (talk) 18:22, 20 June 2009 (UTC)

(Unindenting) The reason for listing some popular-culture references in articles (that's some references in some articles - not every mention; this isn't an all-or-nothing proposition) is to provide information to the reader about how far a concept has spread through society. This can be useful to readers, which is the point of wikipedia.

XKCD probably isn't mainstream enough to perform that job, but a widely printed newspaper comic strip would be a different matter (which is why I tossed out "Zits" as an example). If Erdos Number had reached enough cultural critical mass to be mentioned in Zits (which will never happen, by the way), that would certainly be notable enough to be mentioned in this article in some form. However, if then became a meme-of-the-day that was mentioned many times in many places, they wouldn't need to all be listed. At that point it would require some thought and effort to figure out how to handle the cultural references in the article. - DavidWBrooks (talk) 13:58, 21 June 2009 (UTC)

David expresses the situation well. There are no simple rules which we can use to answer this type of question. Such situations require editorial judgement of the just the kind that David describes. Paul August 15:45, 21 June 2009 (UTC)


Why does the definition use the phrase 'mathematical paper' when everyone seems to ignore that? (And, BTW, xkcd is so cool, that I cannot mention it without violating NPOV.)

Chippo1 (talk) 10:22, 19 August 2009 (UTC)

Actually, the MathSciNet distance calculator only considers mathematical papers (more specifically, papers within the MathSciNet database). —David Eppstein (talk) 15:36, 19 August 2009 (UTC)

Erdős numbers and time[edit]

Article doesn't clarify following case: X and Y collaborate on paper, but don't have their Erdős numbers. Later on, X collaborates with Z and gets Erdős number of n. Does Y get n+1 or do they still have infinity? Can one's Erdős number decrease in time without them writing any papers, as people on their shortest path get smaller Erdős numbers? —Preceding unsigned comment added by (talk) 15:12, 6 November 2009 (UTC)

Yes, it is possible for the Erdős number to decrease in this way. It does not depend on the order in which the papers on the path were written. —David Eppstein (talk) 15:22, 6 November 2009 (UTC)

xkcd cartoon[edit]

User:DavidWBrooks reverted this with the comment: The problem is, there's an XKCD cartoon for 50 percent of WP articles - just like there's a simpsons reference. That appears strange, since WP has at present about 3.700.000 articles. I would therefore expect about 1.850.000 xkcd cartoons to exist. In fact there are 928, which I am reading these days and found about half a dozen illustrative. For instance the Erdős number strip appeared illustrative. I have known people who co-authored a paper, because of the low Erdös number of another author.

But of course, one has to keep things in perspective: The danger of millions of cartoons flooding Wikipedia must be nipped in the bud. As I said before: I can have this type of thing on de-WP any day (see German humour#Stereotypes). This was a test. en-WP fails. --WolfgangRieger (talk) 14:19, 22 July 2011 (UTC)

Geez, everybody knows that "50 percent" is shorthand for "wicked lots of them"!
Did you read [[2]], as was suggested in an earlier Edit summary? It's an excellent discussion of why "lookit, lookit: this topic was mentioned in an XKCD cartoon!" is rarely a useful addition. (By the way, I count myself among XKCD's many fans and read it regularly; this is not a knock on that invariably clever strip.) - DavidWBrooks (talk) 14:31, 22 July 2011 (UTC)
I read it and noticed that it pertains to the "in popular culture sections". A weblink may just illustrate the subject, the requirements for adding sth to the "in popular culture sections" should be higher, namely, that the specific cartoon has an impact for the subject / popular culture. But I do not want to start the type of elongated discussion I know all too well from de-WP. Greetings. --WolfgangRieger (talk) 15:26, 22 July 2011 (UTC)

Berezovsky number[edit]

One editor added the following section to the article:

"M.V. (Mikhail) Simkin suggests that number should be used to represent the distance between Boris Berezovsky and connecting scientists. See this ref.".

Another editor removed the section, with the wp:ES of "Berezovsky number has no scientific significance. The living persons mentioned in Simkin's paper might dislike the association.".

Well, the cited repository looks to me to be a reasonable wp:RS.

The paper is already in the public domain, and Wikipedia (WP) doesn't generally wp:censor existing external content. Why might living persons mentioned in Simkin's paper dislike the association? Even if 1 of them were to, it's still not WP's job to bow to their sensitivities, and WP has means for them to express their views & ban bad sources, if proven.

Hence, I have reinstated the section concerned, for now, until any wp:consensus emerges on this page. As ever, other editors are cordially encouraged to have further say, below here. Trafford09 (talk) 13:51, 4 December 2011 (UTC)

The repository is the collection of preprints, mostly of good quality but also some junk. This work is not refereed as in peer-reviewed journals. Everybody can post there a paper on `Simkin number' for example. Simkin's paper is better suitable as a joke for university party, rather than for publication. In the paper Simkin mentions a number of highly respectable scholars as `accomplices', which might be disliked by the mathematical community. My suggestion is to remove the section completely for the reasons just given.ArthurTheKing (talk) 16:16, 4 December 2011 (UTC)

Hmm - that would seem to be some elaborate joke, then.

Do we have any proof (or suspicion?) that any of Simkin's content is inaccurate? Trafford09 (talk) 17:31, 4 December 2011 (UTC)

We don't need it. What we need to include it are reliable sources for its significance. Arxiv preprints are essentially self-published so, unless they are by a recognized expert in the subject area, they don't count as reliable. —David Eppstein (talk) 18:14, 4 December 2011 (UTC)
It is more than elaborate, it is an unethical and bad joke. Apparently, Simkin aimed to attract attention to his person by means not accepted among scientists. His article is offending and misleading. The last of 16 publications on MathSciNet Simkin refers to is dated 1989, even before the fall of the USSR, thus in which sense the people mentioned could be considered as `accomplices' is not clear at all. The content of Simkin's article is partly inaccurate: he claims that "Berezovsky defenders argue that he is an honest mathematician with an Erdős number of 4". Who are these mysterious defenders (or opponents)? These are no more that Simkin's ill fantasy, as long as he is not able to present concrete "defenders that argue..." the claim is unsourced. Above that, the number introduced by Simkin has no scientific significance. Would you think Erdős were happy to see references to such dubious publications?ArthurTheKing (talk) 20:23, 4 December 2011 (UTC)
But the arXiv article appeared in Significance (a magazine of the Royal Statistical Society) . Must be significant enough. — Preceding unsigned comment added by (talk) 23:05, 12 December 2011 (UTC)

Impact vs effect[edit]

Impact is clearly a reasonable choice here. For example, a Merriam-Webster dictionary definition includes (italics in the original):

2 : the force of impression of one thing on another : a significant or major effect <the impact of science on our society> <an environmental impact study>

and 4 of their 5 examples are non-physical impacts of this kind.

Also, in scholarly work it is common to rate the impact of articles, journals, and authors, even in a formal and scholarly context. See for example this discussion on impact factors. LouScheffer (talk) 11:51, 31 October 2012 (UTC)

Notability highly dubious™[edit]

[3] How do you decide that such games as chess and go are more notable than "What? Where? When?"? The entire concept of Erdős number belongs to the folklore of mathematicians, and here we have a list of its variations. Links provided show an automated system that calculates Snyatkovsky number for almost 80 thousand players with statistics that allow anyone to check the small world phenomenon and discover the distribution of this number between its reached bounds (as I see, other variations are calculated individually and little statistics exist, with suggestions like 'every mathematician should have Erdős number less than 10' or what). I guess it's like a 'serious investigation' in terms of a humorous concept. Androniq (talk) 10:09, 16 May 2013 (UTC)

The article What? Where? When? does not mention the name "Anton Snyatkovsky"
Google has zero results for the phrase "Snyatkovsky number".
Google translate for mentions Bacon number and Erdős number, but doesn't specifically name "F number" as being a "Snyatkovsky number" (and even if it did, we would require an independent Reliable source to verify its notability/veracity as a term).
So, "How do you decide [...]?" - By consulting/finding Reliable sources. –Quiddity (talk) 21:05, 16 May 2013 (UTC)
And since when Google is considered notable source? Just because you can find something in search engine doesn't mean it should have wikipedia entry. Also, it's about Notability, not Reliability. Article and list topics must be notable, or "worthy of notice". I for one fail to see how inside joke of small group of mathematicians is more notable than any random anime distributed in bigger number of copies than amount of matematicians who care about Erdos number. Anon (talk) 09:48, 11 July 2013 (UTC)


I believe the name is Erdös NOT Erdős. If so it is misspelled thoughout the article. Also, the URL for this article does not work, perhaps because the character used in his name is not recognized. This should be fixed. — Preceding unsigned comment added by (talk) 23:53, 14 April 2014 (UTC)

What's your source for this, all the sources I've checked spell it "Erdős". Paul August 00:18, 15 April 2014 (UTC)
Ask any Hungarian mathematician and you will get the same answer: Erdős. There is really no question about it, even if you can find sources spelling it differently. The URL with this character is perfectly legal. Older browsers may fail to handle it properly, but relatively recent versions of all the common browsers should have no problem. McKay (talk) 01:57, 17 April 2014 (UTC)

Misha Collins[edit]

I would like to submit Misha Collins as an honorary co-author to a mathematical paper which would assign him the Erdos number of 1. — Preceding unsigned comment added by (talk) 11:15, 6 August 2014 (UTC)

This is part of an internet prank hosted by that person, who is some sort of actor. - DavidWBrooks (talk) 13:04, 6 August 2014 (UTC)

I too would like to submit Misha Collins as an Erdos connection. Actually, the Misha Collins GISHWHES (Greatest International Scavenger Hunt the World Has Ever Seen) is NOT an internet prank (as another author indicated above). Instead, it is a very timely social experiment in both social media and social sciences. By way of getting tens of thousands of people willfully perform crazy tasks in this international scavenger hunt, he has, perhaps indirectly, influenced people's behavior. Many of the tasks involve performing random acts of kindness (RAOK), and by way of creating international teams, foster teambuilding skills and greater understanding of cultural differences and similarities. AODLLM (talk) 15:00, 6 August 2014 (UTC)AODLLM (Army of Dragonflies loves Lucky Minions) Cite error: There are <ref> tags on this page without content in them (see the help page). : GISHWHES Net Web Scavenger Hunt Gets People Out of Their Comfort Zones.

Nope, Wikipedia is not the place for the latest social media fads. This is an artificial attempt to inflate the importance of one particular person.--♦IanMacM♦ (talk to me) 15:07, 6 August 2014 (UTC)
These comments misunderstand the nature of Erdős numbers and of this page. An Erdős number is not created nor in any way officially recognized by being listed here at Wikipedia. Rather, what we list here are Erdős numbers that are already recognized, either by the Erdős number project or by data in MathSciNet. To find an Erdős number for someone (like Collins), that person needs to have been listed as the co-author on a published scientific paper with someone else that has an Erdős number. If that happens and the resulting Erdős number is at most three, then we can list him. Nothing more and nothing less will work. So all the effort to include content here is pointless. Certainly an Erdős number of 1 is out of the question; that could only happen for a paper co-authored by Erdős himself, and he's long dead. Something larger is not impossible in principle, but it would involve Collins (not someone else in his name) doing actual mathematical research with a published mathematician, and then getting it published, something that seems highly unlikely to happen within the timeframe of this scavenger hunt. "Honorary coauthorship" (adding someone as an author who had no connection to the work described in a paper) is considered quite unethical so I would not encourage trying to accomplish that. —David Eppstein (talk) 16:07, 6 August 2014 (UTC)
May I suggest that the Web Scavenger Hunt pursue its goals by having people do something useful and intelligent, instead of unpaid publicity stunts for B-listers. If you must do it in wikipedia, creating legitimate articles about unreported topics would be wonderful! - DavidWBrooks (talk) 22:17, 6 August 2014 (UTC)

The Erdős Couch Number[edit]

The Erdős Number is known (and the Bacon Number et al. based on it), but a more significant and prestigious number is the related Erdős Couch Number. Mathematician Paul Erdős himself was the Zero. Anyone who has sat on a couch with Erdős is a One. Anyone other than Erdős, and not a One, who has sat on a couch with a One is a Two, and so on. For example, the late logician and philosopher Willard van Orman Quine of Harvard University could claim to be at most a Two, because he once sat on a pew (which counts as a couch) next to me. Erdős had once sat on a couch where I was sitting, in a lobby in Florida. (I didn't converse with either, but that doesn't affect the Number.) Erdős Couch Numbers have been considered too exclusive to reveal on public websites, especially since riffraff might have an incentive to plop down beside one on a couch merely to obtain a better (lower) Number. But, being known for your low Erdős Couch Number could help a shy male pick up chicks (something my Erdős Number never did). Fritz Lehmann. (talk) 12:32, 28 August 2014 (UTC)

Should a Notable Person's Erdos of 4 be highlighted[edit]

Another editor and I are having this agreement about Judge Richard Posner. I contend not, because a huge portion of the chattering classes probably has such a number, and if we list one, we are Wiki-ethically obligated to try to list as many as we can. Do the math! Bellagio99 (talk) 16:18, 21 November 2014 (UTC)

I agree. Maybe it would be worth mentioning if we were talking about the president or the Pope or some incredibly well-known person like that, but otherwise I think that 3 is the limit. (Hey, that's a math joke! Limit - get it? ) - DavidWBrooks (talk) 16:35, 21 November 2014 (UTC)
In general an Erdos number of 4 is nothing to get excited about. But if it shows that some seemingly unrelated field (such as law in this case) then it's significant since it expands the connected component of the Erdos graph significantly, which most other members of strength 4 do not. So the importance here is that the number is finite, not that it's particularly low. LouScheffer (talk) 05:16, 22 November 2014 (UTC)

Include Posner?[edit]

The argument for including Posner is that he's a staple of legal scholarship. This means, of course, that lots of other legal folks will have finite Erdos numbers (including, I strongly suspect, almost all the US Supreme Court Justices, and at least some of the US presidents who were academics at some point in their career). This is the exactly what Castro and Grossman points out in Famous Trails to Paul Erdos. In particular see table 2, noting that some of the famous folks in other fields are shown to have Erdos numbers of at most 7 or 8. LouScheffer (talk) 05:12, 22 November 2014 (UTC)

At least one Supreme Court justice, Ruth Bader Ginsburg, has a finite Erdos number, due to the chain AE Roth -> RA Posner -> L Epstein -> JL Spaeth -> AL Kalleberg ->BF Reskin -> DJ Merritt -> RB Ginsburg, giving Ginsburg a number of at most 10. Each of these links can be found with Google Scholar. Of course this is original research, so I did not include it in the article. LouScheffer (talk) 04:31, 24 November 2014 (UTC)
Continuing this thread, at least one American president had a finite Erdos number. The chain of jointly argued decisions might go RB Ginsburg -> W Rehnquist -> WO Douglas -> Harlan Stone -> William Howard Taft. (all from the Case Law search of Google scholar). LouScheffer (talk) 04:53, 24 November 2014 (UTC)

possible spam[edit]

Isn't the link to Phys Author Rank Algorithm a spam? It is self referential and no source demostrates its relevance. Shouldn't the link be deleted? (talk) 11:25, 7 August 2015 (UTC)

That sentence and its references don't seem to have any substantial connection to Erdos numbers; I have removed it. --JBL (talk) 15:34, 7 August 2015 (UTC)

Collaboration distances over long time scales.[edit]

The article says: Collaboration distances will necessarily increase over long time scales. This is likely to be true (only in a special sense, see below), but not a necessary property of the graph. Suppose with a bit of exaggeration that there are presently hundreds of "isles" of mathematicians working in different fields, with no collaboration connecting the isles (this is a conceivable situation, though not the actual one). The isles might continue to be isolated for centuries; then just a few collaborations between authors belonging to different isles will immediately turn a great deal of infinities to finite numbers. What is true is that:

(1) Given mathematicians X and Y, ther collaboration distance cannot increase with time, but it might decrease.

(2) The Erdos number of some mathematician X in the far future cannot be smaller than (t-T+1)a where t is the time of the first publication of X, T the time of the last Erdos publication, and a is the largest possible publishing time of any mathematician (the difference between the time of the first and the last publication) (or some similar formula, I have not checked exactly full details). (talk) 11:46, 7 August 2015 (UTC)

Point (2) is surely what was intended; I have made the small wording change necessary. --JBL (talk) 15:36, 7 August 2015 (UTC)
Thank you all. (talk) 21:44, 7 August 2015 (UTC)

Graeme Simison - author - Erdos number 4 - Erdos-Bacon number 8[edit]

Link see - Extract - "Nerd fact: My (Kevin) Bacon number is 4 – Self – Dominique Simsion (Voluntary Act) – Beth Child (Push Up) – Meryl Streep (Evil Angels) – Kevin Bacon (The River Wild) My Erdos number is 4: Self – Daniel Moody – Amotz Bar-noy – Nathan Linial – Paul Erdos …. … which gives me the (rare) Erdos-Bacon number of 8. Colin Firth and Natalie Portman have 6s. Partner also 8. Of course." (talk) 08:47, 18 August 2015 (UTC)

You need much better sourcing than your personal website. See WP:RS and WP:V. Sundayclose (talk) 13:30, 18 August 2015 (UTC)

Largest Erdos number[edit]

The last paragraph of the Overview section says 134,007 mathematicians have an Erdos number. So we must know the Erdos number of each of them. What is the largest Erdos number? This would fit well into that paragraph. Loraof (talk) 16:45, 17 August 2016 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just modified one external link on Erdős number. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

As of February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{sourcecheck}} (last update: 15 July 2018).

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 07:53, 22 September 2017 (UTC)

Earliest person to have a finite Erdős number[edit]

The article claims that according to Laplace is such a person with the EN=14. But Laplace published an article in co-authorship with Antoine Lavoisier (which is used to calculate EN for Laplace), and Lavoisier was born in 1743, 6 years before Laplace.

This has a good, reliable source, so I changed it. One the other hand, Lavoisier was a chemist, so Laplace might still be the earliest mathematician. But that's not what the sentence says. LouScheffer (talk) 19:13, 26 September 2017 (UTC)
Since Erdős number link into other fields, the oldest number might not belong to a mathematician. For example, a link to the legal field is Roth->Posner. But Posner is linked to Supreme Court justice Ginsberg, via AE Roth -> RA Posner -> L Epstein -> JL Spaeth -> AL Kalleberg ->BF Reskin -> DJ Merritt -> RB Ginsburg. And once on the Supreme Court, there are lots of co-authored decisions, i.e RB Ginsburg -> W Rehnquist -> WO Douglas -> Harlan Stone -> Holmes -> Harlan -> Clifford -> Catron -> Story -> Washington -> Cushing, where Cushing (EN<=20) is the earliest-born Supreme Court Justice (born 1 March 1732). All these links can be found in Google Scholar. (As a written, co-authored publication on an unsettled question, it seems to me that Supreme Court (or any appeals court) decision should count.) LouScheffer (talk) 15:39, 26 September 2017 (UTC)
Some sources limit the definition of Erdős numbers only to publications on mathematical topics (widely construed), not to academic scholarship more generally. —David Eppstein (talk) 17:17, 26 September 2017 (UTC)

Collaboration distance site is password locked[edit]

The site now requires a login. Does anybody know since when this is and if it is an error by the admins there? --Ralf Muschall (talk) 14:07, 5 October 2017 (UTC)

Don't know if it is an error. The site works from a university IP without a password, using the university's subscription - the same as most of MathSciNet. This could be accidental or an intentional change, of course. Could you see if this link works better for you? [4]. Thanks, — Carl (CBM · talk) 14:19, 5 October 2017 (UTC)
As far as I know it has always required either a log in or an IP address from a subscribing institution. —David Eppstein (talk) 16:26, 5 October 2017 (UTC)
It works for me from my house with no login, if I use the link I posted above. The link that used to be in the article does not work, however. I think they must have moved the page on the AMS website, or consolidated the "free" version of the tools into one place, and our link was not updated to match. — Carl (CBM · talk) 21:29, 5 October 2017 (UTC)


I disagree that scientiometrics is of "dubious relevance". An Erdos number might very well be considered is a scientiomentric measurement. There is no reason that such measurements cannot be fun as well as useful. --Wikiman2718 (talk) 14:31, 20 May 2019 (UTC)

Scientometrics is about serious measures of academic productivity. An Erdős number says nothing about one's productivity and should not be taken seriously as a measure of anything beyond maybe similarity of research interests to Erdős's. —David Eppstein (talk) 16:08, 20 May 2019 (UTC)
I see no reason that scientometrics should not be fun. This source[1] calls the Erdos number "a topic of research in scientometrics". --Wikiman2718 (talk) 16:17, 20 May 2019 (UTC)
The source's title is "Download limit exceeded"? That definitely makes me think it's reliable and on-topic. —David Eppstein (talk) 16:37, 20 May 2019 (UTC)
The source's title is most certainly not"Download limit exceeded." Wikipedia's citation formatted has made an error. If you click the link, you will be able to download a paper which fits the definition of a reliable source based on reliable sourcing guidelines. This source[2] describes the Erdos number as bibliometric in nature, which is the same but more general. This also justifies a link to the bibliometrics page. It's no wonder that Wikipedia has trouble keeping newcomers. Edit warring and mockery are not suitable behavior for an administrator. I am only trying to improve this page. --Wikiman2718 (talk) 16:51, 20 May 2019 (UTC)
If there were a LIKE button on Wikipedia, I'd push it. But I support Wikiman on all counts. The Erdos number is aa lodestar in the sociology of science: how do scholars connect?Bellagio99 (talk) 17:26, 20 May 2019 (UTC)

I propose that the lede be changed to the following in order to recognize the Erdos number's place in the knowledge tree as recommended by the manual of style:

The Erdős number (Hungarian: [ˈɛrdøːʃ]) is a scientometric index which describes the "collaborative distance" between mathematician Paul Erdős and another person, as measured by authorship of mathematical papers.[3] The same principle has been applied in other fields where a particular individual has collaborated with a large and broad number of peers.


  1. ^ "Download Limit Exceeded". Retrieved 20 May 2019.
  2. ^ Uddin, Shahadat; Hossain, Liaquat; Rasmussen, Kim (28 February 2013). "Network Effects on Scientific Collaborations". PLoS ONE. 8 (2). doi:10.1371/journal.pone.0057546. ISSN 1932-6203. Retrieved 20 May 2019.
  3. ^ Smeaton, Alan F.; Keogh, Gary; Gurrin, Cathal; McDonald, Kieran; Sødring, Tom (2003). "Analysis of Papers from Twenty-five Years of SIGIR Conferences: What Have We Been Doing for the Last Quarter of a Century?". SIGIR Forum. 37 (1): 49–53. doi:10.1145/945546.945550. ISSN 0163-5840. Retrieved 20 May 2019.

I would like to hear opinions and obtain consensus. --Wikiman2718 (talk) 18:13, 20 May 2019 (UTC)

  • Oppose The proposed lede would both treat a recreational idea too seriously and make the opening passage of the article more difficult to read by loading it up with jargon. The Erdős number is fundamentally entertainment, and reliable sources (like Hoffman's biography, the standard source for Erdős lore) treat it as such. Note that even our own guidelines for academic notability explicitly avoid putting any weight on it. XOR'easter (talk) 20:01, 20 May 2019 (UTC)
  • Oppose unnecessary jargon in lead per WP:TECHNICAL. This doesn't do anything to help readers understand Erdős numbers; it appears to be intended purely as a way to shoehorn the link into the article. —David Eppstein (talk) 20:11, 20 May 2019 (UTC)
  • Oppose adding it the the lede. I'm fairly technical but did not know the term, so I'm guessing it would be meaningless to many, and the lede is supposed to be easy to read, even by the just curious. I am OK with it being in the "See Also" section, as researchers have used it for serious purposes. LouScheffer (talk) 20:44, 20 May 2019 (UTC)
  • Support: In response to the argument that the link is too technical, I supply the following rebuttal:
1) Wikipedia's manual of style tells us to consider an article's place in the knowledge tree. That is the purpose of this link. Therefore, Wikipedia's technical jargon guidelines must be balanced against this principle.
2) The readers are likely to be interested in this subject. We should place links that interest and inform the readers. I don't think that this is "shoehorning".
3) While WP:TECHNICAL tells us not to fill up the lede with jargon, I don't think that one technical term is too many. The meaning of the jargon is inferable from context, and is not likely to challenge the intellect of the kind of person who reads this article.
  • In response to the argument that the Erdos number is not a "real" scientomentric index, I supply the following rebuttal:
1) The Erdos number provides us with sociological data regarding the sciences. I know that it is typical for scientomentric indexes to provide us data about research impact and such, but there is no reason that it has to work for that purpose. One might argue that this kind of sociological data is not "valuable", but then again, one could just as easily argue that any other piece of sociological data is not valuable. This kind of sociological data is completely valid.
2) Reputable sources do consider it a scientometric index. I doubt that any reputable source will say that it is not. Therefor, as wikipedians, we should consider it a scientometric index without qualification.
3) With these things considered, I do think that the acknowledgement that the Erdos number is a scientometric index is an asset to this article. I welcome new votes, and urge those who have already voted to reconsider.--Wikiman2718 (talk) 21:13, 20 May 2019 (UTC)

I had seen Wikiman2718's edit on my watchlist before David Eppstein reverted. It made my eyebrows go up a bit, but after poking around a bit I decided it was a good addition to the See Also section. In particular, while the Erdős number started as an in-joke, it's definitely taken on a life of its own, and it has in fact been studied by people in the field of scientometrics (or at least people publishing in the journal Scientometrics; e.g. [5] [6] [7] [8] -- slap "erdos" into their search bar to find more). So I would support restoring it to that section. I don't think the proposed change to the lead is good, though -- calling the Erdős number a scientometric index is kind of right but also very misleading, and I agree with XOR'easter about the jargon issue. --JBL (talk) 20:28, 20 May 2019 (UTC)

I don't have any substantial objection to listing it in the "See also". Small cost, small benefit, I'd say. XOR'easter (talk) 22:39, 20 May 2019 (UTC)
  • Oppose the proposed lede change. I agree with David the Erdős number is virtually meaningless as a measure of anything particularly useful (by the way mine is 3). Paul August 22:57, 20 May 2019 (UTC)
  • In recognition of the current consensus that scientometrics is worthy of the "see all" section, I have placed it there. However, I would like the vote as to it's position in the lede to stay open for at least a week so that all interested parties may participate. --Wikiman2718 (talk) 23:15, 20 May 2019 (UTC)
  • Oppose: The Erdős number of a mathematician has little to do with scientometrics.--♦IanMacM♦ (talk to me) 04:53, 21 May 2019 (UTC)
  • Oppose" It certainly doesn't belong in the lede. -DavidWBrooks (talk) 10:56, 21 May 2019 (UTC)
  • There still seems to be disagreement about whether or not the Erdos number is a scientometric index. As far as I can tell, the literature does unanimously refer to it as a scientometric index, and therefor it is one. If anyone has a counterexample, please post it here. Therefore, the issue at hand is how to accommodate the guideline which tells us to link to an article's place in the knowledge tree and the guideline against jargon at the same time. I have boldly placed an indirect link in the lede which accommodates all concerns. Per wikipeida's BRD guidelines "Cycle. To avoid bogging down in discussion, when you have a better understanding of the reverter's concerns, you may attempt a new edit that reasonably addresses some aspect of those concerns. You can try this even if the discussion has not reached an explicit conclusion, but be sure you don't engage in any kind of edit warring." --Wikiman2718 (talk) 15:57, 21 May 2019 (UTC)
For convienience, I have posted the contested edit here:
The Erdős number (Hungarian: [ˈɛrdøːʃ]) is a measure of the "collaborative distance" between mathematician Paul Erdős and another person based on co-authorship of mathematical papers. The same principle has been applied in other fields where a particular individual has collaborated with a large and broad number of peers. --Wikiman2718 (talk) 19:23, 21 May 2019 (UTC)
    • That's not how BRD works. The time for the B part has passed. You made an edit, it was reverted, and we're discussing it here. Don't pretend your concerns have been addressed and reinstate your edits or variants of them when they clearly do not have consensus.—David Eppstein (talk) 16:20, 21 May 2019 (UTC)
David Eppstein: In fact, this is how BRD works. According to this nice quote that I have extracted for you: "Cycle. To avoid bogging down in discussion, when you have a better understanding of the reverter's concerns, you may attempt a new edit that reasonably addresses some aspect of those concerns. You can try this even if the discussion has not reached an explicit conclusion..." I have addressed all valid concerns, as explained in my comment. If you still object, you may start a second discussion about any remaining concerns you still have. Since I believe that I still have one revert left in this 24 hour period, I have decided to use it. Please do not revert the article back. If you have existing concerns about this compromise, you may state them here and seek a second consensus. --Wikiman2718 (talk) 16:28, 21 May 2019 (UTC)
Reinstating the same bad edit to the lead sentence, clearly opposed by the discussion above, and merely disguising it by making a WP:SUBMARINE link on top of it, celearly indicates that you DO NOT "have a better understanding of the reverter's concerns". This is heading towards WP:IDIDNTHEARTHAT territory. —David Eppstein (talk) 16:47, 21 May 2019 (UTC)
@Wikiman2718: So far five or six seven people have opposed adding this link to the lead, and none have supported it. To add it under those circumstances is foolhardy; to repeatedly add it is disruptive edit-warring, and is liable to get you blocked. --JBL (talk) 16:56, 21 May 2019 (UTC)
It is my understanding that this edit was opposed for the following reasons: 1) WP:TECHNICAL: This is a valid concern. I have addressed it completely by making the link indirect. 2) The belief that the Erdos number is not a scientometric index. I do not believe that this concern is valid. While it may seem odd, the sources clearly state that the Erdos number is a that it is scientometric index, and if you disagree you will have to cite a source to challenge it. Only valid concerns must be addressed in consensus. 3) Wikipedia's manual of style tells us to consider an article's place in the knowledge tree. 4)WP:SUBMARINE is a new concern, which was not raised in the original discussion. If you would like to raise it now, you are free to do so in seeking a second consensous. I do not believe that an indirect link violates this principle.
As I have states twice previously, I am following the BRD guidelines: "Cycle. To avoid bogging down in discussion, when you have a better understanding of the reverter's concerns, you may attempt a new edit that reasonably addresses some aspect of those concerns. You can try this even if the discussion has not reached an explicit conclusion..." By reverting my edits, you are not. Additionally, Wikipedia's guidelines require you to assume good faith. --Wikiman2718 (talk) 17:06, 21 May 2019 (UTC)
The salient phrase is even if the discussion has not reached an explicit conclusion. A discussion in which there is a lone advocate for a position that has been opposed by 7 other people can be fairly said to have reached a conclusion. The fact that you disagree with this conclusion is mildly unfortunate, but it does not give you grounds for ignoring it, and especially not for conducting an edit-war to ignore it.
By the way, nothing I have said here could possibly be construed as not assuming good faith: it is obvious you are acting in good faith, you are just wrong about whether this link should appear in the lead sentence of this article. --JBL (talk) 17:13, 21 May 2019 (UTC)
1) The discussion was about whether or not an explicit link should be made. The consensus is that it should not. However, all guidelines must be taken into account when reaching consensus, which is why I have proposed this compromise to address existing concerns. I am following guidelines: "Cycle. To avoid bogging down in discussion, when you have a better understanding of the reverter's concerns, you may attempt a new edit that reasonably addresses some aspect of those concerns. You can try this even if the discussion has not reached an explicit conclusion..." My edit was not made to edit war. In refusing to seek second consensus, as required by wikipedia's guidelines, your behavior may constitute edit warring. --Wikiman2718 (talk) 17:21, 21 May 2019 (UTC)

My understanding of WP:SUBMARINE is that you should not trick the reader by linking to one page when they expect another. The guidelines state: "If a link takes readers to somewhere other than where they thought it would, it should at least take them somewhere that makes sense." I think we have agreed by now the sources indicate that the Erdos number is a scientometric index. Therefore I believe that this link makes sense, and WP:SUBMARINE does not apply here. If it does, it must be balanced against the weight of using links to identify the place in the knowledge tree. --Wikiman2718 (talk) 17:38, 21 May 2019 (UTC)

As I think we have agreed by now the sources indicate that the Erdos number is a scientometric index. I, for one, don't agree. None of the sources presented here have changed my opinion that it is essentially recreational mathematics for professional mathematicians. It is lore; it is entertainment; it is a nod of respect to Erdős. Scientometricians may mention it or compare their own ideas to it, but calling it scientometrical itself gets the mathematical community wrong. XOR'easter (talk) 17:47, 21 May 2019 (UTC)
This source[1] claims that the Erdos number is a scientometric index. If you disagree, you will have to find show a source which claims that it is not. Otherwise, consensus must reflect that it is. It is not a wikipedian's place to challenge sources on the basis that they disagree with them. --Wikiman2718 (talk) 17:51, 21 May 2019 (UTC)


  1. ^ Smeaton, Alan F.; Keogh, Gary; Gurrin, Cathal; McDonald, Kieran; Sødring, Tom (2003). "Analysis of Papers from Twenty-five Years of SIGIR Conferences: What Have We Been Doing for the Last Quarter of a Century?". SIGIR Forum. 37 (1): 49–53. doi:10.1145/945546.945550. ISSN 0163-5840. Retrieved 20 May 2019.
You have already linked to that, and I already read it, and my take on it was already expressed by my comment above. I am not "disagreeing" with it; I am evaluating it. Moreover, I have already pointed to a source that backs up my view: Hoffman's biography of Erdős. XOR'easter (talk) 17:56, 21 May 2019 (UTC)
As for the knowledge tree, I find it is amply watered by the links already in the article, including the "See also" section, so that's a moot point. Nor is over-linking a good way to help the reader. Replacing an overly technical turn of phrase with a submarine link is not progress. XOR'easter (talk) 17:58, 21 May 2019 (UTC)
The source you cited does not support your claim. It does not use the word scientometrics or express any view which inconsistent with the view that the Erdos number is a scientometric index. It may call the Erdos number recreational, which it is, but this only shows that scientometrics can be recreational. Many sources find that the Erdos number is a scientometric index, and I have cited only one. If you cannot refute it in accordance with wikipedia's reliable sourcing guidelines, it stands. Additionally, this article is in no danger of overlinking. Please express only valid concerns. I have already stated that I do not believe this link to be a submarine link: I will paste that text here.
My understanding of WP:SUBMARINE is that you should not trick the reader by linking to one page when they expect another. The guidelines state: "If a link takes readers to somewhere other than where they thought it would, it should at least take them somewhere that makes sense." I think we have agreed by now the sources indicate that the Erdos number is a scientometric index. Therefore I believe that this link makes sense, and WP:SUBMARINE does not apply here.--Wikiman2718 (talk) 18:01, 21 May 2019 (UTC)
Repeating your assertion does not make it true. XOR'easter (talk) 18:07, 21 May 2019 (UTC)
(edit conflict) This is quickly becoming tiresome. There is clear consensus against your edit, in either form. --JBL (talk) 18:09, 21 May 2019 (UTC)
I will post a notice on Dispute resolution noticeboards. --Wikiman2718 (talk) 18:14, 21 May 2019 (UTC)
For the avoidance of doubt, I'm also opposed to the "indirect link" to "scientometrics" (per among other things WP:EGG, WP:UNDUE) in the lede. I would suppose, so are most of the other editors who have opposed above. Paul August 18:26, 21 May 2019 (UTC)
Good point regarding WP:DUE. That's kind of what I was going for above, but it's better to make the point explicitly. XOR'easter (talk) 19:11, 21 May 2019 (UTC)

Here is a second source which states that the Erdos number is a topic of study is scientometrics, just in case the first is disputed.[1] --Wikiman2718 (talk) 19:17, 21 May 2019 (UTC)


  1. ^ Balaban, Alexandru T.; Klein, Douglas J. (1 September 2002). "Co-authorship, rational Erdős numbers, and resistance distances in graphs". Scientometrics. pp. 59–70. doi:10.1023/A:1016098803527. Retrieved 21 May 2019.
Yes, I've already linked to this paper above. I very much doubt that linking to it a second time is going to change anything. --JBL (talk) 19:23, 21 May 2019 (UTC)

Upon further reading, sources do not refer to it explicitly as a scientometric index, but only as an object of study in scientometrics and bibliometrics. In light of this discovery, I have decided to resolve the dispute.--Wikiman2718 (talk) 20:18, 21 May 2019 (UTC)

Wow - a long wikipedia edit tussle ends with a thoughtful, adult response. Is that allowed? - DavidWBrooks (talk) 13:09, 22 May 2019 (UTC)