What's your Erdos Number?

[beergeek279]

I'm starting to get into my research (graph algorithms) and was discussing with my advisor small-world graphs. Naturally, Erdos numbers came up, and she mentioned an Erdos number of 3. One of my math profs has a 2 and there's a person in the department with a 1, though I'm not sure if they're in discrete mathematics or not.

So, what's your Erdos Number? Are there any Erdos numbers of 1 on the SDMB (and if so do you want to publish? )?

[CalMeacham]

I looked this up once -- I had to contact a group that had this stuff tabulated and wait for a reply -- but I don't recall my number. Unfortunately, it's not like your Kevin Bacon number. My Erdos number is finite, but not low -- I have papers in common with at least two others who have a huge number of connections, so I'm likely connected through more than one path.

[DaddyTimesTwo]

He presented a paper at the university I attended, several years after I graduated. And one of the professors was working on a paper with him, but I never took one of his classes. And I'm guessing that means my Erdos number approaches infinity.

[Sattua]

Mine is 4.

[Terminus Est]

One of my graduate committee members has an Erdos number of 3. As I have published a (non-mathematical) paper with him, my Erdos number is 4.

Here's an easy way to compute an Erdos number: http://www.ams.org/msnmain/cgd/index.html

[bluecanary]

That site requires some sort of password, Terminus Est.

[Terminus Est]

Oh, sorry. I must have been given access based on my Cornell IP.

[ftg]

3

(And I'm in CS, not Math.)

[MovieMogul]

I thought the thread title referred to Irv Erdos.

The only way I'd have an Erdos # is if it cross-referenced with the IMDB.

[Strinka]

Uh...

So, just what is an Erdos number?

[HPL]

Originally Posted by Strinka

Uh...

So, just what is an Erdos number?

I've been wondering the same thing. Maybe it's like your rank in the stonecutters?

[RealityChuck]

Originally Posted by Strinka

Uh...

So, just what is an Erdos number?

Paul Erdos was a renouned mathematician and the author of many papers in the field. The Erdos Number is similar to the Kevin Bacon number: if you coauthored a paper with Erdos, your number is 1. If you coauthored a paper with someone with an Erdos number of 1, your number is 2, etc.

The most famous non-mathemetician with an Erdos number of 1 is Hank Aaron.

[MaxTheVool]

Can Erdos numbers be by inheritance? If so, I have an Erdos number of 1 from my maternal grandfather, along with some hilarious Erdos anecdotes ("I have just proofed a sheorem... what sheorem have you just proofed?")

[Strinka]

Ah. Thanks. Mine is infinite then.

[Thudlow Boink]

The Erdos Number Project


I haven't ever published anything, but I have taken classes from at least one guy with Erdos number 1.

[Wendell Wagner]

My Erdos number is 4.

[Hypnagogic Jerk]

Actually, I've heard that Erdos (don't know how to write the appropriate diacritic) numbers are computed only drawing from papers published by two authors. For example, if you publish a paper with someone whose Erdos number is n, you'll get an Erdos number of n+1 if you and this person are the only authors, but if there are three or more authors, this paper will not count towards your Erdos number. There are also extended Erdos numbers which, as you'd expect, are computed taking into account any joint paper with other authors, regardless of their number.

I personally might have a finite Erdos number, but I don't know what it is, and it is probably rather high anyway. "High" is relative: it is estimated that a very large majority of people who have a finite Erdos number actually have it smaller than 15.

This said, can someone tell me what paper Hank Aaron published jointly with Paul Erdos? And maybe send me a link to it, if it is possible?

[beergeek279]

Originally Posted by severus

This said, can someone tell me what paper Hank Aaron published jointly with Paul Erdos? And maybe send me a link to it, if it is possible?

It's generally a joke.........at an event (I think it was at Emory University, IIRC), Paul Erdos and Henry Aaron signed the same baseball, so you could say they "co-published" and thus Aaron has an Erdos number of 1.

[Wendell Wagner]

severus writes:

> Actually, I've heard that Erdos (don't know how to write the appropriate
> diacritic) numbers are computed only drawing from papers published by two
> authors.

I've never heard that before. Every description of how the Erdos number works I've heard up to now says that you can connect via any co-authored paper, regardless of the number of co-authors. I just checked the only two books I own that might speak to this issue, the two biographies of Erdos (yes, there are two biographies of Erdos), and they both say that any co-authored paper is O.K., regardless of the number of co-authors.

> "High" is relative: it is estimated that a very large majority of people who have
> a finite Erdos number actually have it smaller than 15.

Indeed, one of the biographies claims that it's rare to have a (finite) Erdos number over 7.

[BobLibDem]

So exactly what makes this Erdos fellow so important?

[Kimstu]

Erdos wasn't the world's greatest mathematician, but he was a mathematical problem-solver and collaborator extraordinaire. Check out the recent biography of Erdos by Paul Hoffman, The Man Who Loved Only Numbers:

A key factor behind Erdös's many research collaborations was his highly unusual lifestyle. Though Hungarian by birth, he left his home country in 1934 to avoid political persecution, and spent the rest of his life with no fixed home, wandering the world, for the most part relying on other mathematicians to house and support him. He carried around all his worldly possessions in a shabby suitcase and a plastic bag. Typically, he would show up on the doorstep of a fellow mathematician -- often unannounced -- and declare, "My brain is open," his own peculiar way of inviting the colleague to spend the next few days engaged in a feverish orgy of mathematical investigation. [...]

It was hard not to like Erdös. There was a childlike innocence to him, coupled with a wry wit that indicated someone not entirely oblivious to the rest of the world. He was a man whose entire life revolved around mathematics. Short and frail looking, almost cadaverous, he generally dressed in a well-worn gray striped suit, with open sandals, and spoke English with a strong, melodic Hungarian accent. He was able to carry on several mathematical conversations at the same time. [...]

The book does have some weaknesses, mostly I think because Hoffman is not a mathematician. For instance, one thing the biography does not bring out, but which I think is important, is that Erdös was not a "great mathematician" in the sense of say, Andrew Wiles, who solved Fermat's Last Theorem. Erdös's most significant single result was in fact a new proof of a result that was already known -- the Prime Number Theorem. He did not spend years working on a single, major problem, as Wiles did, nor did he appear to know much mathematics. Erdös's greatest strength was as a "problem solver" who had a rare ability to ask questions that had just the right degree of difficulty. By and large, he preferred those areas of mathematics where all that is required is a sharp mind; he avoided those areas where it is necessary to absorb vast amounts of information in order to make any progress. It was very much an innocent, childlike approach to mathematics. I think this point should have been made -- and explained -- because Erdös did have a greatness, and it was a unique greatness. (It also explains why Erdös was able to collaborate with so many others. On occasion, a single conversation resulted in a joint paper.)

(Holy crap, my own Erdos number is only 5! Oh wait, that's if you count co-editorship of an edited volume as co-authorship: I'm not sure that's legit.)

[Hypnagogic Jerk]

Originally Posted by Wendell Wagner

I've never heard that before. Every description of how the Erdos number works I've heard up to now says that you can connect via any co-authored paper, regardless of the number of co-authors. I just checked the only two books I own that might speak to this issue, the two biographies of Erdos (yes, there are two biographies of Erdos), and they both say that any co-authored paper is O.K., regardless of the number of co-authors.

You seem to be right, according to the Erdos Number Project. And I thought this is where I had heard that it counted only papers with two authors.. I guess I was wrong.

By the way, thanks to beergeek279 for telling me about Erdos and Aaron's baseball. Oh, and Kimstu, the Erdos Number Project specifically says that they don't count co-editorships, so I guess you're out of luck.

[Cervaise]

Originally Posted by Kimstu

Check out the recent biography of Erdos by Paul Hoffman, The Man Who Loved Only Numbers

That I, a layman, read this book the year it came out in hardback, and have had it on my shelf ever since, doesn't give me some sort of honorary number, does it?

[mojave66]

WOO HOO!!! I just found out I have an Erdos number of 4 even while I'm flunking real analysis! I am thoroughly unworthy of this number! I've gotta tell my math prof this.

How I got it is that the biostatistician on my project published with a UC Berkeley biostatistician with an Erdos number of 2. Has absolutely NOTHING to do with math, unless you count poking at databases all day as being mathematical.

[Wendell Wagner]

The other biography of Erdos is _My Brain Is Open: The Mathematical Journeys of Paul Erdos_ by Bruce Schechter.

[iamthewalrus(:3=]

My Erdos number is 4. Or, at least, it will be once my first paper is published. Scheduled for Feb of '06 right now.

And that requires that you use the looser definition that counts a multi-author paper.

Still. That's pretty freakin' cool.