Who has the lowest Erdos-Bacon-Epstein number?

Lawrence Summers of course has an Epstein number of 1 (the lowest you can have while still being alive), but his Erdos number is a disappointing 6. I don’t know any easy way to calculate his Bacon number, but Summers does have an IMDB page . . . maybe the best connection will be through this documentary, Panic: The Untold Story of the 2008 Financial Crisis, which also includes Epstein friends Steve Bannon and Donald Trump. Bannon produced this documentary about Sarah Palin, featuring Pamela Anderson and Roseanne Barr. And this site gives Anderson a Bacon number of 2. So Summers’s Bacon number is no more than 4, thus his Erdos-Bacon-Epstein is at most 11. That’s stretching it, though, because the definition of Bacon number seems to require that all the people in the chain have acted in the movies. Being filmed as yourself (as with Summers) or being a producer (as with Bannon) doesn’t really count. So I guess Summers has an infinite Erdos-Bacon-Epstein number. Sorry, Larry! No Nobel Prize and no Bacon number. I guess you can forget about the EGOT too.

Here’s a possibility. Sergei Brin has an Erdos-Bacon number of 5, and in the Epstein files there’s this story:

I don’t know that Brin had any emails with the famed financier, but I think it’s fair to guess that his Epstein number is no more than 2, so his Erdos-Bacon-Epstein number is at most 7.

Also . . . it seems that Bill Gates has an Erdos-Bacon number of 6, and he was a known associate of Epstein, so he also has an Erdos-Bacon-Epstein number of 7.

What about me? I have an Erdos number of 3 and, sadly, an Epstein number of 2 (through this guy and this guy). But my Bacon number is infinity, as I’ve never participated in a movie. If they ever make a film of Recursion, maybe. Come to think of it, Kevin Bacon would make a great Bob Dwyer; I could see it working out.

The Erdos-Bacon-Epstein challenge is that you need a connection in academic publication, a connection in the movies, and an email link to Epstein. That last bit is the easiest: anyone can send an email, and I expect that just about everyone online has a fairly low Epstein number. My mom has an Epstein number of 3! Lots and lots of scientists and writers have Epstein numbers of 2 the same way that I do, if they’ve ever worked with or considered working with the notorious book agent John Brockman. I’m locked out regarding Bacon, and lots of other people have infinite Erdos numbers. Peter Thiel, for example, is an Epstein associate and has some IMBD credits but an infinite Erdos number. Similarly, Woody Allen has an Epstein number of 1 (I expect) and a Bacon number of 2 but no Erdos connection.

But here’s a thought: Stephen Hawking! Another Epstein associate–no email from or to him in the recently released files, but I think it’s safe to assign him an Epstein number of 1. His Bacon number is 2 and his Erdos number is a surprisingly high 4–I guess Hawking didn’t actually publish so many papers during his career–so his Erdos-Bacon-Epstein number is 7.

And . . . Noam Chomsky. He has an Erdos number of 4, a Bacon number of 2, and an Epstein number of 1, so a total of 7. And Noam is still active! With some effort maybe he could find the right collaborator and get his Erdos number down to 3 and have an Erdos-Bacon-Epstein number of 6.

But I’d give Noam an asterisk. I got his Bacon number of 2 from wikipedia, where it says he “co-starred with Danny Glover in the 2005 documentary The Peace!, giving him a Bacon number of 2.” But it seems that he was just a talking head in that movie, not acting, so I don’t know if that really counts.

Here’s an amusing bit from Noam’s IMDB page:

“Trivia” . . . that’s about right!

Any other possibilities? Going over to Epstein’s birthday book, we see some famous names with some academic connections, including Nathan “Albedo boy” Myhrvold and Henry “Harvard” Rosovsky. Both are on Google Scholar and no doubt have finite Erdos numbers, but neither seems to have acted in any movies. Myhrvold hosted some kind of cooking show but that doesn’t really count, sorry. Also some prominent-but-not-famous-scientists such as Gerald Edelman, Stephen Kosslyn, and Lee Smolin, but no IMDB acting credits among them. And my namesake Murray Gell-Mann. I hate to see that! But, again, no relevant screen time.

Epstein’s birthday book also has an entry from someone named “Ace Greenberg.” Hey, what is this, a Damon Runyon story?

Really you have to read that birthday book all the way through. There’s a letter from a “Johnny Boy” who says that Epstein is his kid’s role model. How creepy is that? What a horrible parent. I get that some people are seduced by money, sex, and power and thought that Epstein was cool because he had all three, but to bring your kind into that? Yuck! And Myhrvold’s charming collection of animals in sexual positions.

And then there’s Marvin Minsky, who writes that Epstein is the second-quickest intellect he’s ever met. Wha . . .? Either Minsky’s students and colleagues at MIT were a lot slower than I’d have pictured, or Epstein was much more impressive in person than I could possibly imagine based on anything in his emails, or the famous artificial intelligence pioneer was not such a good judge of intellectual quickness. I’m racking my brain here to think of whatever witticisms or quick replies or deep thoughts Epstein had to offer that would lead Minsky to this assessment. Then again, one of the greatest statisticians of the twentieth century said he really enjoyed meeting with Epstein . . . so I guess the man had some charm that just doesn’t come across on paper.

Also featured in the birthday book is Alan Dershowitz. He’s gotta be a contender here, right? An Epstein number of 1–indeed, given all his connections, he practically has an Epstein number of 0–and he has acting credits on IMDB! He’s in this Rob Lowe movie from 2012 with the amusing-in-retrospect tag line, “A political strategist juggling three clients questions whether or not to take the high road as the ugly side of his work begins to haunt him.”

Don’t worry–the Dersh would only take the high road!

Anyway, this movie also features David Harbour, who actually appeared in a movie with the prolific Kevin Bacon. So Dershowitz has an Epstein number of 1, a Bacon number of 2, and an Erdos number of . . . Let’s go to Google Scholar.

Here are the first three links:

In the first one, he says the death penalty should be abolished. In the second, he defends the use of torture. In the third . . . I haven’t read it, but I guess he thinks the death penalty and torture are ok if Israel does it? But this doesn’t get us any closer to Paul Erdos. We need to find some place where Dershowitz has coauthored with a scientist, or someone who would’ve coauthored with a scientist, etc. His papers are mostly solo authored, so it’s tough. Dershowitz was interviewed in the magazine Litigation by Ashish Joshi, who also interviewed judge Jed Rakoff, who wrote an article on eyewitness identification with neuroscientist Thomas Albright, who’s published lots of scientific papers and so surely has had some collaborators who have worked with mathematicians. I don’t know what Albright’s Erdos number is, but if it’s 5, then this would give Dershowitz an Erdos number of no more than 8, thus an Erdos-Bacon-Epstein total of 11. Not bad! Too bad he couldn’t get closer on the Erdos part of the game. It’s kind of like being a potential triathlete but barely being able to swim.

What Dershowitz should do is publish a paper with Steven Pinker! It would be easy. Pinker famously wrote up a linguistics argument in defense of Epstein as a favor for Dershowitz. Pinker knows his linguistics so I’m sure this would be publishable somewhere. Pinker has an Erdos number of 3 (just like me!), so this collaboration would make Dershowitz an Erdos 4, giving him a combined Erdos-Bacon-Epstein number of 7.

What about Pinker himself? Erdos 3, Epstein 1 (or maybe 2), but no IMDB acting credits, just some TV appearances as himself, so no dice.

Again, the hard part is finding people in Epstein’s orbit who have academic publications and acting credits. Some academics, some people in the entertainment industry, but not many with both. Bacon himself, for example, has no academic publications. (Nor does he have any direct connection to Epstein, but he may well have had email exchanges with Kevin Spacey or someone else in the Epstein orbit.) I don’t think Andrew Mountbatten-Windsor has any academic publications either. I guess he was too busy with his research to get around to writing any of it up.

But wait . . . here’s a dark-horse candidate. Looking again at the contributors to Epstein’s birthday book, we see a bunch of businessmen, some politicians and scientists, some girlfriends, and some names that I didn’t recognize at all, including someone named Stuart Pivar. Hmmm . . . according to wikipedia, there’s person with that name, born 1930, who’s a “chemist and art collector known for his unorthodox views about evolution.” That sounds like someone in Epstein’s circle, and, indeed, scroll down the wikipedia page and you’ll see the connection. And it says here that he found a van Gogh painting from a flea market or something like that! Pivar’s Epstein number is 1 (you get that by contributing to the birthday book) and . . . ummm, yes, he has an IMDB acting credit, having been one of many many people to have played the role of Socrates in the 2010 film, The Death of Socrates (writers are listed as “Plato, Benjamin Jowett, and Natasa Prosenc Stearns“) and featuring Ray Abruzzo, who has a Bacon number of 2. So, like Dershowitz, the somewhat obscure (although not completely obscure, I guess, given that there have been magazine articles about the guy) Pivar is an Epstein 1, Bacon 2. What about his Erdos number? Pivar was a chemist. According to wikipedia, “As an inventor, he made a large fortune in plastics.” On Google scholar, we see this 2016 paper, “Origin of the vertebrate body plan via mechanically biased conservation of regular geometrical patterns in the structure of the blastula,” with David B. Edelman, Mark McMenamin, and Peter Sheesley. At this point, Pivar is considered a bit of a crank, so I doubt these coauthors are serious scientists themselves, but maybe we can follow some links and get to mainstream science, and there to mathematics, and there to Erdos. McMenamin has a Google Scholar page but it all seems pretty narrow . . . hmmm, there’s a paper, “Did surface temperatures constrain microbial evolution?”, with David Schwartzman and Tyler Volk. Schwartzman seems like a bit of a dead end, but Volk, in addition to publishing some cranky-looking things himself (“Gaia’s body: toward a physiology of Earth”) also published a speculative paper in Science with coauthors including earth scientist Klaus Lackner, who I saw speak at Columbia once! Lackner hung out sometimes with Upmanu Lall, who has published a paper with me, and I have an Erdos number of 3. If we suppose that there is a link connecting Lackner to Lall, this would give Pivar an Erdos number of no greater than 9.

But it’s hard to imagine that 9 is the best we can do for Pivar. Another route is through another of his collaborators on that paper, Edelman, who also wrote this article:

Ummmm, I’m skeptical. Evaluating a horserace prediction method based on only 300 races? C’mon. But, hey, all things are possible. Perhaps this Edelman fellow is now rolling in the dough. Maybe he owns a few Arabian thoroughbreds himself!

Edelman (not the Gerald mentioned earlier, unfortunately) also wrote a couple of papers on finance. That seems like a possible route to mathematics, and thus Erdos. There’s a paper with Patrick O’Sullivan on Adaptive Universal Profiles, but his links are all applied finance, no math happening here, also a book, Numerical Methods for Finance, with two coauthors, including a John Appelby who wrote some papers on differential equations . . . whatever. I give up on this one. It’s surprisingly difficult to navigate the publication network. It’s hard for me to believe that we can’t get Stuart Pivar‘s Erdos number below 9, but maybe that’s what it is. I’ll just tag Pivar with an Erdos-Bacon-Epstein number of 9 + 2 + 1 = 12.

Also, I don’t like Pivar. I’ve never met him, but he appears to be a liar. His wikipedia page says, “Pivar was also a well-known friend of the late financier Jeffrey Epstein; however, the two had a falling out prior to Epstein facing charges for sex crimes. Pivar corroborated the account of Maria Farmer, a graduate of the New York Academy of Art in 1995, who stated that she had informed him about her abuse at the hands of Epstein in 1996. According to Pivar, this was when the friendship with Epstein ended.”

But Pivar contributed to Epstein’s 50th birthday book. Birthdays seem to have been a big deal to Epstein; his sycophantic correspondents are always wishing him happy birthday. Anyway, Epstein was born in 1953, so his 50th birthday was in 2003, so unless Pivar wrote that tribute seven years ahead of time, he was lying when he said in 2019 that he’d ended his relationship with Epstein back in 1996.

Also this bizarre bit:

In August 2007, Pivar sued a science blogger named P. Z. Myers and Seed Media Group, which hosted his blog, alleging defamation. Myers had lit into Pivar’s work, calling him “a classic crackpot.” In his complaint, Pivar made a point of mentioning by name two prominent members of SMG’s board: Jeffrey Epstein and Ghislaine Maxwell. The lawsuit was later dropped.

Remember blogging? That used to be a thing.

And . . . the winner!

This is someone you’d never have expected. According to wikipedia, MIT mathematician Daniel Kleitman has an Erdos-Bacon number of 3. Kleitman was at MIT forever, so I bet he has some email exchanges with Noam Chomsky or some other Epstein intimate, in which case his Erdos-Bacon-Epstein would be 5. I knew Daniel Kleitman! He was my freshman adviser at MIT. There was a group of about five of us, and we met with him in his office a few times. He was a nice guy, very blunt spoken–not in a crude way at all, just the kind of guy who would say something was bullshit. I told Kleitman I was interested in doing research, and he connected me with a graduate student, Susan Assmann, who gave me a project to work on. It took me a year to figure it out. I learned a lot from the experience; the story is here.

From a statistical point of view, the lesson here is that low Erdos and Bacon numbers are rare, so the best way to perform this search is not to start with Epstein associates but rather to take Erdos-Bacon champs and then go to Epstein from there. For example, mathematician Jordan Ellenberg has an Erdos-Bacon number of 5. I’ve emailed with Jordan, so his Epstein number is at most 3, giving him an Erdos-Bacon-Epstein upper bound of 8. But Jordan could well have been contacted by Brockman at some point, in which case he’d have a number of 7, tied with various other people listed above.

Does anyone in the world have an Erdos-Bacon number of 6? I don’t know.

What’s the point?

Why do all this? Why did I spend two precious hours of my time on earth tracking down these links and writing all this? Or, maybe more to the point, why did you read all this. (I’m conditioning here on whatever subset of our blog audience has who’ve read this far down on the post.)

The quick answer is that connections can be interesting. You can learn all sorts of unexpected things from this sort of quasi-stochastic search.

Another answer is that seeing these connections of various elite and not-so-elite people gives us some sense of the social world. It’s a core sample of part of American society.

The other interesting thing about the Epstein files is the content. Not the crude sexism: people will say all sorts of things in private, so this sort of thing is hardly shocking. If a cookbook writer / retired technology executive thought it was cute to talk about sex with one of his rich friends, so be it. The part that was more stunning to me was all these luminaries who seemed so impressed by Epstein. In addition to the aforementioned pioneers of statistics and computer science, you’ve got ultra-successful businessmen such as Gates, artists such as Andres “Piss Christ” Serrano, leading physicists (sorry, no Bacon number here; according to IMDB the closest she came was an uncredited role on a TV show, and I think that only movies count), etc.

I get it that lots of politicians got caught in a net: if you’re a politician, you pretty much can’t avoid getting close to lots of distasteful people. I’m not saying that it’s cool that Trump, Clinton, Richardson, Bannon, Thiel, etc. were friendly with Epstein, but it’s also not so clear what the alternative would be. If you’re in politics, you only have a limited number of times you can piss off powerful and well-connected people. But in academia and in business, you can do what you want most of the time. The idea that these people were choosing to hang out with Jeff, going to the trouble to wish him happy birthday . . . it’s just weird. Again, I think Epstein must have had a real ability to talk with lots of different people, making people as different as Bannon, Chomsky, Minsky, and Summers to all think he agreed with them. And that’s kind of interesting.

Finally, I laugh because otherwise I would cry. I joke about all this because that’s a way to deal with disturbing things.