Erdős Number
Saturday October 21, 2006
I’ve been doing a lot of reading and work with networks (or more generally, graphs). Lots of mathematical and conceptual connections with Markov chain theory relevant to MCMC, and, separately, Bayesian belief networks (i.e., statistical models).
Anyway, it turns out I have a finite Erdős number: ok, so none of my publications show up on MathSciNet; but as far as I know my Erdős number is 6 (Doug Rivers to Quang Voung, to Jean-Jacques Laffont, to Eric Maskin, to Peter Fishburn, to Erdős). I am also 6 steps away via a collaboration with Frank Vella (whose link to Erdős also runs through Peter Fishburn).
I have an active collaboration with Howard Rosenthal, who is Erdős 4, through a link to John Ferejohn (Erdős2, he’s co-authored with Peter Fishburn, as has Steve Brams).
In Stanford Stat, Trevor Hastie is Erdős3, and Brad Efron and Charles Stein are Erdős2, because Persi Diaconis’s Erdős number is 1…!
MathSciNet says Gary King is Erdős5, because he has a collaboration with Jeff Gill (Erdős4), via a collaboration with George Casella (3), with Diaconis being the eventual link to Erdős. Then again, Gary has coauthored several times with Andy Gelman (who is Erdős3 via a roundtable/symposium with Radford Neal, and Neal via Persi Diaconis), so Gary is actually Erdős4.
Jan de Leeuw (4), Jim Heckman (4), George Box (3), John Tukey (2), Egon Pearson (3), Don Rubin (3), Kolmogorov (2, via Rényi), and Albert Einstein (Erdős2)…!

