This page was inspired by the corresponding one of Martin Bohner!

One defines by mathematical induction: The Erdös number of famous Paul Erdös is 0. The Erdös number of any other person X is n+1 if there is Y whose Erdös number is n and X and Y have published a common paper, and if in addition no co-author Y of X has Erdös number less than n. Check that it is possible for a person to have no Erdös number at all!

Proof:

1. Erdös, P. and Lorentz, G. G: On the probability that $n$ and $g(n)$ are relatively prime, Acta Arith. 5 (1958) 35--44.
2. Jurkat, W. B. and Lorentz, G. G: Uniform approximation by polynomials with positive coefficients, Duke Math. J. 28 (1961) 463--473.
3. Balser, W., Jurkat, W. B., and Peyerimhoff, A: On linear functionals and summability factors for strong summability, Canad. J. Math. 30 (1978), no. 5, 983--996.

My Erdös number is at least two, since I do not have a joint article with Erdös. Very likely, it is equal to three, but I have not verified this! It is surprising (or is it?) that very many colleagues in Analysis have Erdös numbers that are at most three!!!

Mathematical Reviews now even has a tool for computing the collaboration distance, in German Abstand zwischen Autoren, between any two mathematicians! An extra button allows for computation of your Erdös number. However, be aware that this tool only takes into account articles that have been reviewed by MR. As an example: According to this tool, my collaboration distance to C. F. Gauss is at most equal to 5.