H-index and Erdos number
There must be some mistake here. Professor Smith, of PhD comics has out-performed me by a factor of 3 (papers in Science or Nature) to 17 (PhDs graduated), by every criterion except the H-index of citation impact, where I hold a slight lead, 23 to 19. In other words, 23 of my papers have been cited 23 or more times, so far. And my lead seems to be increasing.
I don't think my H-index is unusually high, so maybe Smith's is unusually low. Perhaps, if he treated his students better, they'd write better papers together?
Also, where's his Erdos number, the degrees-of-separation formula that inspired this XKCD cartoon? Mine is 5, via T.R. Sinclair, R.H. Rand, H.D. Block, and P.C. Rosenbloom. The first two links are via papers in nonmathematical journals, though. I'd be more interested in my W.D. Hamilton number, anyway. Incidentally, Hamilton's H-index is only 15, so maybe it's not such a reliable measure of scientific impact after all. Other approaches to citation analysis have been developed, including "eigenfactors."