Written: May 1, 2013
Larry Shepp was fond of gambling games. Only a few weeks ago, my class and I explored (March 11, 2013) [see Maple code done in class], his deceptively simple, but so intriguing "Shepp Urn", that asks for the best strategy when to stop in a game where you draw blindly, from an urn containing m "minus balls" and p "plus balls", earning a dollar if you pick a plus ball, and losing a dollar if you draw a minus ball. If someone asked me (or Larry), at the beginning of this calendar year, to bet on who would be the next-in-line to go to a better world, neither of us would have picked Larry. And perhaps he is not really gone?
In his wonderful tribute to Roland Dobrushin and insightful essay on the statistical thinking of Tukey and his good friend and colleague, the late Yehuda Vardi, he talked to them as though they are still with us. And even though Larry was far from religious in the conventional sense (he once shocked me when he offered to meet for lunch on Rosh Hashana, that for him was a normal working day), these essays imply that he believed in the immortality of the soul, so let's hope that he was right, and as a corollary, his soul is with us for ever.
Larry was loud, and knew it (see line 35 on the second page (p.269) of the above-mentioned essay), but in such a charming way, and he was the opposite of obnoxious, in spite of his very strong opinions, not only on math, statistics and science, but also on politics, that he passionately loved.
And boy was he brave! He went to Iran and told his hosts that their regime is no good. Way back he went to the Soviet Union (before it became former), and also told them that their communist regime is bad.
And not only with global matters. Larry was such a mensch. Once a graduate student at another university complained to him that his advisor gave him a dissertation problem, that he believed he solved completely, only to find out that he was scooped by his very own advisor, who in a joint paper with his grand-advisor (his advisor's former advisor) completely solved this thesis problem. When Larry asked for clarification from that advisor, the latter replied that "there is no way that the grad student could have proven it", since the problem "turned out to be too hard for (at least this particular) grad student". Larry felt that even if the student's proof is incomplete, it is not right for an advisor to "compete" with his students, and Larry said so (to that advisor) in his inimitable loud and opinionated way. It may or may not be a coincidence that after Larry's confrontation, he got "punished" (by a relative of the above-mentioned advisor, who was in charge of scheduling classes) by all of a sudden having to teach basic courses for undergraduates, something that he never had to do before.
Larry was a brilliant theoretical probabilist, but an even greater applied mathematician, who contributed very significantly to tomography (and was a member of the Institute of Medicine!), but his vision and attitude transcends any narrow classification.