James C.: << DA> Has anyone listed the largest terms in Hans's data? The GM? . . . . . .

...

Those are good questions, but they weren't mine. (I asked about how the CF expansion of pi compared with a "statistically expected" one . . . and then someone pointed out that dependence among nearby quotients makes this tricky to define. But I'm still curious about the answer, if someone knows a better way to ask the question. And maybe in the limit, it suffices to just examine the frequency of various quotients, anyway. Since I think the effect of one quotient on the distribution of nearby ones diminishes quickly.) --Dan _____________________________________________________________________ "It don't mean a thing if it ain't got that certain je ne sais quoi." --Peter Schickele