I wrote:
> [log in to unmask] writes:
> > Back to QCTerm, Wirt.  The answer is 15 times .5**6, or .234375.
>
> Seems reasonable, but I just flipped 26,000,000 coins and ended up with an
> empirical value of .1226 or thereabouts.  Maybe I just can't count, or
> there's something wrong with my coin.

Stan and I just independently determined that there are 128257 of the
possible 1048576 cases of flipping 20 coins that include at least one run of
at least six 1s, resulting an a probability number of 0.12231541... (or
exactly 128257/1048576).  Sooooo, at least some of us are really wrong
somewhere :-)

Interesting problem.

G.