[seqfan] Re: Probability of identical sequences

[David Radcliffe]

There is an interesting example in the OEIS. The sequence A078608, defined
by a(n) = ceiling( 2/(2^(1/n)-1)), coincides with floor( 2*n/(log 2) ) for
all n from 1 to 777451915729367 but differs at 777451915729368.

Of course there is a reason that the sequences agree for so many terms. It
is not merely a coincidence. But this does not prove that they always
agree. I would be wary of probabilistic arguments in such a case.

- David