Re: [PrimeNumbers] prime +/- 2^n = primes: A tight sequence and a potential c...
> So the chance of each of them being prime is approximately 2/log(N).On further reflection, the correct statement (for the bit-flipping variant)
the chance of each of them being prime is approximately 2*C_2/log(N), where
C_2 = 0.66016... is the usual "twin-primes constant".
See e.g. Chris Caldwell's
So, as N -> oo, the expected no. of primes is 2*C_2*log2(N)/log(N) =
1.9048... And for the +/-2^n procedure, it is just twice this, or 3.8096...,
which is in rather good agreement with Jens' and my results.
[Non-text portions of this message have been removed]