26 Sep
2003
26 Sep
'03
6:28 p.m.
On Fri, 26 Sep 2003, Dan Hoey wrote:
John Conway <conway@math.princeton.edu> writes:
... I presume 7/32 beats 5/24, but don't have time to check.
Yes of course it does [...]
Let's just do 2^2n:3. The order spectrum of that is 1 , 2^2n - 1, 2^(2n+1), and the squares are precisely the elements of odd order. The argument I gave shows that the number of "bad" triples is (2^2n - 1) x 2^(2n+1), as against a total number of product triples of 2^4n x 9, making a proportion of (2^2n - 1)/(9 x 2^(2n-1)). JHC