On 25 November 2015 at 15:22, Alan Burlison <alan.burlison at gmail.com> wrote: > > Hmm, do you have any feel for how long touches have to be before a > significant proportion of them start being false? I'd say 5041 is a good upper bound. :-) Martin