use Math::BigApprox 'c'; my $bits = 2**32; my $inserts = 1e5; print 1-(c($bits-$inserts)^c($bits-1))/c($bits)**$inserts, $/;