I find that any calculation that would result in a number smaller than 0.000001 either rounds up to 0.000001 or down to 0. Since numerics are accurate up to about 10^15, what gives with the fairly limited decimal portion?