Your simulation did not contain enough data to be considered satistically significant most likely.

It is interesting that you would try something like this though.

Something like this, only on races with a certain predefined tote distribution based on some number theory, that would be very interesting.