So close yet so far

1) As your said your first 5 are giving are giving worse performance, so picking up 10 would increase your returns and not dilute them.

2) Nope, random implies no pattern to your picking. With enough sample size, your edge, if it exists, will converge reasonably to what it should be. Any patterns you seen in randomness is your own bias. Its entirely possible that you don't have an edge, in that case.
Run the numbers and find out ...
So from 500 to now 10 stocks there definitely is a pattern to picking them.....but you mentioned from the 10 pick 5 out randomly right?