Anecdotal/empirical evidence
Which is insubstantial when verifying the success of a script.
I can run a script 2 times, get a positive profit on the second run and claim a 100% chance of success.
My sample size contains only that very last one. It's anecdotal evidence.
Excellent point. But if I profit 100 times for myself and 100 times for investors, even if it is against math, I'm happy. And it is not against math - check out combined probability.