[FlFishOn]

" Statistically the underdog on these types of hands should win about 27%. To date i've observed and recorded 271 such hands and the underdog has won 100 or 37%, 10% higher than they statistically should."

Two points. 27% vs 37% is not 10% different, it's 37% different (higher). This is big. If maintained you need very little data to prove your hypothesis.

Secondly, by grouping all these dominated hands together you build in some error, maybe as much as 5-10%. To do it right you need to calculate each exact match-up, a ton of work.

Where did/do you play to gather this data?