That is always a risk, and given I 'only'* had 48 warriors to look at (Jan - August), certain things cannot be said with any certainty--if only two warriors had a given legacy, it would be dangerous to make generalizations about the deathfulness of that legacy, particularly if I wished to control for race and cabal. If I get approached to hero-imm someday I'll request a grep and put together a more comprehensive study, but for now, although Autumn Harvest may or may not be statistically good for rager smurfs, I can't say anything conclusive for that one taller dude who took it or the svirfnebli who chose against it.

That said, some statistically interesting things appeared that really cause me to question some people's pet theories about balance and play behavior.

Do I hope to provide any answers with this study? Not really, but maybe, MAYBE it will help phrase the correct questions to be asked during game design and testing.



* Reading through all those warrior PBFs seemed like a helluva lot of work, not mentioning entering 48 cases x 49 variables.