Football (soccer) definitely belongs on this list for me. It looks "easy" with all the stats/data out there, but the variance is brutal - one injury, bad weather, or overrated form streak against weak teams can flip everything.I've been building a side project (Betora) that tries to show clear probabilities + explanations (e.g., how weather quietly disrupts passing teams more than odds adjust, why old H2H is useless, injuries often underpriced). Mid-70% overall accuracy, but past predictions stay public (including losses) - no hiding. It's not about "best bets," just decision support.Curious if anyone here uses probability tools like this to filter matches or spot edges in football, or if you avoid them completely because of the variance?(Full disclosure: I made it. If interested: Betora | AI-Powered Football Predictions:Ad REmoved by admin)