But all games have a rate of refunds, including refunds due to rage quitting. Our null hypothesis should be that the refund rate should be the same as any other game (controlling for difficulty).
I don’t know what a normal rate of refunds is, that is probably data that is out there. But if the refund rate for this game is significantly higher than normal, then we can say that SOMETHING about this game is different.
If we control for every factor besides hltb, and it’s still higher, only then could we say with high confidence that the difference is due to people finishing and refunding. But we technically can’t ever say that with 100% certainty unless we go to every person’s house and watch them do it.
You’re just making me wanna play this game lol