It's also an interesting form of a regressive tax in the form of a confirmation bias in selecting richer debate partners. Maybe the GP is just more well-off th

It's also an interesting form of a regressive tax in the form of a confirmation bias in selecting richer debate partners. Maybe the GP is just more well-off than their interlocutor, who might not make rent this month if they'd lose a $100 bet they're 99% sure they'd win, whereas the GP is only 90% sure, but would find losing $100 a minor inconvenience.