To be fair, betting isn't about being right every time either. To really make bank on a site like DraftKings you need to be good at the meta strategy, not just good at predicting sports outcomes.

Obviously it's not a good idea to require a bet to make a public statement, but I also wouldn't mind seeing scientists be held more accountable for important decisions. How often are study sections evaluated after the fact, for example?

If we're talking public policy decisions, I don't think scientists should be punished when acting on the best available, but incomplete data, especially when two equally reasonable experts within the field can disagree on its interpretation.