Hacker News Reader

That's a feature, not a bug, I believe. Results that contradict the dominant theory are in general much less likely than one's that confirm it or extend it, and as such will have much lower odds. As a result, if you have significant evidence that such a result is true, then there is a large upside in betting on it, enough to outweigh the unlikelihood.

View on HN

That's interesting. You're looking at it from the point of view of the bettor. My concern is just that non-dominant lines of inquiry are already intensely selected against and this just institutionalizes that bias. But I see your point that the lopsided upside incentivizes contrarians. That's very interesting .. I have to think about that!

[Edited as enlightenment dawns ...]

That's as it should be, though. Extraordinary claims require extraordinary evidence, and studies should be regarded more skeptically the more surprising their results are.

Reply on news.ycombinator.com