Friday, April 30, 2010

Degrees of crazy

I recently read an article in Skeptic Magazine ("Crazy Ideas 101", Vol. 15, No. 1) that began with a simple and thought-provoking question:

Which is more plausible: That aliens are visiting and abducting humans, or that there is some large creature living in the Loch Ness which caused the Loch Ness monster sightings?

I would say that Nessie is more plausible than UFOs.  And I don't just mean that Nessie is a few times more likely, I mean by several orders of magnitude.  Cryptozoology may be wacky and based on sketchy evidence, but UFOlogy requires at least a couple huge and improbable conspiracies.

But based on polls (if I remember the article correctly), roughly equal numbers of people believe in both these things.  Of course, it's old news that public opinion doesn't always reflect reality.  But the larger point is that people don't distinguish between different degrees of crazy.  There's a category of crazy stuff, and a category of non-crazy stuff.

To some extent, it doesn't really matter.  If the likelihood of claim A is 1 in 1010 and the likelihood of claim B is 1 in 1020, well so what?  Claim A may be ten orders of magnitude more likely than claim B, but they're still both effectively impossible.

But it does matter, because there are some things that are slightly more in the gray area.  For example, I think cold fusion (the claim that nuclear fusion can be caused at room temperature through electrolysis of heavy water using palladium electrodes) is almost definitely the result of experimental error and/or fraud.  But if the cold fusion community ever breaches its way into more legitimate scientific literature, and the results become widely replicated, then that's that.  Then I was wrong about cold fusion.  I would be greatly surprised, but I wouldn't feel betrayed by science or skepticism for having mislead me.  Sometimes the evidence can mislead us, despite our best efforts.  That's why science is hard.

On the other hand, if Autodynamics were proven correct, I'd think the world had gone crazy.  Because E is not mc3.  And 3 m/s plus 4 m/s does not add up to 5 m/s.  That's just stupid.

The point is that skeptics may be occasionally wrong.  So when we're wrong, we better have a good idea of which claims we're most likely to be wrong about.  And we should be able to convey this information in any skeptical analysis.  There's no need to rank individual claims, but it's good to keep in mind how much evidence there is opposing each claim.  But you were doing that already, right?