I'll hang out the shingle here for folks who want to talk about sea ice wagers with me. Keep in mind that I don't bet real money, quatloos instead. Call them points of honor.
I don't take any of this very seriously. Seriously enough to mention that so far, I have won all my wagers of quatloos.
The serious part is how rapidly pseudoskeptics leave the room if one raises even a casual honor bet that takes their statements seriously. People who claim that temperatures are just as likely to go down as up demand 50:1 odds if challenged to a bet on whether temperatures will go up or down. If they believed what they were saying, they'd take 1:1 odds, and be happy with either side of the bet.
Last fall as I was evaluating the seaice ... statements from Wattsupwiththat , Joe Bastardi declared (over there) his belief that September 2011 ice would be at least 5.5 million km^2. If anyone can find Bastardi's address, I would be happy to bet quite a few quatloos with him. If he believed what he said, he should pay off infinity to my 1 if ice is less than 5.5 million km^2. I'll pay him 1 if it's over.
"People who claim that temperatures are just as likely to go down as up demand 50:1 odds if challenged to a bet on whether temperatures will go up or down. If they believed what they were saying, they'd take 1:1 odds, and be happy with either side of the bet."
ReplyDeleteI don't know if I buy that. I'm risk averse: I would probably stay away from a 50/50 bet without some pretty good odds - 5 to 1 or so. It might not be a rational use of expected value analysis, but then... a really good expected value analysis takes into account irrational reactions. I kind of like the approach (I can't remember who right now - Stoat? Tamino?) of having a "neutral zone" with no winner, where that neutral zone would ideally be the intersection between my likely range and my opponent's likely range. So I'd only lose with sea ice so high that I considered it very unlikely, and my opponent would only lose with sea ice so low that they considered it very unlikely.
-M
M:
ReplyDeleteI think you're considering real money, rather than quatloos. For real money, I don't make bets, regardless of the odds.
It is Stoat who introduced the neutral zone. He did so after losing a bet with me :-) It was a matter of some hundredths of a million km^2, perhaps only one. But the number was on my side (low) of the critical.
Introducing the neutral zone doesn't fundamentally change the computation. What we're still, in principle, doing is saying that the odds we're willing to accept are related to how much more likely we think something is than the other person.
This relates to Alastair's comments as well, and I see from your comment and his that this warrants some more discussion. New post to come tomorrow (22 July 2011).