I remember reading an article here a while back about a fair protocol for making a bet when we disagree on the odds, but I can’t find it. Anyone remember what that was? Thanks!
Assume there are n people. Let S_i be person i’s score for the event that occurs according to your favorite proper scoring rule. Then let the total payment to person i be
Ti=Si−1n−1∑j≠iSj
(i.e. the person’s score minus the average score of everyone else). If there are two people, this is just the difference in scores. The person makes a profit if T_i is positive and a payment if T_i is negative.
This scheme is always strategyproof and budget-balanced. If the Bregman divergence associated with the scoring rule is symmetric (like it is with the quadratic scoring rule), then each person expects the same profit before the question is resolved.
edit: no, I don’t think that’s it. I think I do remember the post you’re talking about, and I thought it included this anecdote, but this isn’t the one I was thinking of.
I remember reading an article here a while back about a fair protocol for making a bet when we disagree on the odds, but I can’t find it. Anyone remember what that was? Thanks!
From the Even Odds thread:
Assume there are n people. Let S_i be person i’s score for the event that occurs according to your favorite proper scoring rule. Then let the total payment to person i be
Ti=Si−1n−1∑j≠iSj
(i.e. the person’s score minus the average score of everyone else). If there are two people, this is just the difference in scores. The person makes a profit if T_i is positive and a payment if T_i is negative.
This scheme is always strategyproof and budget-balanced. If the Bregman divergence associated with the scoring rule is symmetric (like it is with the quadratic scoring rule), then each person expects the same profit before the question is resolved.
http://lesswrong.com/lw/hpe/how_should_eliezer_and_nicks_extra_20_be_split/ ?
edit: no, I don’t think that’s it. I think I do remember the post you’re talking about, and I thought it included this anecdote, but this isn’t the one I was thinking of.
edit 2: http://lesswrong.com/lw/jgv/even_odds/ is the one I was thinking of.
Great—thanks! (Thanks to badger below too)