View Single Post
  #94 (permalink)  
Old 19-November-2005, 08:27 PM
grant hutchison grant hutchison is online now
Order of Kilopi
 
Join Date: Jul 2005
Posts: 7,599
Default

Quote:
Originally Posted by Ken G
... your formula tells you that when you encounter the $10 subset, you should pay $10. But as soon as it does so, you no longer have any expectation of breaking even in that particular subset of trials! This last remark is crucial.
But only if you believe that this subset of trials has any relevance to the real world. I say it hasn't.

Quote:
Originally Posted by Ken G
Someone says, "let's have all the people who were among the first 5% born to one corner of the room." Of course there's going to 1/20 of the attendees in the corner. Now they say "OK, can I have every being that was the 10 billionth born to raise their hand". Why would you expect 1/20 of those people to be in the corner? Generally, that would not be the case!
I'm very happy with that. I can't see why there should be any correlation between these two groups. If most populations are very long-lived, then a lot of those with the 10-billionth birth number are going to be in the corner. Carter doesn't care, so long as 5% of each population was wrong in its assertion that it was in the last 95% of all births.
I think you're chasing some sort of symmetry that doesn't have to be there for Carter to work.

Grant Hutchison
Reply With Quote