home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!pipex!bnr.co.uk!uknet!mcsun!sun4nl!tuegate.tue.nl!rw7.urc.tue.nl!wsadjw
- From: wsadjw@rw7.urc.tue.nl (Jan Willem Nienhuys)
- Newsgroups: sci.math
- Subject: Re: probability = 0
- Message-ID: <6830@tuegate.tue.nl>
- Date: 6 Jan 93 10:24:38 GMT
- References: <9301051655.AA22441@aplpy.jhuapl.edu>
- Sender: root@tuegate.tue.nl
- Reply-To: wsadjw@urc.tue.nl
- Organization: Eindhoven University of Technology, The Netherlands
- Lines: 37
-
- In article <9301051655.AA22441@aplpy.jhuapl.edu> louis@aplpy.jhuapl.edu (Louis Vasquez) writes:
- >Hello,
- > My friend and I got in a debate about the probability of
- >something happening. The event was that another friend of ours
- >could win the superbowl. Not that he would play on a winning team
- >in the superbowl but that he would win it.
-
- [rest deleted]
-
- Strictly speaking, there's hardly a way that you can define the probability
- of such an event. If you want to discuss the probability of a certain
- *type* of event, you must have some kind of well-defined universe of
- possibilities and some kind of known or plausible chance mechanism.
- If the situation leading up to the "event" can be experimentally
- reproduced (in principle) then you would have (in principle) a way
- to check whatever you say about the probability.
-
- In most cases such "probabilities" only are defined in suitably chosen
- models of the situation. So the throw of a coin is simplified to
- *postulating* that both sides have equal chance. The possibility that
- the "throw" consists of initializing the coin as spinning very rapidly
- around a vertical axis on a very smooth surface is left out from the
- model. Similarly the unequal mass distribution (due to the different
- numbers of pips) of a die is left out in the model.
-
- However in the case of the poster, where hypothetical changes of
- rules are envisaged, there is no clear model. I'd say that for
- all practical purposes the chance is 0. If you were to bet, then
- 1 cent would be the lowest stake. But there is also a highest stake,
- namely the amount you can credibly raise in case you lose. I guess
- that in each credible model of the situation, the chance of the event
- is much less than the ratio between the lowest and highest stake.
- So from a theoretical point of view there's no probability (because
- a clear model is lacking) and from a practical point of view (rational
- betting) the probability is zero.
-
- JWN
-