Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: Notesfiles; site hpcnoe.UUCP
Path: utzoo!linus!decvax!genrad!mit-eddie!think!harvard!seismo!hao!hplabs!hpfcdc!hpfcla!hpcnoe!jeff
From: jeff@hpcnoe.UUCP
Newsgroups: net.games.board
Subject: Re: Re: Seattle Rules Rail Baron
Message-ID: <37800028@hpcnoe.UUCP>
Date: Fri, 20-Dec-85 13:54:00 EST
Article-I.D.: hpcnoe.37800028
Posted: Fri Dec 20 13:54:00 1985
Date-Received: Tue, 14-Jan-86 07:49:18 EST
References: <270@tekigm2.UUCP>
Organization: 20 Dec 85 11:54:00 MST
Lines: 27

I posted the following:

> The  original  rules  destination  chart was  designed  so that the most
> popular cities and areas  corresponds to the "real world"  popularities.
> The Seattle rules does not have any consistency.  For example, Pocatello
> has the same probability as Seattle (12.5%) but Billings has a much less
> (4.2%).  This boosts the importance of UP and lessens that of SP.

I should have clearified, the probability quoted is the probability of
getting that destination once the Region has been chosen.  The reason 
that I didn't post "full probabilities" is that I believe that the
idea of P(region)*P(city given region) is inaccurate.  The above formula
ignores that if the region thrown is the same as you are in, then you
get to pick a region.  When you get to pick a region, than it screws up
the probability because what you really want to know is what is the
probability when you have no choice in the matter.  A second order
approximation would have the formula:

    (P(region) - P(in region)*P(region)) * P(city given region)

It subtracts out the number of times when you get a choice.  Which is
the probability of being in a region (approximate P(region)) times the
probability of throwing that region (P(region)).  You can refine this
more by saying that the probability of being in a region is not P(region)
because of the choice factor, and so on and so on...

-- Jeff Wu