Re: "A" Railroads Base Fleet

Anthony Thompson <thompson@...>

Charlie Vlk wrote:
These statisical studies of car distribution assume appearance based on proportion of ownership of cars, which in a perfect free interchange environment might be viable and probably are somewhat useful as a starting point.
Here we go again. Charlie, there WAS a free interchange environment, for CERTAIN car types. To say this fact is "somewhat useful" is like saying "sometimes summer days can be somewhat warm."

However, there were favored connections and bias built into the system because of origin of and destinations of certain commodities and other human and economic factors.
Read previous comment. OF COURSE assigned cars are NOT free runners and cannot fit the Nelson-Gilbert model, and they said so, at the beginning and frequently thereafter. Please don't ignore what the originators of this idea THEMSELVES explained. Otherwise you open yourself to the accusation of not grasping what you are commenting on. No one has or will apply the N-G model to tank cars, etc. etc.
As for favored or hostile connections, yes, it does introduce a small bias. This has been acknowledged repeatedly on this list and hardly bears repetition.

Wheel reports, photos of entire trains, or overall photos of yards for the road you are modeling or similar ones in the region you are interested in, in your era, would be more instructive than a mathematical model.... if you can find such documents.
Yes and no. Yes, a wheel report or film of an entire train does give you a superb snippet of data, and if you wish to model that train or trains, fine. Ditto for yard shots. But the Gilbert-Nelson hypothesis does not deal with those points, but instead attempts to answer, what is the OVERALL pattern over a long time period. To say "wheel reports are more instructive" is like saying apples are better than oranges. Each to his own taste, of course; but please don't muddy the waters.

Tony Thompson Editor, Signature Press, Berkeley, CA
2906 Forest Ave., Berkeley, CA 94705
(510) 540-6538; fax, (510) 540-1937; e-mail, thompson@...
Publishers of books on railroad history

Join to automatically receive all group messages.