Re: Freight car distribution


--- In STMFC@..., Bruce Smith <smithbf@...> wrote:


You are basing your analysis on the incorrect assumption that most
NYC cars were loaded on NYC rails. Those NYC boxcars you see on the
PRR in WWII could just as easily have been loaded on PRR rails, or NP
rails or AT&SF rails, etc... as NYC rails. Boxcars were basically in
a national fleet for most of the steam era. During WWII, when many
of the car service rules were suspended for other classes, those
classes also operated in a national pool, most notably flats and
reefers. Ultimately, those of us modeling WWII really have it quite

Boxcars - national pool - model the fleet (with a slight home road
bias, as these cars did get sent home for repairs etc)

I find this hard to believe, but I am a relative novice.

Three separate issues:

1) If a freight agent in LA has 2 shippers requesting 40' XM MT's, one
load destined for Syracuse, and the other for Harrisburg, and he has a
bunch of empty 40 foot XM's - including one PRR, and one NYC, would he
just flip a coin? I've got to believe he would send cars towards their
home roads, national pool or not, even during WWII. That should bias
the movements towards home road cars. Wouldn't a PRR freight agent in
Harrisburg deliver an MT PRR 40' XM to a shipper with a load for
Pittsburgh, and deliver an MT ATSF 40' XM for another shipper's LA
bound load? Or do the agents not know the load's destination, only
that a shipper needs a 40' XM?

2) I'm completely uncertain about this and looking for guidance, but
if a NYC box car out of New England and destined for southern Illinois
hit a NYC yard in New Jersey, is it more likely that the NYC yard
clerks would have it routed over the NYC's water-level route, rather
than to the PRR, which could get it there over a slightly shorter
distance? If it was a PRR car on the same route, would the NYC clerks
route it down to Greenville for the PRR to deal with if they were
overloaded with traffic, rather than try to claim the mileage revenue?

3) There is still the issue of statistical variations. I was once
responsible for buying a phone system for the medium size engineering
firm I worked for (the curse of being a network literate engineer in
the early 90's). Once installed, what amazed me was the in-bound phone
call patterns. Most calls were direct in, so our receptionist only
received 30-40 calls per hour, and the average duration was under 10
seconds. Yet there were several times, EVERY day, when she would have
to handle three calls at once, and occasionally four (which affected
programming the receptionists console), and then be idle for the next
5 minutes. These excursions from the "average" rate were quite
significant, frequent, and initially surprising.

A little research found that Bell Labs/AT&T spent a ton of money
developing the operations research for modeling these sorts of
problems (it directly impacts the size of telephone systems). Turns
out the deviations from average are often significant for many
processes, especially if they involve processes that are not purely
random. This is the sort of phenomena that Mike Brock has observed in
his reports - LARGE deviations from the average may be the NORM, not
the exception.

Hence, unless someone operates their layout every day with enough
traffic to generate statistically significant traffic volumes, I don't
think our wheel reports need to be that close to annual averages of
the nationwide fleet. Obviously some movements could defy belief
(string of Rutland box cars in southern California, or ATSF boxcars
ALWAYS outnumbering NYC boxcars in Harrisburg), but I'm not sure we
need to be that close to the nationwide fleet averages to still seem

For a 400 box car fleet, statistically over 60 reporting marks would
represent a national average and would include one each of WM, D&H,
and WP. 90% of the fleet would be from 35 reporting marks, and 50%
from just 10 railroads, including CP and CN. On a PRR layout centered
around Altoona I think it is reasonable to see two D&H cars go by out
of every 400 boxcars, and maybe even two EB loaded WP 50 footers in
one train (although not very often). I think it will be ok if I only
have a few CP and CN cars, even though, statistically, they
represented 12% of the North American fleet (that would be 50 out of
400 boxcars). Only having one or two of the required six 36 foot
Southern boxcars would probably be OK too. (I still haven't seen a
picture of one of these cars on PRR rails, but they were 1.7% of the
US boxcar fleet during WWII)

I think considerable latitude is possible. It would be interesting to
take wheel report data in 400 car blocks, and look at the statistics
of each block, to determine how much variation from the "average" is
normal, and reasonable, to model.

Dave Evans
PS - I haven't reviewed all of Tim Gilbert's posting in the message
archives, but I have not come across large amounts of data in the
posts I have reviewed. I have found data summaries. I'm still
relatively new to this group - am I missing something in my archive

Join to automatically receive all group messages.