In a message dated 4/16/2004 8:31:03 PM Central Daylight Time,
CBarkan@... writes:
You guys are assuming that a particular type of car appears randomly and
independently in a train (or in a days worth of trains). There are probably
reasons to think that is not the case.
Yep.... Not to bore you with my little project again: the Glass Plant traffic
in Corning....

Ah, you heard that story already...

My conclusion is that theoretical data does only get you so far. When it
comes to YOUR prototype location, the only thing you can really rely on are
photographs, and actual records for that line and location...

Even if I would average over all plants in the Corning environment, I would
make serious "mistakes", since each plant has its own pattern. Sometimes I
catch myself in analysis paralysis...

Especially since I only have half the data: only incoming materials, not
outgoing product.
Now one could assume that the clean box cars bring in material could ship
boxes of product out, but what about covered hoppers? I'll apply the statistics
rules there...
Or more likely, the "what's in the yard? I.e. my boxes..."

Martin

Martin