There are some interesting nuances here that deserve consideration… While I agree overall that if one focuses on the oddballs, that is problematic. However… on average, oddballs are quite common. What I mean by that is that the number of unique cars or occurrences is such that you might see something like one per train, on average. The issue with modeling that, it that you don’t want to see the same one on each train! So, if you have the luxury of having storage space, an alternative concept is to have a box or two of “oddballs” that each get seen about once a year (about the memory span of most operators) to continually rotate onto and off of the layout. For me, these might be boxcars, hoppers, or gondolas from fleets that constituted a minor percentage of the total and fail to make the minimum fleet size cut-off for my “fleet”.

As for foreign stock cars -  These were not unusual on the PRR and thus should appear with a regular frequency in PRR stock trains. These may also appear as small shipments of breeding stock at the head end of expedited trains

Regards,

Bruce

Bruce Smith

Auburn, AL

     I have the paperwork somewhere, however no access to it at the moment -- about an AAR directive ordering the B&O to send 25 (IIRC) stock cars west of the Mississippi to relieve temporary shortages in western territories.

     One has to apply Thompson's Corollary* here I would believe.  I had stated on this list many years ago that I was a member of a club layout in the Twin Cities where some of the members would bring in old Trains magazines with the "There's a prototype for everything column", and then proceed to
model it on the layout.

      Even at RPM's I still hear that phrase from the attendees in presentations and having experienced first-person former club members use that phrase to justify their aberrant modeling, it always used to rub me the wrong way to the extent of leaving that club as Rule 4** was continually violated and I realized
that no matter how congenial these guys were, they were not prototype modelers, just toy trainers. 
                                                                                                                                                                         James Dick - Roseville, MN 

 * I would add Thompson's corollary (Tony Thompson, that is): Model only what's plausible. Never model anything that causes other knowledgeable modelers to raise their eyebrows or requires that you be constantly explaining it, even if you have irrefutable evidence that it actually occurred. Richard Hendrickson
** Rule 4 (Nehrich's Rule): A layout filled with exceptions does not an exceptional layout make.