An analysis of my freight equipment, by manufacturer

[wcfn100]

I do admit to one area of challenge, and this covered hoppers....about the only car that *commonly* appears both in general manifest freights and also unit trains.   So accounting for those is always tricky.  Applying the formulas straight out tends to give me a bloated number, but then when I consider spotted cars (also above average probably given what I model) and the chances of two unit grains passing each other (a very common occurrence) then I think maybe it's not so bloated.

I did think the other day that one thing I like about the era I model is not having 'unit' type trains.  That's a lot of resources into basically a single trains' worth of cars.  With that said, I have my own numbers buster in the form of all the interchange cars cars needed for Waterloo.  Within a roster of 300, I'm trying to get 60-80 cars that can be used in some form as interchange cars with the Waterloo Railroad, Illinois Central and the RI (to a lesser extent), plus there's the local switching.  So again, I'm faced with the situation, as an example,  where I'm going to want 4-6 cars just for tractor loading (half loaded, half empty) when in the big scheme of thing, 1-2 cars fits better.

Jason

[conrail98]

I do admit to one area of challenge, and this covered hoppers....about the only car that *commonly* appears both in general manifest freights and also unit trains.   So accounting for those is always tricky.  Applying the formulas straight out tends to give me a bloated number, but then when I consider spotted cars (also above average probably given what I model) and the chances of two unit grains passing each other (a very common occurrence) then I think maybe it's not so bloated.

Even having no unit trains, I'm still finding this hard when looking at my roster. There was a thread on the Card Card Yahoo group the other day asking about the AAR code, LO, and someone brought up they expanded it to include number of bays. So you'd have LO2, LO3, LO4 for the different # of covered hopper bays on the cards. I'd probably think of taking it a step further and have it like LO3P or LO4G for plastics or grains, respectively. Give some kind of letter designation on the end of it to indicate the type of commodity in that covered hopper,

Phil

[bbussey]

Jason,
Keep in mind that the segment of railroad you are modeling also plays a factor. Otis AFB is in the Cape Cod area if I remember correctly, which is not only off the Shoreline route but way east of my eastern-most modeling point (New London). Segment of road is more important than total traffic contents overall.

However you bring up a valid point regarding the stock cars. Sometimes, you will have to make some exceptions to have certain desired models on the layout. Given that on any specific day, the consist makeup can vary from the norm, I don't think 5 cars as opposed to the standard (by ratio) 3 cars is a deal-breaker. The fleet overall still would be representative of the prototypical equipment makeup.

[wcfn100]

Jason,
Keep in mind that the segment of railroad you are modeling also plays a factor.

I always do.   

BTW, thanks for this thread.  It got me looking at my collection and I've got 40 more MTL cars to go on the block. 

Jason

[Denver Road Doug]

AAR code, LO, and someone brought up they expanded it to include number of bays. So you'd have LO2, LO3, LO4 for the different # of covered hopper bays on the cards. I'd probably think of taking it a step further and have it like LO3P or LO4G for plastics or grains, respectively. Give some kind of letter designation on the end of it to indicate the type of commodity in that covered hopper,

Phil, that's a really great idea.  Worth the price of admission to this thread, maybe twice that!  Seriously, thanks I will likely end up using that.

I'm still trying to grasp how operations is going to go down on my next layout. (which won't exist for a few years yet, but I'm a planner...  )    The main n-scale layout I operate on does a "one-to-one" type of deal, where the train is basically stacked with cars that mimick the spots.   So *for the most part* it's intuitive and he has enough work to keep the crew busy even with the "shortcut". (done due to the difficulty of reading car numbers in N, and obviously not having car cards)   I'm inclined to go that route, but the prospect of car cards with photos of the cars on them to help identify (using weathering or graffiti or other visual cues besides the numbers) the rolling stock is intriguing.

[bbussey]

I always do.   

BTW, thanks for this thread.  It got me looking at my collection and I've got 40 more MTL cars to go on the block. 

Yeah, but it feels good to unload non-prototypical models. 

BTW, even the 11 tanks are too many for me technically, but 3% instead of 2% isn't noticeable and I still get to have the tanks I want to have present. Also, my multitude of home-road 40' boxcars is subdivided both by bodystyle (XIH, PS1, and two AAR types) and schemes (script, St of ME, red-orange New Image, and two black New Image) as well as detail variations (brakewheels, roofwalks, trucks) which removes the effect of one monolithic block of cars dominating the fleet.

[ljudice]

Is it any wonder that ads for OCD drugs are showing up at the top of this page?

  

[bbussey]

An unforeseen positive of this approach is that it helps you not buy every single "shiny new object" that comes to market. I've been recently unloading product from 30 years of accumulating and it is astounding how much there is. So focusing on a set roster configuration results in major budget relief as well. This holds true for motive power and passenger equipment also.

[Denver Road Doug]

An unforeseen positive of this approach is that it helps you not buy every single "shiny new object" that comes to market. I've been recently unloading product from 30 years of accumulating and it is astounding how much there is. So focusing on a set roster configuration results in major budget relief as well. This holds true for motive power and passenger equipment also.

That's pretty much exactly what got me started me down the path of my OCD freight car research.   I didn't have money, but I did have time so I could go through and whittle down exactly what was appropriate and what was not and could be culled. (and consequently lay the groundwork for new announcements)   And naturally I slid my era a little and had to go back and review everything.   But to me it's not only money/space savings but it's a lot of fun doing the research.

[CVSNE]

A few years back a few of us in the Central Vermont society took it upon ourselves to tabulate the total number of boxcars going through White River Jct, VT over a 4-5 day period in 1954. (The period and locale were chosen because we had relatively complete train lists for that period of time). We also figured it would help us determine which cars we needed to have and what percentage by railroad we needed.

The sample total was 3,605 cars. There were about 60 or different reporting marks represented (basically, name a North American railroad of the time and it was on there...)

By far the most common roadname, with 50% of the total, was CN. (no suprise there).
The remaining 59 reporting marks totaled either much, much fewer cars - or so few they were insignificant.
And the percentage of the total by roadname didn't reflect the percentage of the national fleet, although they do seem to reflect some regional "bias" (greater percentages of New England/Northeastern region road names, but not by much). I was especially shocked at how few NYC and PRR cars (based on the % of these roads rosters compared with the national fleet at the time) appeared in our sample data.

Here's some of the numbers:
#2 GTW (5% of total)
#3 NYC (4.4% of total)
#4 PRR (3.1% of total)
#5 CV (2.5% of total)
#6 NH (1.9% of total)
#7 B&M (1.8% of total)
#8 B&O (1.7 % of total)
#9 SOU (1.6% of total)
#10 CP (1.6% of total)

Not sure what I learned from this exercise, except that out of a fleet of 100 cars fully half should be CN, with almost any other road name represented, but not too many of any one road. The thing is, if one were to model a roster to these percentages and then compare the resulting trains to prototype photos, the result may be defendable as statistically "authentic" but still wouldn't look right!
Marty

[wcfn100]

The remaining 59 reporting marks totaled either much, much fewer cars - or so few they were insignificant.
And the percentage of the total by roadname didn't reflect the percentage of the national fleet, although they do seem to reflect some regional "bias" (greater percentages of New England/Northeastern region road names, but not by much). I was especially shocked at how few NYC and PRR cars (based on the % of these roads rosters compared with the national fleet at the time) appeared in our sample data.

That's great info Marty and something people should really let sink in.

On my list of 10,000 cars over about a month, I have 240 different road names, but once you get past the top 20, everything else is less than 2%.

The only thing I can say about your list is that's a pretty small sample of days.  If it were extended out, some of the 'insignificant' cars could show otherwise, especially depending on what they are hauling.  Something like a certain road hauling Dry Milk or whatever may only show once on your list, but could be every 3 days (and you caught it in the middle) on a larger scale which could quickly move it up the list of importance.

Jason

[Ed Kapuscinski]

The thing is, if one were to model a roster to these percentages and then compare the resulting trains to prototype photos, the result may be defendable as statisically "authentic" but still wouldn't look right!

And therein lies the issue with the statistical approach. Rail traffic is not a random thing. Cars move in paths and patterns, and those things are not equally distributed across an entire railroad.