Are There Different Rarity Factor "Systems"?

Rarity Factor is often discussed in relation to cancellations and covers. The Unitrade Catalogue gives RF ratings to a small number of early issues and they are rated on a RF 1 to RF 10 scale. I've seen some posts here in the forum which refer to a RF 2.5 to RF 250 scale which seems a bit excessive to me. How could you possibly rate the difference between a RF 127 and RF 128? Maybe I misunderstood the scale and I don't recall any "real world sightings" of this system.

Many listings for Canadian covers and cancels on ebay and other auctions use a rating of RF A through RF E. Is there an "official" source (similar to the Unitrade ratings) which helps to list and define how these ratings are applied? Or is a rating just someone's independent assessment?

US perfins A-to-F perfectly useful
US Flag Cancels 1-to-100 overkill
US precancels used to be dollars up to 90 then C,B,A,
but recently replaced with dollars up to 300 then A & AA (one only known)

There is no standard "Rarity Factor" scale. The author of each of the specialized catalogues derives their own RF system and normally explains it in the catalog. For instance, Moffatt and Hansen use rarity factors from 2.5 to 250 in their Squared Circle catalogue while the RPO catalogue of Lew Ludlow uses RFs from 5 to 500. In order to give credibility to the system used, the collector needs to understand that particular system and how it was derived.

Agreed. And I guess that's my question. Specifically, my interest is in older (late 19th and early 20th century) British Columbia cancellations and PO marks. Some covers I've looked at are labeled RF C or RF D. Just trying to understand how these ratings were applied.

Why not just ask the seller of the material where the rating came from. Which book, reference or catalogue was it from and how does that individuals system work or what does it compare to?

The rarity factors for British Columbia cancellations comes from:

"British Columbia Post Offices" edited by William Topping (1991).

In the introduction, the rarity factors are defined as:

A - presently open, very common
B - open, but harder to find, or closed but quite common
C - closed - found in most collections
D - closed - harder to find, but not rare
E - closed - rare, usually less than ten known
U - unreported to 1991-06-30
* - unreported except in proof book.

These rarity factors are the same for all books in this series (so Alberta, Saskatchewan, Manitoba, Territorial Post Offices).

Jan

Here's the Rarity System developed by members of the Vermont Philatelic Society for Discontinued Post Offices in Vermont. The system was initially developed in 1994, revised in 2006 and may be revised again in the next few years.

In 2006 the Vermont Philatelic Society printed a booklet with all the discontinued post offices and rarity number for each member of the society.

However, the value of many of the rarer items has decreased in recent years as the number of active collectors has decreased.

2 = plentiful < $5.00
3 = common $5.00 - $10.00
4 = uncommon $10.00 - $25.00
5 = scarce $25.00 - $50.00
6 = very scarce $50.00 - $100.00
7 = rare $100.00 - $300.00
8 = very rare $350.00 - $1,000.00
9 = unique $1,250.00 - $2,000.00
10 = unreported

The freely downloadable Queensland Steel Datestamps (Toowoomba Stamp Club) has the most interesting scale, in my opinion. Their "ratings" are:

C-Common
S-Scarce
R-Rare
2R-Rarer
3R-Rarer Still
4R-Rarest

They freely admit that this is based on a "feeling" and that they are conservative based on new discoveries being constantly made.

Thank you, Jan. I think this is exactly what I was looking for. This is the rating system I have been coming across in the pieces I have looked at.

Re the Vermont listing

However, the value of many of the rarer items has decreased in recent years as the number of active collectors has decreased.

Using a value seems very a strange way of calculating a RF - something is rare or not does not depending on how may people want it. It also do not take account of inflation. In that case it would be much better to just drop the $ sign and use the numbers.

Or even better just to publish the raw data listing the number of items known from a given sample size and the date that the list was created
.

AQ