Question on survival estimates and those "R" numbers

I was able to get back to the coin shop today and I managed to get a couple of decent upgrades for my type set. One of them was a pretty buffalo nickel and the other was a capped bust dime. After getting home, I started to look into my current capped bust dime (a very well-circulated one) and PCGS has this particular coin at something like 150 survivors and at R7.

So what I am curious about is the accuracy of these estimates, including that R number. For clarity, it's a 1823 Large E dime that is getting bumped out for a 1837. I'd guess the 1837 might be in the XF range, and the 1823 maybe Fine on a good day.

Anyways is their survival estimate something that might be construed as under- or over-estimated and would that then change the rarity number associated with it? I recall before when I had posted about my California fractional and mentioned that R number someone else had said they saw an entirely different number elsewhere.

Curiosity sometimes gets the better of me so I find myself coming back here and starting a new topic to be short-lived and hopefully walk away a bit wiser. And... so you aren't left reading all of this without seeing any pictures of any coins, here's my new additions!

 

Log in or Sign up to hide this ad.

Check that Buffalo for an RPM. The MM placement looks really close to RPM-009.
http://varietyvista.com/03 Buffalo Nickels/RPMs 1938-D.htm

 

I'll take a look at it, thanks!

 

R-numbers are always estimates. Sometimes they're based on a lot of data, like rigorous surveys of collections, and sometimes they're estimates. The problem with them is that once they're in a book, you can't change them. You can publish an update later, but then you have both numbers out there for people to use.

With VAMs, the numbers were very rough estimates due to the sheer number of silver dollars that there are that couldn't be examined. On top of that, the level of detail used to define a specific variety increased over the years to where the originally published rarity scale was no longer useful. Add to that people incorrectly inferring value from these estimates and the decision was made to no longer use them at all. The study of 1878 8TF is probably mature enough by now that they could be reintroduced for those varieties only, but the bad information is still out there, and that would confuse matters. Relative or ranked rarity is more useful measure.

Don't get me started on the "URS" numbers. I'll start and finish by simply saying it's a bad scale and leave the rant for another time.

 

Well, that was a quick Google search (URS numbers) that I read the result and quickly thought "no thanks!" It seems very similar to what confused me from the beginning, which you put into perspective rather well.

I've used coinfacts a lot and read those little excerpts at the bottom and at least several times they mentioned the era in which they were from. When that bulk of information is 30+ years old, it makes me wonder if those R numbers and survival estimates are also as old and, as you pointed out, could or could not have been updated. I think it's helpful to have information like that available, even if it's dated, but having something like a "last updated" date on it would be helpful, too.

 

In order for the information to be useful, it has to be managed like a population report. A single, comprehensive source updated at predictable intervals. I think EAC might do something like this. Maybe also JRCS and BHNC. Popular early federal stuff fits that paradigm well. Seeing dates attached to individual R numbers now makes you have to wonder if an old number is wrong or stable.

 

To succinctly answer the OP's question: Generally all R values are estimates. The exceptions are for quite rare coins in the range of R-6 to R-8 plus unique survivors. The estimates range from fairly credible to not very credible at all. Depends on the issue and the amount of serious collector attention over time. This is where deep dives into the specifics of collecting a particular issue pays off.

JRCS regularly conducts census surveys of the various denominations of the silver denominations from 1794 to 1837. Admittedly, these are only surveys of those who choose to reply (mostly JRCS members) and do not include all survivors. Many raw and unattributed coins are not identified. But the surveys are not just numerical counts, they are leavened with the knowledge and experience of the members. Thus, sometimes previous census rarities are changed, usually reducing the rarity, say from an R-5 to an R-4 as new known survivors are identified.

The same, more or less, can be said of EAC for early US copper and C4 for pre-Federal and Colonial issues.

As for the 1823 dime in the OP's post, there is no way that a coin with 150 known survivors has been assigned a rarity of R-7. The Sheldon scale (both the original from 1949 and more modern versions) assigns R-7 to a known survivorship of 4 to 12 examples.

My references list only three die marriages for 1823. JR-1, R-3. JR-2, R-4. JR-3, R-2. The large E (E of STATES) varieties are either JR-2 or JR-3 both using Reverse W.

R-2 is 501-1000 survivors.
R-3 is 201-500 survivors.
R-4 is 76-200 survivors.

Now, sometimes a coin might be a very common R-1 but have a conditional rarity in grades above or below a certain point. There are even particular coins or series where circulated coins are rarer than uncirculated coins. But if someone claims a R value for a conditional rarity, they need to specify that it's a conditional rarity otherwise it's just a lie.

I'm familiar with the URS scale and I see its logic but I fail to see the need for it. I have the same feelings for it as I do for the NGCX decimal grading scale - OK but why? All these scales are arbitrary anyway except for the original genesis by Dr. Sheldon of his R scale which was designed to arithmetically "fit" his intended pricing scheme.

Anyway, hope this sheds some light and no confusion.

 

I don't understand why they put up those R numbers if they're wholly inaccurate, but they're there for certain.

 

PCGS uses a different scale -- the PCGS CoinFacts Rarity Scale™. It's a classic example of precision without accuracy, having 92 different rarity numbers from 1.0 (> 1,000,000) to 10.1 (none known). As the PCGS CoinFacts Rarity Scale™ is a registered trademark of PCGS, it's unlikely that anyone else will use it. The traditional 8 numbers is sufficient and well-understood (unless you're a CWT person, then you have 9).

 

When it comes to Bust Dimes I always refer to the 'Bust Dime Variety Identification Guide' by Zack, Scuderi and Merrill.

The guide describes 4 varieties for the year, as @Publius2 says. From your pictures you appear to have an JR-2, which is rated R3 in the guide, ie, scarce with 201-500 estimated examples extant.
The scarcest variety for the year is JR-1, which is described as R4- (76-200 estimated).

The year 1823 has 3 die varieties. Two of them are large E's.
Jr-2 is one of the Large E's. It is an overdate 1823/2. This is the scarcest variety for the year at R4.



The guide is already about 8 years old so current numbers may differ somewhat. However, as previously mentioned, the publishers base their estimates on their extensive expertise with Bust Dimes, specially when it comes to the varieties which most collectors seek like 1809, 1811/09, 1822, 1814 and 1820 STATESOF, 1824/2 JR-2, and 1829 Curled Base 2. Maybe a few others.

Finally, a word of wisdom (not from me) regarding R numbers: Over time, they will invariably always decrease.