This article examines long-standing rarity scales, their applicability to various numismatic collectibles and how different scales are used. We also introduce REAL™, a supplementary rarity system you can us to monitor your own key collectibles.

R-1 to R-9 rarity scales are used consistently in many aspects of numismatics. Few other hobbies have the same well-established rarity or numeric grading standards of coins, medals and tokens. Many coin hobbyists have a simple goal of getting everyone on the same page, so that when you find a medal, token or coin, you'll be able to evaluate its rarity and thus—hopefully—establish a common ground with others who collect, buy or sell the same item. A more ambitious goal would be to move all rarity scales to a common system.

The answer depends on what you're evaluating. If you're purchasing an original painting, you evaluate it based on the artist and the desirability of his/her work and prices realized for other paintings. An "R" scale wouldn't be very applicable.

If you're eyeing a coin from Canada, the U.S. or Portugal, you can consult mintage figures—in the tens of thousands (rarely), millions (certainly) or hundreds of millions (often). Price guides tell you the mintage and the grade, along with wholesale and retail prices.

As a collector, you know the criteria and resources used to value the collectibles in your areas of interest or expertise. But when you move into the field of tokens, medals and other related numismatic items, you'll need to shift your thinking about valuation from mintage to rarity. When coin mintages are extremely low or you are dealing with a specialty token or medal and the number in existence is relatively small, the R-Scale comes into play.

So-Called Dollars, Civil War patriot or storecard tokens, medals, early proof coins and other numismatic and pseudo numismatic issues don't address rarity in MINTAGE, even if mintage is known. The number extant is what drives value.

Consider the dictionary definition:

It's interesting that the dictionary uses a manuscript as an example; museums and historians are comfortable with that familiar reference. Obviously, we added the second line (in red). That is where a bit of ambiguity enters the picture.

A range of 5 to 20 extant for an item can be a huge range in some cases, making it difficult to gauge true value. Consider any token or medal you collect, curate or sell: If you know that five exist, is your evaluation of rarity and dollar value the same as if you know 20 exist? Certainly not. Knowing you might be purchasing one of only five extant could "feel like" a $500 item; one of 20 might "feel" more like a $100 to $200 item to you. And regardless of what the price guides say, if they are available and if you believe their data, that "feel" is what might govern your price decision more than anything else.

It's also important for anyone basing value on an R1–R9 scale to know which scale is being used. The accompanying table (which we realize is a little challenging to use) incorporates the major rarity scales you'll encounter. Clearly, if there are 500 or 1,000 of the item extant, it doesn't really matter whether someone call it "Not uncommon" or "Not scarce." It is equally relatively unimportant if it is called an R-2 or an R-3.

This question and answer are never black and white, but what is clear is that the higher the grade, the more desirable the item regardless of rarity. Conversely, the greater the rarity, the less important the grade. While this sounds contradictory, it isn't. Rather, you should think of grade and rarity as complementary factors, with the importance of one increasing as the other declines.

This might best be expressed in a table that has very simplified statements of grade and rarity. Of course, when you are on the floor of a show or in an internet or phone auction debating an acquisition, it will never be so straight-forward or simple. But we highly recommend having a personal "cheat sheet" in hand so you are sufficiently armed with information to help you make a decision when the situation arises. We don't say "if," because eventually, you will find this dilemma—guaranteed! We just can't tell you how close your situation will come to the hypothetical for which you prepare.

The accompanying chart lists "gross" grades; as we all know, there are myriad split grades, and of course numerical grades, within each of those used for illustration. The chart provides a simple starting point, i.e., the range in which you are interested in purchasing. Of course, you don't want a cull, and if you happen to find an MS68 in R-1, it certainly looks better than an AU in R-3. Judge the rarities for yourself, and judge the price, as well.

If you find an R-1 and an R-5 for $50 each, and the R-1 is MS60 and the R-4 is AU58, we're assuming you'll purchase the R-5 unless that specific R-5 is known to exist mostly in high-grade, uncirculated specimens. If you find an R-4 in AU for $50 and an R-6 in Very Good for $80, we assume you'd choose the R-4.

The choice between an R-8 in Extra Fine for $200 and an R-4 in MS66 for $140 is a little trickier. Will you be able to find another R-8 . . . ever? For that matter, will you find a comparable R-4 ever again? In these cases, you have to be the sole arbiter in the decision. The money comes out of your pocket and you probably know best which one of these specific items will ever or never pass your way again.

Below is a tool to start your decision process. Approach it as a process, not a simple decision.

EVALUATING GRADE VS. RARITY         

We've made two basic assumptions in this guide: 1) You generally want nothing below a VF in your collection and 2) You generally do not collect, say, R-1 or R-2. The blue shading represents your target purchases. Even though you generally don't collect R-2 or R-1, perhaps a super high grade would make you reconsider.

You might consider creating a "dollar" chart to overlay on this one, utilizing the same template and substituting what you're willing to spend. If your upper limit is $500 and your "no problem" low is $100, you could easily overlay a series of amounts ranging from $100 to $500 and end up with a table showing that, just hypothetically, the R-9 in VF is $250 moving across the R-9 line to $500 for an MS65. We're then assuming that the left-to-right movement of dollars would follow, with the amounts becoming progressively lower as you moved down from R-9 to R-1.

As collectors, dealers and researchers, we assume many of our readers, if not most, are familiar with Rarity Scales. Important names to recognize are William Sheldon, whose numismatic literature is highly respected and who created his scale for use in various applications; he is the one who, more than 60 years ago, devised the coin grading system of 1–70, best known for the various MS60–MS70 uncirculated grades. Q. David Bowers is arguably the foremost numismatic author who has been a dealer, collector, and writer in the hobby for 50-plus years.

J. Hewitt Judd provided us with identification of many varieties, most notably early proof coins, and his Rarity Scale is widely recognized/used. George Fuld is most recognized for Civil War numbers, and Kanzinger's Civil War Token Collectors Guide was the most recent attempt to provide an update on rarities and values for CWTs, published decades after Fuld's scale. David Bowers developed his own Universal Rarity Numbers in 1992, which run in the opposite direction to all other rarity numbers. Interestingly, even someone as popular and well-respected as Bowers hasn't been able to change a half-century of thinking, and thus his scale has received little interest among collectors.

The Hibler & Kappen So-Called Dollars book published in 1963 provided overly simplistic rarity ranges (far right column in the chart below), but fortunately the editors of the revised 2008 edition adopted the Sheldon scale instead, providing a more realistic reference for collectors.

As you can see, jumping from one scale to another can be problematic. If you're a longtime CWT collector, an R-5 has a concrete meaning in your mind. Fuld tells you that an R-5 token means 75–199 are extant. If you collect proofs and Judd is your mainstay, R-5 still means 75, but instead of being at the low end, it's at the high end with a meaning of 31–75. Continuing down the line, Judd and Fuld scales overlap but are essentially one grade off in practical terms.

COMPARING RARITY SCALES

A lack of this knowledge routinely causes novices to overpay for a quality piece, equating quality with rarity. The Rarity Scales ONLY MEASURE KNOWN PIECES EXTANT; they do not address quality. Two different pieces each with say, 500 known, may be worlds apart in known high-grade pieces. It is not unreasonable to expect, for any given item or type of item, to find 10 percent of one R-5 group to be uncirculated, MS60 and higher, and another to have 75 percent of its extant numbers known in uncirculated grades. The reasons for and details describing the apparent anomalies for every example would take pages and pages to explain.

Likewise, it would take volumes to evaluate the details and nuances of the more than 1,000 So-Called Dollars listed in the Hibler & Kappen "bible" (and only guide to SCDs). The original H&K was issued more than 40 years ago, and there have been mostly little-known or unworthy attempts at approximating that volume since. Then, in 2008, a team of editors collaborated on updating the original H&K (we were one of a number of experts contributing to the new book) and also devised a price supplement. Perhaps the best thing that the editors accomplished in the second edition was adding a Standard Rarity Scale, replacing Hibler and Kappen's original, silly scale that was tantamount to "pretty rare, kind of rare, not so rare" and was "kind of sort of" useless.

The new price guide is helpful, but only "kind of" again. For example, take the common HK154 and HK155 from the World's Columbian Exposition. Both are listed at $5–$30 circulated and $20–$75 uncirculated. That may provide a little help, but it's vague. The silver and gilt SCDs from the U.S. Centennial, HK20 and HK21, likewise have a price range broad enough to accommodate the Titanic.

HK20 is $50–$150 circulated, $200–$1,000 uncirculated; HK21 is $30–$75 and $75–$300. Will that help you make a purchase? What should you pay for an MS63 of either? How about an MS65 or an AU50?

Let's look at a typical and real example from H-K. HK559 is an R-2 celebrating the centennial of Oregon statehood. Circulated specimens are listed at $5–10, which is about as inexpensive as any SCD listed in the book. Uncirculated specimens are listed as $10–$150. The editors could hardly provide a comprehensive listing for every grade and every listed item without producing essentially an entirely new book. And what about the other nearly 50 unlisted Oregon Centennial SCDs? Clearly the book will need to double the number of SCDs listed before it does a credible job of really "updating" the original edition.

Collectors have always been concerned if, for example, something is an R-8 (which, depending on the rarity scale you're using, could mean that as few as two or as many as nine are known to exist) and yet you've seen, say, two in the last year. How could that be? Have you seen the same one twice? Have others been discovered and come to the marketplace? Could it be you've seen 40–60 percent of all that exist? If you believe that the rarity scale citing 2–3 extant is the one you're following, how on earth could you—an average collector—have seen ALL of these known to exist? Possible, but highly unlikely.

The History Bank formed a subsidiary in 1987 to conduct various types of research for our clients and for our company. Laing Research Services has been retained by museums, universities and private companies to take on some very interesting research questions in the past two to three years. This issue of rare, rarer or rarest is one that we have pondered for many years and decided to tackle on our own, rather than as a consultant.

We have created a method for evaluating rarity based on YOUR personal estimates as a collector, appraiser, curator or archivist. It may seem a little cumbersome, but it's actually quite simple and requires only basic arithmetic based on how often you see an item.

The REAL ESTIMATE and ANALYSIS LOG (REAL™) for determining rarity is intended to provide you with a tool to compare your estimates, findings and gut feelings to the established rarity assigned to a given coin, token, medal or other collectible.

The greater your involvement in an aspect of a hobby, the more reliable your findings with this method. If you attend one show and one auction a year and cruise by eBay every couple of months, this method will do you NO good. If, on the other hand, you attend a show monthly and follow eBay weekly, the validity of your findings may be better than the R-rating.

The level of your involvement will also affect the accuracy of your observations—and your ability to judge the R-rating. If you are moderately active in the hobby and you NEVER see an item, it doesn't necessarily mean the item is rare, because you simply haven't had enough exposures or opportunities to see it. On the other hand, although it may seem contradictory, if you're hitting multiple shows and are highly involved in auctions and you never see the item, it doesn't necessarily mean you can gauge the rarity more accurately; it DOES mean you can determine how accurate the published rarity might be.

If you attend two shows and two auctions every month and you see five examples of a coin that is allegedly an R-9, you can't say for sure what it should be. But you can say what it isn't—and it isn't an R-9!

If you only hit a few shows each year, have little exposure in the hobby and see an item once or never, you don't have enough information to disprove an R-9 designation; if, on the other hand, you have 50 such experiences and never see the R-9, we probably can believe it is an R-9 or nearly so.

So, please read through the methodology and see how it might apply to your own findings and feelings about rare items. We've tested and massaged this formula to arrive at what we think is a user-friendly, effective tool for evaluating your views on rarity as a supplement to assigned R-numbers.

Begin by keeping a log, formally or informally, on any number of given rarities among proof coins, So-Called Dollars, tokens and medals. This system is designed to make sense and be of help if you fill out the following matrix—preferably accurately, but even estimated from memory. "I saw three to four last year." That's fine. Following the matrix below we have outlined examples to calculate the numbers and give you an estimate to compare with a Rarity Scale number.

YOUR REAL™ WORKSHEET

Now that you've given serious attention to how many tokens or medals you've seen and you've worked through the REAL™ calculation process, what do you do with the numbers? Following are two hypothetical examples that will illustrate the application.

A—You saw NO examples on eBay = 0
B—You attended 20 shows and saw 2 examples = .1
D—You saw 2 for sale in publications/ads read weekly = 4
You have 18 years of experience = 2
G—The total—and REAL™ computation that can be used to help analyze current R ratings = 8.2

A—You saw 6 examples on eBay = 36
B—You attend 12 shows and saw 10 examples= approx .8
D—You read publications weekly and saw 10 examples = 20
You have 18 years of experience = 2
G—The total—and REAL™ computation that can be used to help analyze current R ratings = 113.6

In Hypothetical #1 you are an experienced person with 18 years as a collector and your extrapolations provide a first step in estimating the number extant at 8.2. Does this mean that only 8.2 are known to exist in the world? No, but it gives you information on which to base your analysis—and this is the part of the process where you have to apply your own innate logic to evaluate "G" at 8.2. Considering that you are a serious, experienced numismatist who is VERY involved on a daily basis, you probably see what's out there in the hobby about as well as anyone. Using that innate logic it is reasonable to say that you might have seen 20 percent of all such items that exist in the marketplace, or 5 x 8.2. That number, 41, would translate to approximately an R-5. What's the book on this item? If it's an R-7 or an R-8 the odds are that the rating is overrated; if it's only an R-1, the odds are strong that the rating is an error. Our system provides you the tool by which you can analyze your findings and compare them to known R ratings.

In Hypothetical #2, with you having the same background and seeing 20 percent of all the examples out there, the item would be at 5 x 113.6 or 568 total. This is an R-3, and as the number in our REAL™ methodology goes higher, there is less guesswork. We know that you've seen at LEAST a number that we feel is an R-3. If the book on this piece is R-5 or R-7 it's out of whack, with little doubt. If the book says R-1 and you have it at an R-3 with your own experience, odds are likely that it's no better than R-3 and PROBABLY less.

The key here is using this method to give you information and insight, not to give you hard facts. We can only speculate rarity based on your experience. But consider this: This is more scientific than any given book assigning its rarity. We're not criticizing anyone or any guide assigning rarity, but each guidebook assigning hundreds—thousands!—of rarity to say, Civil War tokens or SCDs, is doing so by asking a few experts to share their gut feeling, and we're providing real tools.

Obviously, the more examples of an item ones sees, the better this system can guide you.

However, when you have no information, you're in quite a quandary: If something is an R-7 and you've seen NONE, what does that tell you? Only that you can't be sure or that MAYBE it's even an R-8. But when you see an item yourself 3, 4 or 5 times in a year, and the odds of it being the same piece resold in a short time are quite slim, you know one thing: It's not an R-9! And the REAL™ system will help you decide on down from there!

Experts are constantly reevaluating rarities assigned to SCDs and Civil War tokens. Naturally, the more frequently a guidebook is published, the more up-to-date such figures will be. It is reasonable to see a guide calling a Civil War die pairing an R-9 in 1960, but when a revised edition or new book is published in say, 1995, perhaps new information was discovered and that die pairing thought to be unique is actually fairly common. The new book calls it an R-5. Such revisions happen more than occasionally, if not frequently. In all aspects of numismatics, like all hobbies, new information or even new stashes of items are often found.

We have seen an item with only 5–6 known for decades, only to have someone find a hoard of 10–20, which changes the entire complexion of that subset of a hobby. Rarities of and knowledge about SCDs, Civil War tokens, medals, proof-only issues and so on can be turned upside down in a heartbeat with discovery of a cache of 20 or 40 coins or items, a box of proofs, a couple of rolls of mint state issues and so on.

The HK154 (left) and HK155 (right) SCDs have a variety of differences, but the most obvious is the size of type across the top arc of the medals. HK154 has large letters in "U.S. Govt Building" and HK155 has fancier script and much smaller letters. But the biggest difference? No way should they both be R-2. HK155 is moderately to substantially rarer, depending on your research.

So-Called Dollars can be even more suspect; the more we see of a given example, the more we question the often outmoded R-values assigned by Hibler & Kappen. Just selecting easy examples, look at HK154 and HK155 that are both listed as R-2—and have the same dollar value in the new price supplement added to the revised book in 2008.

If you are not familiar with them, ask someone who is: The consensus is that 155 is substantially rarer than 154. We applied the REAL™ evaluation for ourselves to see how it might compare. Naturally, we did not keep an annual log of sightings, but used our best guesstimate.

Here is my personal observation of both HK154 and HK155, and with the market slowdown, I've done estimates for both 2008 and 2010. See how yours compare.

COMPARING HK154 and HK155

Our experience with estimates without prejudice has HK155 at a REAL™ number of 88.67 in 2008. As active as we are in WCE collectibles, we probably saw 5–10 percent of those in the marketplace. And we know collectors by the dozens who have 10–20 of them. This might easily mean a market of 20 x 88.67 or 1,773—and that is indeed an R-2. In 2010 our number is 95 x 20, or 1,900—still an R-2.

We have long held that HK155 was much rarer than HK154. But our own numbers in '08 and '10 show HK155 to be only 2.65 and 2.21 times rarer than HK154. Our gut feeling, despite our own numbers, is that the ratio should be at least 5 to 1. What do you think?

Either way, HK155 is clearly at least an R3 if HK154 is an R2.

The REAL™ method enhances what you know and what you suspect; it looks more comprehensively at the assigned rarity charts and gives us more information with which to assign a rarity level.

We included the World's Columbian Exposition's HK154 and HK155 with good reason. We know the medals well and we can compare our own ideas with our own experiences. We have long felt that they are both not deserving of the same R-2 rating. Everyone in the hobby "believes" that the HK155 is rarer than its HK154 mate, but now we can see validity. This is an excellent example of how the REAL™ system can help.

It is impossible to have one person's observations determine the rating for any item. If the R ratings have been determined a) by asking a few hobby experts and b) perpetuating what has been assigned for years, then we have provided a method to ensure that virtually every R rating will be more accurate than in the past.

We say virtually because if the observer using this system sees NO EXAMPLES we cannot contribute statistics or even estimates to enhance current ratings. But even a bottom line of zero says something! If the observer is a seasoned hobbyist and sees NO examples, we know it's rare; if it currently is labeled as an R-1 up through, say, an R-5, something's askew. If it's an R-8 or R-9, there is good reason for not seeing any.

With a start like this, and developing a "clearinghouse" approach so that others receive the information, we will begin to see much more accurate R ratings from top to bottom.

Tell us how it works for you—we'd love your feedback!