From time to time we see references to the Sheldon 70 point grading scale when talking about the grading system that is currently being used in Numismatics today. For the benefit of those who are not familiar with what Dr. Sheldon did and why he did it, I’d like to offer a short summary. William H. Sheldon was born in 1898. He first became interested in Early Large Cents at a young age while looking through his Fathers cigar box full of those big old coppers. By the time he was a teenager, he was pricing Early Large Cents for a local dealer and according to him, using a 70 point system. As the years went by he became quite the aficionado of Early Large Cents (those from 1793 through 1814). On three different occasions he assembled sets with the last containing all 290+ known collectable varieties at the time. To put that more into perspective, that collection contained at least 16 (sixteen) 1793 Large Cents! In 1948 he published EARLY AMERICAN CENTS detailing much of his research as well as introducing his 70 point concept to the hobby. (Technically speaking, it’s not even a 70 point system since Dr Sheldon only used 22 of the 70 numbers as compared to the 30 or so in use today.) Although Dr Sheldon is well respected by the numismatic community for his research, if not for his honesty, the general consensus at the time was that the 70 point concept was a “whimsical” idea. In 1958 he published a revised edition and apparently as a lighthearted gesture in response to the criticism renamed it PENNY WHIMSY. In his 70 point grading system, Dr Sheldon was attempting to equate the value of Early Large Cents by using a numerical grade. To accomplish this he first determined the value of the various Cents in what he called “basal” state. This he defined as “identifiable and unmutilated”. For basal state he assigned the numerical grade 1 (one). Everything from here on is predicated on the basal value. By examining auction records and dealer price lists, along with a little interpretation on his part, he determined that a coin grading “good” was worth four times basal so he assigned the numerical grade 4. For a coin grading “fine” he determined that it was valued at 12 times basal and assigned the numerical grade 12 and so on up to “perfect uncirculated” which he determined to be 70 times Basal. So in effect, the 70 point system was chosen because it provided the multipliers he needed for his system to work (in most cases) with regard to common dates and varieties that were readily available. If a particular coin was scarce, rare or a rare variety, additional calculations were necessary in order to determine value. This he accomplished by devising a set of 10 “rules” to be applied as needed based on the condition census at that time. For example, rule #5 states “If known to be in the first half dozen but not among the first three, and not full MS, value equals condition times basal times 1 ½”. In some cases, rule #10 would need to also be applied. “If tied for honors with one or more coins, simply calculate what the total value of these coins would be when placed in descending order of rank as if there were no tie, and divide by the number of tied coins”. What that means is that in some cases, several rules would need to be applied in order to get the desired result. Of particular interest is the fact that his system was designed for Early Large Cents of 1793 through 1814 and not for the remaining issues of Large Cents, nor any other series of US coins. Even at that, there were certain issues of Early Large Cents where his system would not generate the desired result. Also, the multipliers he determined for the scarce and rare varieties were skewed due to the fact that the information available for condition census coins was very limited at the time. From 1965 to 1990, more than a thousand Condition Census coins have surfaced. For many of the varieties, the finest known in Penny Whimsy would not be in the top six today. Dr Sheldon was also rather vague in defining what constitutes a certain grade. He assigned three numbers for “good” coins (4, 5, & 6), but only has one description. The same for other adjectival grades on up the ladder. Clearly his objective was not so much to define the grades themselves, but to assign numbers to accommodate his system. Some of his adjectival grades are different than those being used today. In addition to basal for grade 1, he uses very fair for grade 3. The descriptions he offers for the various grades offer interesting reading as well. For example, he defines AU 50 as “Close attention or the use of a glass should be necessary to make out that the coin is not in perfect Mint State. Typically, the AU 50 coin retains its full sharpness but is darkened or is a little off color”. The descriptions for 60, 65 and 70, the only three MS grades he used, are very vague as well. Here’s what he says about MS 60. “For condition 60 a minor blemish, perhaps some microscopic injury, or light trace of discoloration may be tolerated”. One interesting note is that in his breakdown by variety he lists 9 (NINE) MS- 70’s in the condition census for the 1795 Large Cent! As you can see, aside from the numbers themselves, today’s numerical grading standards have little in common with Dr Sheldon’s. We can only speculate as to what the current system might look like today had Dr Sheldon needed a multiplier of 93 or 117 for his “perfect Uncirculated”.
Thank you for the article. To further complicate Coin grading today, we now have Star (*) designations, plus (+) designations and of let's not forget the new kids on the block, CAC .
Well written article...wonder what happened to those MS-70 large cents? They must have been quite nice...proofs maybe? What do y'all think?
Thanks for the historical commentary, Vavet! I actually didn't see this article just now, but just commented on another thread here about technical grading vs market grading that it's no mistake that the original intent of the scale was as a standardized method for calculating the value of large cents via multiplication. The system broke down, of course, because markets change and there are certain coins where the difference between EF and MS is much larger than for others; Sheldon liked to classify things, and like his attempt to classify humans into categories of endomorphs, mesomorphs, and ectomorphs based on their physical characteristics (with particular intelligence levels and other traits associated with each category) trying to make predictions based on mass groupings of things with widely varied origins tends to fall apart. You can read more about his physiological/psychological research here: http://www.trutv.com/library/crime/criminal_mind/psychology/crime_motivation/6.html Also, as Vavet alludes with his comment "if not for his honesty," Sheldon was no superhero- he has since been discovered/accused of stealing from the American Numismatic Society's collection of large cents. http://ansmagazine.com/Summer04/Cabinet. An interesting guy, to say the least! --- "Silver and gold are not the only coin; virtue too passes current all over the world." http://myworld.ebay.com/cannoncoins
The Sheldon linear scale (Price = constant x grade) broke down within a few years of the publication of Early American Coppers. If we look at price versus adjectival (Good, VG, Fine, VF, EF, AU, MS60, & MS63) and plot adjectival grade (not numerical) vs. Price (on a semi-log scale), you actually will see a reasonably linear relationship. This relationship has held pretty much linear for over 30 years (e.g., 1981 - 2011). For type coins, each step up in grade is usually accompanied by a ~75% increase in price (obviously this slope will vary depending on the coins series. For example, if you look at the different series of Large Cents, you will see that the slope (increase in price for each step up in grade) will vary (these are relative price increases for an arbitrary year between 1981 and 2012—while the absolute price of a VF-20 coin will vary, the price of an EF-40 relative to the VF-20 stays pretty much the same): Chain 67 ± 8% Wreath 65 ± 3% Liberty Cap 79 ± 9% Draped Bust 106 ±10% Classic Head 111 ±8% Coronet 79 ±12% Braided Hair 76 ±16% 85-coin Red Book Type Set (excludes clad coins) 75 ± 4% The data has been published in three articles to date: ——“Pricing Relationships of United States Type Coinage”, The American Journal of Numismatics Second Series, 23, 257 – 263 (2011) ——“The Sheldon Scale and Price-Grade Relationships in Large Cent Type Coinage”, Penny-Wise (publication of Early American Coppers), XLVI, 5 – 15 (2012) ——“The Sheldon Scale and Price-Grade Relationships in Half Cent Type Coinage”, Penny-Wise (publication of Early American Coppers), XLVI, 31 – 39 (2013) As to why the behavior is exponential (semi-logarithmic), I don't have a clue. All I know is that the pricing relationship has held pretty constant.
Very much agree with the last two columns! us EAC types also use G-5 (sometimes as a substitute for Good), F-15, VF-25, VF-35, and I've also seen G-6.
Thanks for the link. Very interesting article. To the OP: Thank you for this thread. While I don't collect this series, I learned a lot today. I vaguely knew about Sheldon and the reason behind the 70 point scale. You've added some additional detail for me. Thanks.
In that chart the column "What Sheldon intended" is too large. He never intended to use anywhere near that many grades. He just used 1,2,3,4,8,12,15,20,25,30,35,40,45,50,55,60,65, and 70. Later G-5,6, VG-7,10, 63, and 67 were added.
When we (Dennis Fuoss, Bill Eckberg, Ray Williams and myself) were writing the Grading Guide to Early American Copper Coins http://www.eacs.org/GradingGuide/GradingGuide.html (shameless plug), we had to decide on what grades to illustrate. We looked at thousands of images of colonials, half cents and large cents to chose representative images and the main criteria became, at what step up in grade could you see a difference? If we looked at two different coins that had slightly different numerical grades (e.g., VG-7 and VG-8), but couldn't see a difference, we ignored the "odd" grade. We ended up with P-1, Fr-2, AG-3, G-4, VG-8, F-12, VF-20, VF-30, EF-40, EF-45, AU-50, AU-55, MS-60, MS-63, MS-65 (and a single MS-67). Grades intermediate to these (at least for early copper) were hard to distinguish between the next lower and upper grades. I can see the inclusion of additional MS grades (especially MS-64), but for me, when I grade coins, I'll stick with the ones I just listed. As far as the upper end of MS (68-70), my pockets aren't big enough to carry around the microscope needed to distinguish between them.
