I haven't tried it, don't think it will work. Bicker and barter, even with slabbed coins adds excitement to the hobby. But for the purpose of perhaps our new auction and classified sections, it is the responsibility of both the seller and buyer to be reasonable with a starting bid and for the buyer, theirs to responsibly participate or not. ( agree with starting bid and the grade ,or not)
Yes. Yes. Since my personal definition of uncirculated basically says, "No wear", then it doesn't matter what the computer thinks as long as it indicates a wear-like area (it also may not be able to identify an original planchet surface from a weakly struck coin). Under my simple model the computer flags the area for futher scrutiny by a human. Then the human would decide: - not fully struck - circulation wear - cabinet friction (if differentiation from circulation wear is important). Human analysis will probably always be necessary.
You might be interested in this patent on computerized coin grading that was assigned to PCGS in 1993, 16yrs ago and still not utilized ( publicly) today. There are abstracts, description, full text ( lots of good math here) http://www.patentstorm.us/patents/5220614/description.html http://www.patentstorm.us/patents/5220614/fulltext.html Jim
I guess I should hypothesize here that the main importance of this patent lies in it's ability to block any efforts that would infringe on it by other companies who might want to try. IMO. Jim
Mark, I have a question for you because I value your take on the subject. I was reading in the PCGS Guide to Coin Grading and Counterfeit Detection that PCGS uses a grading set to compare submissions to. I have been told here in this forum that you can't grade across a series due to strike characteristics of different mints, quality of strikes for the year, etc. If they can't have a grading set across the series then how do they do it? and if they do have a grading set across the series, as it pertains to key dates what is the difference between market grading coins and bumping them a grade because they are scarce? In other words, why do they seemingly go easier on key dates?
They do have a grading set, as you say, across the series. A grading set is a complete set of coins of any series. That means an example of every date/mint combination in vitually every grade. In the beginning I think there was a bias towards key dates. In other words people tended to be more forgiving because the keys were harder to find and because many people, professional graders included, get somewhat excited when they see one. In todays world I think the value aspect of market grading plays a much larger part. The grade is assigned based on what the value of the coin is - not what the condition of the coin is. Why do you think the term gradeflation originated ? It came into being because in the old days coins were graded by the TPG's based on condition, with the bias towards keys that I mentioned. As time went on, that bias towards keys inched ever upwards and thus the term gradeflation came into being. If anyone can remember, it was the key dates and the rarities that were first discussed and written about in the press as being over-graded. And it was not so long ago when it first started, it was in this decade. But in today's world, that gradeflation has widened to encompass nearly all coins, not just the keys. In today's world coins are priced - they are not graded by the TPG's. Now some will argue that they were always graded based on price/value. But what I have mentioned above contradicts that. In other words, results say otherwise.
You bet it's a lot of coins. Ever seen one of the grading sets at a coin show ? And they only display small portions of the set at any one time. Now rather obviously Mike there are more than few coins where it just isn't possible to have one in every grade, the coins don't exist. And with moderns there are few if any examples in some of the lower grades. And of course you aren't going to see MS70's and MS69's either - again they don't exist. But with the common stuff, they pretty much have it all.
You know what just struck me, Doug? It's funny, but where the grading sets are lacking is likely in the key dates, hence the overgrading. Maybe true, maybe just ironic. I dunno.
I believe, that because each coin is unique, grading sets have very severe practical limitations. You can spend all day trying to match a given coin to one in a set, but there will always be differences, subjectivity, biases and inconsistency involved. Allowances are made for differences among various dates and mints. So, for example, if you have a date/mint that is typically weakly struck, a high mint state grade can/will still be given to an example which is lacking a full strike. On the other hand, if a certain date and mint is normally very well struck, a weakly struck one can/will be penalized. And like it or not - I don't! - the grading companies are definitely more forgiving when grading especially rare/valuable coins. Doug, they couldn't possibly (and don't) have grading sets like that for most series. The coins simply don't exist for many dates and/or grades that would need to be included. And even if they did, the cost would be beyond prohibitive. Edited to add: What they probably do have is grading sets for major series by type, but certainly not by date and mint. Below is a link from a 2007 article about the PCGS grading set. http://www.pcgs.com/articles/article_view.chtml?artid=5114&universeid=313
20 years ago is a long time. Computing power today is almost infinite compared to then. Whatever can be detected with the human eye and interpreted with the human brain is or will be possible with advanced optics and computers. Cabinet wear and circulation wear can be identified using the same basic algorithims used by the brain. It's just a matter of including the rule set in the program. I think it would be more correct to say that it isn't economical to develop the programing and hardware to accurately grade coins. It isn't correct to say it isn't possible --- all in my humble opinion since we have no way to really know.
Hmmmmm - from your article - "PCGS Grading Set coins displayed in Long Beach included multiple examples of all grades from About Good-3 to Mint State-67 for many major U.S. coin series;"
Doug, that isn't by date, it is by type - I have seen some of the "sets". The article mentioned about 1000 coins. For example, a complete Mercury Dime grading set by date, consisting of "only" Fair, AG, Good, VG, Fine, VF, XF, AU, MS60, MS63, MS65 and MS67 (if it even existed from the entire population, and I bet it doesn't) would be close to 1000 coins all by itself. Think for a minute about the staggering numbers and value of coins it would take to form grading sets by date. Again, in most cases it would be literally impossible - check the PCGS Population Report if you think I'm mistaken.
Fair enough, you certainly know more about it than I do. I can only go by what I have seen and what they have told me at the shows.