I not only know its possible having been implemented over 20 years ago, but think it could now be an app on the app store.
With one exception - eye appeal. A machine cannot tell beauty. So beauty and eye appeal would be left up to the buyer and not a part of the grade - just like it is today. An ugly MS64 slabbed coin has more problem being sold than a pretty one.
A grade could be made for the coin based just upon wear/marks etc. The grade would be a lot more
consistent and
accurate than modern humans can produce ("The best graders only get it right
* 80% of the time -
former PCGS president Ron Guth). The link above (form the 90s) even says it was 90% accurate. Imagine today's machines capabilities compared to those of 20 years ago.
Eye appeal would be left up to the buyer and seller -
like it is now. Are we talking about GRADING or about assigning a value? An ugly MS65 does not sell as easily as an MS65 with eye appeal.
Feasibility with modern technology:The new iphone instantaneously and automatically takes in 30,000 data points with an infrared sensor to ID its owner and power on. There is no worry about proper lighting since its infra-red. Those 30,000 data points which can ID the owner even if they are wearing glasses, a new mustache etc, certainly could qualify the data it collected off of a coin.
As to mistakes in the programming - and this is from someone who used to teach programming - there are innumerable programs in use around us in the world that have to have flawless accuracy beyond what would be needed or assigning a grade to a coin.
The developing stages for an app. would include testing and retesting by letting the iphone "see" the same coin over and over from different angles, just like it does its owners face, and working with it until it assigned the same grade each time - just like it can ID its owner and not be fooled by a photogragh.
The algorithms for the actual grading could be made public so as to allow people to understand how a certain grade is applied. The main problem would be the initial deciding (which PCGS already did according to the above article from the 90s) of how the algorithms would handle the data it collected and what percentages of wear/damage would correspond to what assigned number grade.
Why do it?1.
B/c it gets rid of human error and makes for an actual, scientifically sound, verifiable grade.
When the mint needs a certain purity of gold for making its bullion coins, does it rely upon opinions of people who are trained to look at the metal (and get it right 80& of the time if they are good), or is it scientifically tested?
2.
Contradicting data from the same TPG would be eliminated. Look at the following thread and realize threads like this (or similar) are surely not hard to find:
http://goccf.com/t/322364 3. The
coin is re-instated as the object of the coin collecting hobby. No more cracking out and resubmitting to try for a higher grade. The PCGS guarantee mentions resubmissions for numerous reasons. If the same grade is assigned over and over no matter which iphone the coin is put under, and the grade is scientifically verifiable (even make the program give a detailed output of its findings), then there is no way people would be gambling their money again. Again, the system would make the
coin what is being considered and
not the label.
5.
People of today are much more likely trust a computer grading system today than people of the 90s:Backk in the 90s computers certainly were not in every home, and there was still a large mindset in the populace that hesitated to trust computers (the upcoming "neveer-waste-opportunity-for-a-crisis" Y2K scare did little to help this). But nowadays almost everyone carries a computer in their pocket that makes the Windows '95and '98 machines of yesteryear seem like a, abacus. Trust in a new computer grading system likely would not be any harder than offering it for a 1.00 on the app. store. Once a person tested it and saw it would give the same grade no matter what, they likely would equae it with the phones ability to ID them and automatically turn on and no questions would be asked.
What is slowing this down from becoming reality?1. Its a horrible business move for grading companies to move towards. There are only so many coins out there and so many grades to be given. If a grade is made rock solid and verifiable, then the subjectivity is gone. Once graded, always graded...period. No repeats. No money for re submissions trying to get a highger grade. No crossovers. And the TPGs also lose their perceived status as the coin experts siknce its now a machine doing the job (and doing it with more accuracy).
2. There is a legitimate section of the hobby out there that enjoys the slabbing system as it is. Hobbies are about
fun. Think Beanie Babies - would anyone who took the time to
think about it
actually believe a mass produced piece of cloth stuffed with plastic pellets retain its multi-thousand dollar value through the ages? Fun was the key to the fad. And some people had a blast!
3. People have, on faith, put an awful lot of
money into slabs over the years. Subjectivity is something that, once money has been invested, makes for people who will fight tooth and nail to make sure they keep their investment (can you blame them?). To them it does not matter if the new system (in this case old, but upgraded) is more reliable, logical, and verifiable.
Being someone who always enjoyed tech and wanting verifiable data, I would love to see someone home brew an app. I have no doubt a legit app could be made with today's in-pocket tech. Most certainly it could be revived with incredibly much better accuracy on a big business scale.
-
*Mr. Guth said these best graders get it "right" 80& of the time. However, what a lot of people miss is the definition of the word "right" in that context. He would have been more accurate saying the best graders are able to ascertain what the final grade assigned by
PCGS standards will be 80% of the time. I am sure ANACS, NGC, etc. would all say that their own companies get the grade "right" though all use separate systems and standards.