Yes, I think it's one of those many early Computer Age dreams that still hasn't been realized.
... and never will be.
I tend to agree. It's not that you couldn't get a computer to generate an evaluation. But your subjective eye would never agree with its objective evaluation. And if you give the AI subjectivity, I'm not sure what you have but a tempermental AI. LOL
Charles Hoskins was one of their consultants helping to develop this system. He was the first Director of the ANA's Certification in DC and chose to stay in the DC area when the company was moved to CO. Shortly after, Hoskins became one of the originators of the International Numismatic Society and Director of that organization's Authentication Bureau which later became the first grading service in the U.S.
IMO, the Compugrade system was a failure. Another attempt after that did not appear to be successful either. IMO, any more money chasing this "dream" of machine grading will be a $$$,$$$$ loss.
However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
Unfortunately, this might change and limit TPGS profitability over time.
@Insider2 said:
... However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
@Insider2 said:
Charles Hoskins was one of their consultants helping to develop this system. He was the first Director of the ANA's Certification in DC and chose to stay in the DC area when the company was moved to CO. Shortly after, Hoskins became one of the originators of the International Numismatic Society and Director of that organization's Authentication Bureau which later became the first grading service in the U.S.
IMO, the Compugrade system was a failure. Another attempt after that did not appear to be successful either. IMO, any more money chasing this "dream" of machine grading will be a $$$,$$$$ loss.
However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
Unfortunately, this might change and limit TPGS profitability over time.
Who are these experts who are so absolutely expert that we should trust their consensus opinion for all time?
What if this "small group" are not unanimous in their evaluation?
Given that TPGS grading is market grading not technical grading, doesn't it have to evolve with the market?
Personally, I'd be glad for all those MS coins with rub to be downgraded to AUs, but I don't think it will be just the TPGS's that are unhappy.
@Insider2 said:
... However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
Interesting example you chose, there.
Just about any coin professionally graded by a TPGS since the late 70's (through 1995 and much later) has increased in grade. In 1999, I was told it is because the very experienced and knowledgeable professionals doing the grading have learned more over time by seeing more coins so that they have become even more experienced and knowledgeable. I picked an 1804 dollar as this grade escalation can be easily demonstrated and coins as this may be in the first group that should be "computer assisted" graded.
I would absolutely love to get the human element out of grading in numismatics. No more bad Mondays or Good Fridays or hangovers, or being PO'd due to arguments with the wife, or bias of types of coins, toning, etc, etc... Set "and agree" on the standards, (there's the rub) develop and "release the app". Of course PCGS, NGC and CAC aren't going to be advocates.
@Insider2 said:
Charles Hoskins was one of their consultants helping to develop this system. He was the first Director of the ANA's Certification in DC and chose to stay in the DC area when the company was moved to CO. Shortly after, Hoskins became one of the originators of the International Numismatic Society and Director of that organization's Authentication Bureau which later became the first grading service in the U.S.
IMO, the Compugrade system was a failure. Another attempt after that did not appear to be successful either. IMO, any more money chasing this "dream" of machine grading will be a $$$,$$$$ loss.
However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
Unfortunately, this might change and limit TPGS profitability over time.
Who are these experts who are so absolutely expert that we should trust their consensus opinion for all time?
What if this "small group" are not unanimous in their evaluation?
Given that TPGS grading is market grading not technical grading, doesn't it have to evolve with the market?
Personally, I'd be glad for all those MS coins with rub to be downgraded to AUs, but I don't think it will be just the TPGS's that are unhappy.
Do you live under a rock? Who do you think they would be? I'll bet if I offered you $100 to name just a few "experts" that could do a wonderful job, you could spit out a dozen that we would all agree with in the time it took me to one finger the rest of these letters including the punctuation and emoji!
That patent was a glimpse of what the future must eventually become. It is possible, and, IMO, inevitable that the system will become reality, Objectivity must replace subjectivity if grading is to be respected. Just witness the disagreements we constantly see here....Witness the 'upgrades' we see constantly....Standards must become reality. Cheers, RickO
A coin's assigned grade by a panel of "experts" would have ABSOLUTELY NOTHING to do with the market. Assign a grade, image the coin, and that dollar, once graded XF, and now graded MS-62 in 2019 will still be graded MS-62 fifty years from now. Only its price/value will have changed both up and down over the years in between!
@Insider2 said:
Charles Hoskins was one of their consultants helping to develop this system. He was the first Director of the ANA's Cat we should trust their consensus opinion for all time?
What if this "small group" are not unanimous in their evaluation?
Given that TPGS grading is market grading not technical grading, doesn't it have to evolve with the market?
Personally, I'd be glad for all those MS coins with rub to be downgraded to AUs, but I don't think it will be just the TPGS's that are unhappy.
Do you live under a rock? Who do you think they would be? I'll bet if I offered you $100 to name just a few "experts" that could do a wonderful job, you could spit out a dozen that we would all agree with in the time it took me to one finger the rest of these letters including the punctuation and emoji!
I guarantee that you could NOT find a list everyone agreed on. I simply point out the CAC controversy. JA would be on anyone's short list, but that doesn't mean they want to give him and his pals FINAL word on a coin.
@Insider2 said:
A coin's assigned grade by a panel of "experts" would have ABSOLUTELY NOTHING to do with the market. Assign a grade, image the coin, and that dollar, once graded XF, and now graded MS-62 in 2019 will still be graded MS-62 fifty years from now. Only its price/value will have changed both up and down over the years in between!
I point you to EAC grades vs. TPGS grades. The conundrum already exists. You start down-grading all the "market acceptible" 63 with rub down to an AU and you will have chaos. What you'll end up with is a 2nd or 3rd number: you'll have a PCGS number, a "panel of expert" number and them maybe an EAC number.
@Insider2 said:
A coin's assigned grade by a panel of "experts" would have ABSOLUTELY NOTHING to do with the market. Assign a grade, image the coin, and that dollar, once graded XF, and now graded MS-62 in 2019 will still be graded MS-62 fifty years from now. Only its price/value will have changed both up and down over the years in between!
I point you to EAC grades vs. TPGS grades. The conundrum already exists. You start down-grading all the "market acceptible" 63 with rub down to an AU and you will have chaos. What you'll end up with is a 2nd or 3rd number: you'll have a PCGS number, a "panel of expert" number and them maybe an EAC number.
I point you to the Husak collection where this actually happened.
I'll choose to ignore your point as EAC grading is stupid. Net grading will not be tolerated by the "expert' group assigning the "forever grade" put into the computer.
EAC GRADING is a compartmentalize, offshoot of normal grading devised by some unknown person for a small group of copper specialists with (by their own admission in print) absolutely no relationship to the way coins are graded by the TPGS's and everyone else.
They have their system. It is complicated and while it is not very precise by its nature (the same goes for commercial grading) it works for them.
Nevertheless, They have folks who would be in the group of "experts" because their opinion of a coin's grade is excellent UNTIL the "net" crap they do "kicks in" causing a different outcome from the rest of us.
@Insider2 said:
I'll choose to ignore your point as EAC grading is stupid. Net grading will not be tolerated by the "expert' group assigning the "forever grade" put into the computer.
EAC GRADING is a compartmentalize, offshoot of normal grading devised by some unknown person for a small group of copper specialists with (by their own admission in print) absolutely no relationship to the way coins are graded by the TPGS's and everyone else.
They have their system. It is complicated and while it is not very precise by its nature (the same goes for commercial grading) it works for them.
Nevertheless, They have folks who would be in the group of "experts" because their opinion of a coin's grade is excellent UNTIL the "net" crap they do "kicks in" causing a different outcome from the rest of us.
I know you don't know anything about the coin market. But go to the Heritage large cents section of their auction and you will note that they usually quote both the EAC and PCGS/NGC grade. This is the conundrum I refer to. You can't simply dismiss the EAC number because it does still matter to copper people. You don't have consensus, you have bifurcation.
Adding your own POE (panel of experts) will just become an alternative voice to the already crowded field. And throw in CAC, you will really have chaos. When your panel of experts assigns an MS62 to a coin PCGS calls MS64 and CAC beans the 62 but not the 64, is the coin really a 63? If your POE lacks sufficient market clout to shut down PCGS and NGC and maybe CAC as going entities, you simply have an option to PCGS/NGC that the market may or may not embrace. If a PCGS CAC sells for more than a POE non-CAC, people will continue to send to PCGS. Because, hey, y'know, if you are selling a coin, the market grade is going to matter more than the technical grade.
@Insider2 said:
Charles Hoskins was one of their consultants helping to develop this system. He was the first Director of the ANA's Cat we should trust their consensus opinion for all time?
What if this "small group" are not unanimous in their evaluation?
Given that TPGS grading is market grading not technical grading, doesn't it have to evolve with the market?
Personally, I'd be glad for all those MS coins with rub to be downgraded to AUs, but I don't think it will be just the TPGS's that are unhappy.
Do you live under a rock? Who do you think they would be? I'll bet if I offered you $100 to name just a few "experts" that could do a wonderful job, you could spit out a dozen that we would all agree with in the time it took me to one finger the rest of these letters including the punctuation and emoji!
I guarantee that you could NOT find a list everyone agreed on. I simply point out the CAC controversy. JA would be on anyone's short list, but that doesn't mean they want to give him and his pals FINAL word on a coin.
I guarantee that a poll done on CU to pick out a dozen numismatists that would set the final, "computer-recorded grade" on a coin for all time could be done - possibly at the embarrassment of all concerned.
I'd also point out this fact: There are expert graders who specialize in a particular series. Additionally, there are folks I would put up against the "best" professionals in the business that no one has ever heard of! It worked that way with our authentication consultants. I have found that long-time professionals, admired and considered to be "experts" by the whole world couldn't reliably authenticate themselves out of a paper bag! LOL!
I know you don't know anything about the coin market. [EXACTLY! I wish I could tell you what I found out this weekend about the coin market and...] But go to the Heritage large cents section of their auction and you will note that they usually quote both the EAC and PCGS/NGC grade. This is the conundrum I refer to. You can't simply dismiss the EAC number because it does still matter to copper people. You don't have consensus, you have bifurcation.
Holy....Can't you read?
That's exactly what I posted! EAC grading is not relevant. It is different from the rest of the world! The FACT that an auction company has finally printed both grades for copper - probably to help eliminate problems/returns, proves it!
@Insider2 said:
I know you don't know anything about the coin market. [EXACTLY! I wish I could tell you what I found out this weekend about the coin market and...] But go to the Heritage large cents section of their auction and you will note that they usually quote both the EAC and PCGS/NGC grade. This is the conundrum I refer to. You can't simply dismiss the EAC number because it does still matter to copper people. You don't have consensus, you have bifurcation.
Holy....Can't you read?
That's exactly what I posted! EAC grading is not relevant. It is different from the rest of the world! The FACT that an auction company has finally printed both grades for copper - probably to help eliminate problems/returns, proves it!
I can read. Now your turn. EAC grading IS RELEVANT which is why Heritage gives the EAC grade in addition to the PCGS/NGC grade. If it weren't relevant, why would they also give the EAC number which is not on the slab.
I'm glad you are LOLing and ignoring the point, but just because your almighty POE won't consider the EAC grade is irrelevant. The MARKET does consider the EAC grade.
The fact that we're even having the argument proves that your POE won't end the argument.
I happen to equate grade inflation with monetary inflation to a degree. After all the prices on all but the best coins have remained fairly lineal while the grades have increased.
@Insider2 said:
... However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
Interesting example you chose, there.
Just about any coin professionally graded by a TPGS since the late 70's (through 1995 and much later) has increased in grade. In 1999, I was told it is because the very experienced and knowledgeable professionals doing the grading have learned more over time by seeing more coins so that they have become even more experienced and knowledgeable. I picked an 1804 dollar as this grade escalation can be easily demonstrated and coins as this may be in the first group that should be "computer assisted" graded.
Fact: A group could be chosen.
Fact: CAC would be eliminated as it would no longer be needed.
Fact: EAC will still have an opinion on the "set-in-stone" computer recorded grade. Everyone including you and I will have an opinion on the "group" grade. That's fine and good - Just as we do now for all slabs. However, the computer grade (with possibility of group review for any obvious error but hard to imagine) will not change over time as assigned grades do now.
PS I'm enjoying our fantasy solution and your questions. I've been through all this a very long time ago when I developed the "true" technical system never used outside of Washington, DC.
@Insider2 said:
Fact: A group could be chosen.
Fact: CAC would be eliminated as it would no longer be needed.
Fact: EAC will still have an opinion on the "set-in-stone" computer recorded grade. Everyone including you and I will have an opinion on the "group" grade. That's fine and good - Just as we do now for all slabs. However, the computer grade (with possibility of group review for any obvious error but hard to imagine) will not change over time as assigned grades do now.
PS I'm enjoying our fantasy solution and your questions. I've been through all this a very long time ago when I developed the "true" technical system never used outside of Washington, DC.
Could someone contact Mr. Albanese and see how he feels about Fact #2?
I'm not saying you couldn't develop such a system. I'm saying such a system just becomes another voice. It doesn't end the disagreements.
Inevitably, machine learning will come to the TPGs and could eliminate the graders (excepting the finalizer). If graders wages are 250K + 50K in benefit cost each, that’s 900K annually in savings.
@Insider2 said:
Fact: A group could be chosen.
Fact: CAC would be eliminated as it would no longer be needed.
Fact: EAC will still have an opinion on the "set-in-stone" computer recorded grade. Everyone including you and I will have an opinion on the "group" grade. That's fine and good - Just as we do now for all slabs. However, the computer grade (with possibility of group review for any obvious error but hard to imagine) will not change over time as assigned grades do now.
PS I'm enjoying our fantasy solution and your questions. I've been through all this a very long time ago when I developed the "true" technical system never used outside of Washington, DC.
Could someone contact Mr. Albanese and see how he feels about Fact #2?
I'm not saying you couldn't develop such a system. I'm saying such a system just becomes another voice. It doesn't end the disagreements.
First, JA would be on the panel.
As we both posted - NOTHING will end disagreements. Everyone has an opinon; however, some opinion actually count.
Fact; If you and I were at a table with all our lights and "stuff," I'll guarantee we would be able to agree on each + or - that we found on the coin and then assign a number or adjectives to describe the coin's condition of preservation from the day it was made. With a good set of eyes, and adequate instruction, anyone can do it. They could look at a coin and describe everything they see on it.
The major problem with commercial grading comes when we try to adjust what we actually see on the coin in order to place a realistic value on it with a grade. Even if we don't try to value the coin, subjectivity and ignorance play a part due to the qualifications of the examiner- their eyesight, tools, knowledge, etc. With that, try to hit the moving target of gradflation.
The "assisted computer grading" would work because once the "experts" grade the coin it is done. What anyone else thinks does not matter or change anything. A hundred folks could send the coin in to the TPGS a hundred times over the decades yet get the same result. Computer Grading. LOL.
PS A coin's surface would need to be highly altered to receive a different result. I assure you that even a large majority of MS-70 SE coins, even have microscopic differences.
@Catbert said:
Inevitably, machine learning will come to the TPGs and could eliminate the graders (excepting the finalizer). If graders wages are 250K + 50K in benefit cost each, that’s 900K annually in savings.
I wonder how much over 900K it will cost to be able to place a coin in to a box and get a grade?
Grading is something I have always considered to be an opinion. I never quite bought into the idea that grading is an art or anything close to a science. There can be an art to using knowledge of grading systems to buy and sell, but that is different than a grading opinion itself.
While there are standards and some aspects can be objective, it will always be subjective to some extent because people have differing points of view.
Computer grading was pretty thoroughly argued about in the "letters to the editors" to the weekly coin papers in the 1990's. A key point was that someone or group had to put their standards into the program to begin with. Right there, others are going to disagree on that and about how every factor should be handled. (Toss in the revisions to the program version for more complexity which is similar to "was this slabbed while the grading was loose or tight?".)
This all would just lead to different brands of computer grading, just as there are different brands of grading services now. Nothing would actually change, except less resubmission revenue for grading companies if each coin consistently gets only one grade ever.
It may be possible to have the same program grade the same unchanged coin the same way repeatedly. But that does not mean that the market or any particular person or group of people will agree with that grade.
There is no computer program that can tell you what you like in a coin.
In the future I expect these grading discussions to pop up again and again just as they have in the past. That is the one thing that will always be consistent.
Next up, let's create a computer program to tell us which belief systems and political views are the right ones. I'm sure no one will disagree with the results because it's an objective computer with no human input or bias, right?. Surely that's safer than coin grading arguments.
"To Be Esteemed Be Useful" - 1792 Birch Cent --- "I personally think we developed language because of our deep need to complain." - Lily Tomlin
@ricko said:
That patent was a glimpse of what the future must eventually become. It is possible, and, IMO, inevitable that the system will become reality, Objectivity must replace subjectivity if grading is to be respected. Just witness the disagreements we constantly see here....Witness the 'upgrades' we see constantly....Standards must become reality. Cheers, RickO
I think you are correct. But....how can a computer subjectively evaluate TARNISH? I suspect that you don't care. I know I don't!
I'm in healthcare and they've had computer assisted diagnosis of radiology images for years. Computer assisted grading of coins doesn't seem that unreasonable and what could it find that might escape an actual in hand grading? It may also be able to detect substances or lack thereof too. It's not far fetched when you think about it.
@WinLoseWin said:
Grading is something I have always considered to be an opinion. I never quite bought into the idea that grading is an art or anything close to a science. There can be an art to using knowledge of grading systems to buy and sell, but that is different than a grading opinion itself.
While there are standards and some aspects can be objective, it will always be subjective to some extent because people have differing points of view.
Computer grading was pretty thoroughly argued about in the "letters to the editors" to the weekly coin papers in the 1990's. A key point was that someone or group had to put their standards into the program to begin with. Right there, others are going to disagree on that and about how every factor should be handled. (Toss in the revisions to the program version for more complexity which is similar to "was this slabbed while the grading was loose or tight?".)
This all would just lead to different brands of computer grading, just as there are different brands of grading services now. Nothing would actually change, except less resubmission revenue for grading companies if each coin consistently gets only one grade ever.
It may be possible to have the same program grade the same unchanged coin the same way repeatedly. But that does not mean that the market or any particular person or group of people will agree with that grade.
There is no computer program that can tell you what you like in a coin.
In the future I expect these grading discussions to pop up again and again just as they have in the past. That is the one thing that will always be consistent.
Next up, let's create a computer program to tell us which belief systems and political views are the right ones. I'm sure no one will disagree with the results because it's an objective computer with no human input or bias, right?. Surely that's safer than coin grading arguments.
Here is something to think about...
The more strict a grading system is the more precise it will be. This old story: "While there are standards [they are too loose, not followed, and have been allowed to change over time with a wink & nod] and some aspects can be objective, it will always be subjective. [every part of the grading computation IS objective (it can be measured and counted) except for the magnification and lighting; skill/knowledge of the examiner; value; and most of all - EYE APPEAL!
Something else...
In 1986 the big wheels in the industry started PCGS. If computer assisted grading as I imagined and wrote about with all the "players" we agree on involved was done under the blanket and endorsed as the "final word" by the ANA and it's members,there would be no NGC. any competition, or any CAC.
Finally this: Just as with grading, most of the political disagreement in this country comes about because folks are either not informed or misinformed. We are products of our training and environment. JA or anyone else here was not born with a great "eye." Folks cannot be blamed for their beliefs because unlike grading class/guides there is no way for a lazy or busy person to form an informed opinion that may differ from what they learned.
@Aotearoa said:
FWIW, it's been a long time since I've noticed a new Heritage listing with an EAC grade.
I just checked Heritage’s upcoming December U.S. coin sale and EAC grades are included.
OK. First time in a while then. I've been checking out essentially every DBLC listing for quite a while.
They have the EAC grades in all their major sales. I don't think they have the EAC grades in the weekly auction.
Why do you think that is? Did my guess, avoid problems, make sense? If the major dealers and collectors of copper use EAC "Net" grading I should think they would attract more bidders to the auction.
@Aotearoa said:
FWIW, it's been a long time since I've noticed a new Heritage listing with an EAC grade.
I just checked Heritage’s upcoming December U.S. coin sale and EAC grades are included.
OK. First time in a while then. I've been checking out essentially every DBLC listing for quite a while.
They have the EAC grades in all their major sales. I don't think they have the EAC grades in the weekly auction.
Why do you think that is? Did my guess, avoid problems, make sense? If the major dealers and collectors of copper use EAC "Net" grading I should think they would attract more bidders to the auction.
I think it is probably set-up time. The EAC grades must be generated in-house, after all. They might not have time to bother for the weekly auctions. But for the pricier coins in the major auctions, it is necessary, apparently.
Why do they put the EAC grade in the major auction: "Why do you think that is? Did my guess, avoid problems, make sense? If the major dealers and collectors of copper use EAC "Net" grading I should think they would attract more bidders to the auction."
I'm not going to get into a pi**ing contest with anyone. The fact remains that there is too much money involved in numismatics for a universally agreed upon grading system for coins to take place. It's similar to the concept of the limit in calculus.
"Vou invadir o Nordeste, "Seu cabra da peste, "Sou Mangueira......."
@Insider2 said:
Why do they put the EAC grade in the major auction: "Why do you think that is? Did my guess, avoid problems, make sense? If the major dealers and collectors of copper use EAC "Net" grading I should think they would attract more bidders to the auction."
Why don't you ask them? Clearly, copper folks care.
@Elcontador said:
I'm not going to get into a pi**ing contest with anyone. The fact remains that there is too much money involved in numismatics for a universally agreed upon grading system for coins to take place. It's similar to the concept of the limit in calculus.
From my selfish point of view, this entire discussion is about:
A Computer grading patent.
Computer grading that does not work - so far - and into the limited future of my lifetime.
A computer could be used to ID coins that are "correctly" graded enough to eliminate any knowledgeable disagreement.
The obstacles for something like this are not technical. Many people have the ability to create a fully automated grading system. The greater challenge is to get the world to adopt this new system.
@neildrobertson said:
The obstacles for something like this are not technical. Many people have the ability to create a fully automated grading system. The greater challenge is to get the world to adopt this new system.
Have any names for the "MANY PEOPLE you know of?"
BTW, IMHO, if one of your "many people" come up with this mythical "magic" machine (MMM) that actually works good enough to be accepted, a major TPGS would probably want to purchase the rights to control/use it.
@neildrobertson said:
The obstacles for something like this are not technical. Many people have the ability to create a fully automated grading system. The greater challenge is to get the world to adopt this new system.
Have any names for the "MANY PEOPLE you know of?"
I could probably do much of it. Lots of tricky bits, though. The biggest is data. I would have to have enough data for every type of coin to be able to classify it into it's appropriate grade. The more fine-grained the difference between grades (e.g., 69-70 as opposed to 63-64 or 10-12), the more data would be needed. Oh, and there's more to data than just TrueViews and a corresponding grade. The data has to be able to fully represent wear, surfaces, strike, luster, eye appeal, and problems. This includes stuff like hairlines that hide in the light, but can take 4 points off a proof grade or bodybag a MS coin. Once all that data is available to train a grading system, then in order to actually use it, you'd have to be able to acquire the same data for your subject coins.
As for being "good enough," that's another story. I've heard from a retired grader that something like 75-77% correct is the goal for graders. If you're going through all the trouble to build this system, you'd want over 90%. Technically achievable, I suppose, but a big task.
A better automated task might be related to another one of @Insider2 's threads -- detecting AT coins. Much easier task. Don't need as much training data and not as sophisticated data acquisition.
If two graders can make 75% each and the finalizer checks them I should think with the overlap that the rate would be at least 85% to 90%.
Something interesting to me is when someone shows me a PCGS or NGC slab I can hit their grade 95% of the time OR explain why my grade was different. I don't do as well guessing an ICG grade; yet I helped grade the coin! I cannot guess ANACS grades consistently - possibly because I rarely see them.
The R&D and Marketing involved would be orders of magnitude higher than that of a normal human eye grading service. I think most would agree to that.
A limited and stagnant market
How many active submitters are there? Is that number likely to grow in five years, ten years, twenty years?
Expected Return
will it be profitable? where is the break even point?
Longevity
If this works then the majority of coins worth grading will be graded with the non changeable grade within a few years. Customer base will shrink as material with non changeable grades decrease...
This is a good idea but it looks to be self defeating to me.
As I've stated before which might have been lost in the discussion, a computer most likely (probably 95% of the time,) diagnosed your loved one's mammogram before the radiologist even sees the images. It highlights areas that it believes are abnormal and provides an initial diagnosis. I don't think it would be difficult to duplicate a similar thing for coin grading with a finalizer acting like the radiologist and confirming the grade/diagnosis
I've seen things that computers are doing now and seen presentations by futurists. Any routine work done by a human will be taken over by computers in the future. Its already happened in some industries. Who would have thought of computers driving cars but that will happen!
jmlanzaf
Posts: 31,997 ✭✭✭✭✭
Comments
Who holds it?
Compugrade Inc, it says. 1987.
Yes, I think it's one of those many early Computer Age dreams that still hasn't been realized.
... and never will be.
I tend to agree. It's not that you couldn't get a computer to generate an evaluation. But your subjective eye would never agree with its objective evaluation. And if you give the AI subjectivity, I'm not sure what you have but a tempermental AI. LOL
Charles Hoskins was one of their consultants helping to develop this system. He was the first Director of the ANA's Certification in DC and chose to stay in the DC area when the company was moved to CO. Shortly after, Hoskins became one of the originators of the International Numismatic Society and Director of that organization's Authentication Bureau which later became the first grading service in the U.S.
IMO, the Compugrade system was a failure. Another attempt after that did not appear to be successful either.
IMO, any more money chasing this "dream" of machine grading will be a $$$,$$$$ loss.
However, if a small group of professionals meet to assign a grade on EACH coin - beginning with true rarities such as 1804 dollars, and then "fingerprint" that grade using computer imaging so the grade will never change, they can claim to have a "computer assisted" professionally graded coin.
Unfortunately, this might change and limit TPGS profitability over time.
Interesting example you chose, there.
Keeper of the VAM Catalog • Professional Coin Imaging • Prime Number Set • World Coins in Early America • British Trade Dollars
Who are these experts who are so absolutely expert that we should trust their consensus opinion for all time?
What if this "small group" are not unanimous in their evaluation?
Given that TPGS grading is market grading not technical grading, doesn't it have to evolve with the market?
Personally, I'd be glad for all those MS coins with rub to be downgraded to AUs, but I don't think it will be just the TPGS's that are unhappy.
Just about any coin professionally graded by a TPGS since the late 70's (through 1995 and much later) has increased in grade. In 1999, I was told it is because the very experienced and knowledgeable professionals doing the grading have learned more over time by seeing more coins so that they have become even more experienced and knowledgeable. I picked an 1804 dollar as this grade escalation can be easily demonstrated and coins as this may be in the first group that should be "computer assisted" graded.
I would absolutely love to get the human element out of grading in numismatics. No more bad Mondays or Good Fridays or hangovers, or being PO'd due to arguments with the wife, or bias of types of coins, toning, etc, etc... Set "and agree" on the standards, (there's the rub) develop and "release the app". Of course PCGS, NGC and CAC aren't going to be advocates.
Do you live under a rock? Who do you think they would be? I'll bet if I offered you $100 to name just a few "experts" that could do a wonderful job, you could spit out a dozen that we would all agree with in the time it took me to one finger the rest of these letters including the punctuation and emoji!
That patent was a glimpse of what the future must eventually become. It is possible, and, IMO, inevitable that the system will become reality, Objectivity must replace subjectivity if grading is to be respected. Just witness the disagreements we constantly see here....Witness the 'upgrades' we see constantly....Standards must become reality. Cheers, RickO
Successful transactions with : MICHAELDIXON, Manorcourtman, Bochiman, bolivarshagnasty, AUandAG, onlyroosies, chumley, Weiss, jdimmick, BAJJERFAN, gene1978, TJM965, Smittys, GRANDAM, JTHawaii, mainejoe, softparade, derryb
Bad transactions with : nobody to date
A coin's assigned grade by a panel of "experts" would have ABSOLUTELY NOTHING to do with the market. Assign a grade, image the coin, and that dollar, once graded XF, and now graded MS-62 in 2019 will still be graded MS-62 fifty years from now. Only its price/value will have changed both up and down over the years in between!
I guarantee that you could NOT find a list everyone agreed on. I simply point out the CAC controversy. JA would be on anyone's short list, but that doesn't mean they want to give him and his pals FINAL word on a coin.
I point you to EAC grades vs. TPGS grades. The conundrum already exists. You start down-grading all the "market acceptible" 63 with rub down to an AU and you will have chaos. What you'll end up with is a 2nd or 3rd number: you'll have a PCGS number, a "panel of expert" number and them maybe an EAC number.
I point you to the Husak collection where this actually happened.
Keeper of the VAM Catalog • Professional Coin Imaging • Prime Number Set • World Coins in Early America • British Trade Dollars
I'll choose to ignore your point as EAC grading is stupid. Net grading will not be tolerated by the "expert' group assigning the "forever grade" put into the computer.
EAC GRADING is a compartmentalize, offshoot of normal grading devised by some unknown person for a small group of copper specialists with (by their own admission in print) absolutely no relationship to the way coins are graded by the TPGS's and everyone else.
They have their system. It is complicated and while it is not very precise by its nature (the same goes for commercial grading) it works for them.
Nevertheless, They have folks who would be in the group of "experts" because their opinion of a coin's grade is excellent UNTIL the "net" crap they do "kicks in" causing a different outcome from the rest of us.
I know you don't know anything about the coin market. But go to the Heritage large cents section of their auction and you will note that they usually quote both the EAC and PCGS/NGC grade. This is the conundrum I refer to. You can't simply dismiss the EAC number because it does still matter to copper people. You don't have consensus, you have bifurcation.
Adding your own POE (panel of experts) will just become an alternative voice to the already crowded field. And throw in CAC, you will really have chaos. When your panel of experts assigns an MS62 to a coin PCGS calls MS64 and CAC beans the 62 but not the 64, is the coin really a 63? If your POE lacks sufficient market clout to shut down PCGS and NGC and maybe CAC as going entities, you simply have an option to PCGS/NGC that the market may or may not embrace. If a PCGS CAC sells for more than a POE non-CAC, people will continue to send to PCGS. Because, hey, y'know, if you are selling a coin, the market grade is going to matter more than the technical grade.
I guarantee that a poll done on CU to pick out a dozen numismatists that would set the final, "computer-recorded grade" on a coin for all time could be done - possibly at the embarrassment of all concerned.
I'd also point out this fact: There are expert graders who specialize in a particular series. Additionally, there are folks I would put up against the "best" professionals in the business that no one has ever heard of! It worked that way with our authentication consultants. I have found that long-time professionals, admired and considered to be "experts" by the whole world couldn't reliably authenticate themselves out of a paper bag! LOL!
I know you don't know anything about the coin market. [EXACTLY! I wish I could tell you what I found out this weekend about the coin market and...] But go to the Heritage large cents section of their auction and you will note that they usually quote both the EAC and PCGS/NGC grade. This is the conundrum I refer to. You can't simply dismiss the EAC number because it does still matter to copper people. You don't have consensus, you have bifurcation.
Holy....Can't you read?
That's exactly what I posted! EAC grading is not relevant. It is different from the rest of the world! The FACT that an auction company has finally printed both grades for copper - probably to help eliminate problems/returns, proves it!
I can read. Now your turn. EAC grading IS RELEVANT which is why Heritage gives the EAC grade in addition to the PCGS/NGC grade. If it weren't relevant, why would they also give the EAC number which is not on the slab.
I'm glad you are LOLing and ignoring the point, but just because your almighty POE won't consider the EAC grade is irrelevant. The MARKET does consider the EAC grade.
The fact that we're even having the argument proves that your POE won't end the argument.
I happen to equate grade inflation with monetary inflation to a degree. After all the prices on all but the best coins have remained fairly lineal while the grades have increased.
Fact: A group could be chosen.
Fact: CAC would be eliminated as it would no longer be needed.
Fact: EAC will still have an opinion on the "set-in-stone" computer recorded grade. Everyone including you and I will have an opinion on the "group" grade. That's fine and good - Just as we do now for all slabs. However, the computer grade (with possibility of group review for any obvious error but hard to imagine) will not change over time as assigned grades do now.
PS I'm enjoying our fantasy solution and your questions.
I've been through all this a very long time ago when I developed the "true" technical system never used outside of Washington, DC.
Could someone contact Mr. Albanese and see how he feels about Fact #2?
I'm not saying you couldn't develop such a system. I'm saying such a system just becomes another voice. It doesn't end the disagreements.
Inevitably, machine learning will come to the TPGs and could eliminate the graders (excepting the finalizer). If graders wages are 250K + 50K in benefit cost each, that’s 900K annually in savings.
First, JA would be on the panel.
As we both posted - NOTHING will end disagreements. Everyone has an opinon; however, some opinion actually count.
Fact; If you and I were at a table with all our lights and "stuff," I'll guarantee we would be able to agree on each + or - that we found on the coin and then assign a number or adjectives to describe the coin's condition of preservation from the day it was made. With a good set of eyes, and adequate instruction, anyone can do it. They could look at a coin and describe everything they see on it.
The major problem with commercial grading comes when we try to adjust what we actually see on the coin in order to place a realistic value on it with a grade. Even if we don't try to value the coin, subjectivity and ignorance play a part due to the qualifications of the examiner- their eyesight, tools, knowledge, etc. With that, try to hit the moving target of gradflation.
The "assisted computer grading" would work because once the "experts" grade the coin it is done. What anyone else thinks does not matter or change anything. A hundred folks could send the coin in to the TPGS a hundred times over the decades yet get the same result.
Computer Grading. LOL.
PS A coin's surface would need to be highly altered to receive a different result. I assure you that even a large majority of MS-70 SE coins, even have microscopic differences.
I wonder how much over 900K it will cost to be able to place a coin in to a box and get a grade?
Grading is something I have always considered to be an opinion. I never quite bought into the idea that grading is an art or anything close to a science. There can be an art to using knowledge of grading systems to buy and sell, but that is different than a grading opinion itself.
While there are standards and some aspects can be objective, it will always be subjective to some extent because people have differing points of view.
Computer grading was pretty thoroughly argued about in the "letters to the editors" to the weekly coin papers in the 1990's. A key point was that someone or group had to put their standards into the program to begin with. Right there, others are going to disagree on that and about how every factor should be handled. (Toss in the revisions to the program version for more complexity which is similar to "was this slabbed while the grading was loose or tight?".)
This all would just lead to different brands of computer grading, just as there are different brands of grading services now. Nothing would actually change, except less resubmission revenue for grading companies if each coin consistently gets only one grade ever.
It may be possible to have the same program grade the same unchanged coin the same way repeatedly. But that does not mean that the market or any particular person or group of people will agree with that grade.
There is no computer program that can tell you what you like in a coin.
In the future I expect these grading discussions to pop up again and again just as they have in the past. That is the one thing that will always be consistent.
Next up, let's create a computer program to tell us which belief systems and political views are the right ones. I'm sure no one will disagree with the results because it's an objective computer with no human input or bias, right?. Surely that's safer than coin grading arguments.
"To Be Esteemed Be Useful" - 1792 Birch Cent --- "I personally think we developed language because of our deep need to complain." - Lily Tomlin
FWIW, it's been a long time since I've noticed a new Heritage listing with an EAC grade.
Smitten with DBLCs.
I just checked Heritage’s upcoming December U.S. coin sale and EAC grades are included.
Mark Feld* of Heritage Auctions*Unless otherwise noted, my posts here represent my personal opinions.
I think you are correct. But....how can a computer subjectively evaluate TARNISH? I suspect that you don't care. I know I don't!
OK. First time in a while then. I've been checking out essentially every DBLC listing for quite a while.
Smitten with DBLCs.
I'm in healthcare and they've had computer assisted diagnosis of radiology images for years. Computer assisted grading of coins doesn't seem that unreasonable and what could it find that might escape an actual in hand grading? It may also be able to detect substances or lack thereof too. It's not far fetched when you think about it.
Here is something to think about...
The more strict a grading system is the more precise it will be. This old story: "While there are standards [they are too loose, not followed, and have been allowed to change over time with a wink & nod] and some aspects can be objective, it will always be subjective. [every part of the grading computation IS objective (it can be measured and counted) except for the magnification and lighting; skill/knowledge of the examiner; value; and most of all - EYE APPEAL!
Something else...
In 1986 the big wheels in the industry started PCGS. If computer assisted grading as I imagined and wrote about with all the "players" we agree on involved was done under the blanket and endorsed as the "final word" by the ANA and it's members,there would be no NGC. any competition, or any CAC.
Finally this: Just as with grading, most of the political disagreement in this country comes about because folks are either not informed or misinformed. We are products of our training and environment. JA or anyone else here was not born with a great "eye." Folks cannot be blamed for their beliefs because unlike grading class/guides there is no way for a lazy or busy person to form an informed opinion that may differ from what they learned.
They have the EAC grades in all their major sales. I don't think they have the EAC grades in the weekly auction.
Why do you think that is? Did my guess, avoid problems, make sense? If the major dealers and collectors of copper use EAC "Net" grading I should think they would attract more bidders to the auction.
I think it is probably set-up time. The EAC grades must be generated in-house, after all. They might not have time to bother for the weekly auctions. But for the pricier coins in the major auctions, it is necessary, apparently.
Why do they put the EAC grade in the major auction: "Why do you think that is? Did my guess, avoid problems, make sense? If the major dealers and collectors of copper use EAC "Net" grading I should think they would attract more bidders to the auction."
You two guys are like oil and water!
I'm not going to get into a pi**ing contest with anyone. The fact remains that there is too much money involved in numismatics for a universally agreed upon grading system for coins to take place. It's similar to the concept of the limit in calculus.
"Seu cabra da peste,
"Sou Mangueira......."
Why don't you ask them? Clearly, copper folks care.
From my selfish point of view, this entire discussion is about:
The obstacles for something like this are not technical. Many people have the ability to create a fully automated grading system. The greater challenge is to get the world to adopt this new system.
IG: DeCourcyCoinsEbay: neilrobertson
"Numismatic categorizations, if left unconstrained, will increase spontaneously over time." -me
Have any names for the "MANY PEOPLE you know of?"
BTW, IMHO, if one of your "many people" come up with this mythical "magic" machine (MMM) that actually works good enough to be accepted, a major TPGS would probably want to purchase the rights to control/use it.
I could probably do much of it. Lots of tricky bits, though. The biggest is data. I would have to have enough data for every type of coin to be able to classify it into it's appropriate grade. The more fine-grained the difference between grades (e.g., 69-70 as opposed to 63-64 or 10-12), the more data would be needed. Oh, and there's more to data than just TrueViews and a corresponding grade. The data has to be able to fully represent wear, surfaces, strike, luster, eye appeal, and problems. This includes stuff like hairlines that hide in the light, but can take 4 points off a proof grade or bodybag a MS coin. Once all that data is available to train a grading system, then in order to actually use it, you'd have to be able to acquire the same data for your subject coins.
As for being "good enough," that's another story. I've heard from a retired grader that something like 75-77% correct is the goal for graders. If you're going through all the trouble to build this system, you'd want over 90%. Technically achievable, I suppose, but a big task.
A better automated task might be related to another one of @Insider2 's threads -- detecting AT coins. Much easier task. Don't need as much training data and not as sophisticated data acquisition.
Keeper of the VAM Catalog • Professional Coin Imaging • Prime Number Set • World Coins in Early America • British Trade Dollars
If two graders can make 75% each and the finalizer checks them I should think with the overlap that the rate would be at least 85% to 90%.
Something interesting to me is when someone shows me a PCGS or NGC slab I can hit their grade 95% of the time OR explain why my grade was different. I don't do as well guessing an ICG grade; yet I helped grade the coin! I cannot guess ANACS grades consistently - possibly because I rarely see them.
issues with @Insider2 business model
The R&D and Marketing involved would be orders of magnitude higher than that of a normal human eye grading service. I think most would agree to that.
How many active submitters are there? Is that number likely to grow in five years, ten years, twenty years?
will it be profitable? where is the break even point?
If this works then the majority of coins worth grading will be graded with the non changeable grade within a few years. Customer base will shrink as material with non changeable grades decrease...
This is a good idea but it looks to be self defeating to me.
As I've stated before which might have been lost in the discussion, a computer most likely (probably 95% of the time,) diagnosed your loved one's mammogram before the radiologist even sees the images. It highlights areas that it believes are abnormal and provides an initial diagnosis. I don't think it would be difficult to duplicate a similar thing for coin grading with a finalizer acting like the radiologist and confirming the grade/diagnosis
I've seen things that computers are doing now and seen presentations by futurists. Any routine work done by a human will be taken over by computers in the future. Its already happened in some industries. Who would have thought of computers driving cars but that will happen!