Member thoughts: "...coin grading is a subjective art, and not an objective science...." 1 1

[RWB]

Is the title statement correct? Is it an "art" by any standard?

Can objective and subjective be separated? Would that return integrity to independent "grading?"

[The title quote is by grader Don Bonser in a PCGS post earlier today.]

Edited April 20, 2022 by RWB

[Coinbuf]

Coin grading was closer to an objective science when the ANA and TPG's first started, so yes, coin grading can be as simple as give me a grade biased on the coin's level of preservation and condition and leave the personal opinion out.    However, that is not what we have and I'm not at all certain that most of the market would want to return to that system.

 

[VKurtB]

I like the system as it is right now. “Nick picking” grading drives me nuts. I hate it. The four criteria now used in MS coins are all necessary. 

[physics-fan3.14]

Coin grading is both an art and a science. (Hence the title of my book )

There is some objectivity in the strike, luster, contact marks, wear, etc. 

But, there is some subjectivity in the evaluation of eye appeal. 

You have to have both for the current method of grading (sure, old timers talk about "technical grading" or EAC grading, or things like that - but you have to recognize that mainstream coin grading in the current market is not that). 

[FlyingAl]

I agree with the standpoint that it is both.

There is an objective sense in that one has to be able to identify something like counterfeit or genuine, die polish vs. hairlines and such, but a lot (majority) of coin grading is subjective.

It should also remain this way. For example, a computer cannot be easily programmed to identify the difference between positive and negative eye appeal. As such, there is really no way to eliminate human error as a human really does have to grade coins, and as such, the verdicts produced will have some degree of error. 

[RWB]

On 4/20/2022 at 5:23 PM, FlyingAl said:

....a computer cannot be easily programmed to identify the difference between positive and negative eye appeal.

True. But that is, after all, an opinion - something not easily enabled in a machine. Would you want it too, anyway?

[Woods020]

I agree that it is, and honesty should be, both to some degree. However my biggest ask is consistency. If all coins are graded with the same criteria I can adjust to the criteria. It’s when it’s inconsistent that the problems occur. 

[VKurtB]

Consistency is fine…, as long as it works well in practice. Grading schemes change over time for one reason - they create their own problems in the workability department. Once we find a system that works FOR THE MARKETPLACE (not the dogmatic theoreticians, who can go to h—-), then consistency becomes a pure good. Until then, all consistency does is lock in yesterday’s errors. 
 

I think we’re pretty close now, but we were nowhere close in the mid-1980’s. 

Edited April 21, 2022 by VKurtB

[coastaljerseyguy]

I wouldn't underestimate the ability of programming & computers to gauge eye appeal in addition to the technical grade.  What is eye appeal anyway:

• Great strike above whatever criteria you want to input - this is measurable
• Great luster - measurable
• Appealing toning - images can be fed into the program to determine what the TPGs & collecting community feels looks great, good, ok, horrible and computers can do the combination/permutations
• Pure whiteness or lack of toning, spots, etc.

And a human can always do the final once over for approval and consistency.  All modern coinage should be done by computers.

   

[GoldFinger1969]

I think I read about computers being introduced with scanning DECADES ago -- not years ago -- but I am not sure what happened with that experiment or roll-out.  I suspect the technology wasn't up to speed.  I can't recall where I read that piece, but I know it dealt with a long time ago, not in recent years when we have affordable hi-def scanning capabilities.  I think I recall 1980 or 1990 but don't quote me.

I do like the idea of a hi-def scan recording all coins minute characteristics so that coins can be tracked and employed for anti-theft purposes.

 

[RWB]

A lack of standards in coin grading is a severe impediment to consumers recovering losses on over graded coins. Courts will not support claims without consistent standards.

[coastaljerseyguy]

Although I agree that today, coin grading is both objective science and a subjective art, I believe it can be a science or close to one.  The definition and classification of good/bad: art, music, beauty, etc. will always be an art since it is ever evolving, so unique,  and new forms coming into existence. Not so for coins. The universe of classic coins has been defined, will not change, and there is a close ideal of what is perfection.  Have all coins been reviewed, no.  But the proper classification based on standards can be defined & graded, including eye appeal, since thousands of existing images can be classified by the TPGs and professionals whose opinions we value.  These can then be imaged, programmed, and repeated.

Although this change would require significant time and money by the TPGs, there would appear to be a cost savings in the long run.  However there would also be the loss of revenue since regrades would be non-existent since 99.999% of the time the same grade would be returned every time.  God forbid.   

[Coinbuf]

On 4/21/2022 at 9:22 AM, coastaljerseyguy said:

Although I agree that today, coin grading is both objective science and a subjective art, I believe it can be a science or close to one.  The definition and classification of good/bad: art, music, beauty, etc. will always be an art since it is ever evolving, so unique,  and new forms coming into existence. Not so for coins. The universe of classic coins has been defined, will not change, and there is a close ideal of what is perfection.  Have all coins been reviewed, no.  But the proper classification based on standards can be defined & graded, including eye appeal, since thousands of existing images can be classified by the TPGs and professionals whose opinions we value.  These can then be imaged, programmed, and repeated.

Although this change would require significant time and money by the TPGs, there would appear to be a cost savings in the long run.  However there would also be the loss of revenue since regrades would be non-existent since 99.999% of the time the same grade would be returned every time.  God forbid.   

I agree that the technology exists to make computer grading a reality, however no company wants to willingly cut its own revenue stream.

[Coinbuf]

On 4/21/2022 at 8:03 AM, GoldFinger1969 said:

I think I read about computers being introduced with scanning DECADES ago -- not years ago -- but I am not sure what happened with that experiment or roll-out.  I suspect the technology wasn't up to speed.  I can't recall where I read that piece, but I know it dealt with a long time ago, not in recent years when we have affordable hi-def scanning capabilities.  I think I recall 1980 or 1990 but don't quote me.

I do like the idea of a hi-def scan recording all coins minute characteristics so that coins can be tracked and employed for anti-theft purposes.

 

There have been multiple attempts by current and past TPG's, some died on the vine like compugrade and others were most likely more about talking points than a true concerted effort.

[LINCOLNMAN]

My preference would be to assign a grade based solely on the state of preservation and strike, what I think is called, in part anyway, a "technical grade". The more subjective determinants of market price should depend on the taste of the buyer. This is what happens anyway, judging by the divergent prices realized for coins of similar grades at similar times in similar venues. This will never happen of course because most people want to be told what is "good" in all fields of life. Furthermore we have the moronic chase for registry status, which is based solely on a TPG grade, now with a CAC opinion thrown in (brilliant marketing schemes BTW). 

[GoldFinger1969]

On 4/21/2022 at 1:16 PM, Coinbuf said:

There have been multiple attempts by current and past TPG's, some died on the vine like compugrade and others were most likely more about talking points than a true concerted effort.

The cost and qualitative improvements today with Hi-Def and 4K optics has to be light-years better than even 20 years ago when the PC and internet age just started.

I don't know about AI being able to determine "luster" and "eye appeal" but I am sure that today's technology can image coins better than a human eye can and look for distinguishing blemishes, cracks, spots, etc. to uniquely identify each coin.

[Woods020]

I have been working with OCR (optical character recognition) tech for atleast 10 years now. We use it a lot for data entry of our high volume standardized forms. It wouldn’t be a leap at all for the coins to be imaged and logic to be created to grade them. I’m sure it could count the marks, locations of the marks, etc and make deductions. 
 

Where I haven’t seen it be possible would be the strength of strike or eye appeal as others have said. Now I’m not on the cutting edge of tech by any means so it may exist. But the standard OCR stuff wouldn’t be able to give opinions unless it was a programmable algorithm, which I don’t see for certain things.  

[VKurtB]

On 4/21/2022 at 10:36 AM, RWB said:

A lack of standards in coin grading is a severe impediment to consumers recovering losses on over graded coins. Courts will not support claims without consistent standards.

So it’s about suing someone? What a disgusting thought! BTW, there is nothing about cartwheel luster that can’t be quantified if a standard light source can be mandated. We have a problem, though. THERE IS A STANDARD LIGHT SOURCE, BUT VERY FEW USE IT. And I’d venture to say many collectors (and authors, even very busy ones) would never accept the industry’s pre-existing standard because they have their own they prefer. 

Edited April 23, 2022 by VKurtB