Repost of one of my Coin Week Counterfeit Articles

Yes, you would have to have a database of known counterfeit lineages including the dates they were used, i.e. the counterfeit lineage would be generic to the coin type I guess. Same with known die pairs or known types, for example the 1900+ "wide shoulder gap" Morgan dollar reverse being used on a pre-1900 dollar. Problems would arise for discoveries that have not bubbled up to the TPGs. For example, the 1900-1901 dime reverse type is unknown outside of BCCS journal followers, and 1899-P dimes exist with that 1900 reverse - so the system would flag all those 1899s as counterfeit. Same with previously unknown die pairs, etc. Their "varieties" database would have to be more comprehensive, accurate, and kept up to date.

 

Log in or Sign up to hide this ad.

justafarmer, posted: "This would also require a grading standard that remained static."

It absolutely WOULD NOT!!!

There is no static grading standard NOW and thousands of coins are
"professionally" graded" every day. Once a coin is graded by "whatever the team of professional graders decide the grade is finalized into the computer forever UNTIL something is done to it to fool the computer or change its grade. Coins would need to be sent to another major TPGS for any chance to change the grade. If that TPGS also graded coins their way and "fingerprinted" the coin into a computer its grade would be set forever at that TPGS. That's what the obsolete "Technical System" approached. Coins were graded very strictly by their condition only. Any other factors that influenced its sight unseen ID were added. A pristine, flawless, well struck 1884-O was graded the same as as a pristine, flawless, flatly struck 84-O. One was graded MS-65 (the highest grade at the time) while the other was graded MS-65, Flat Strike. Both were absolutely virtually perfect as struck coins BUT ONE WAS WORTH A LOT MORE MONEY. Add splotchy, brown, low eye appeal to the flat coin and it became an MS-65, flatly struck, splotchy toned coin. Worth much less until it was dipped and the eye appeal was raised adding to its value again. We devised Technical Grading to ONLY IDENTIFY a coin, not to place a value on it. Coin dealers can do that.

It is virtually impossible to have two coins with the same "fingerprint." You and other authenticators use "fingerprint" to ID marks. While this is also a part of computer grading everything else on the coin counts for the computer. Tat includes luster, color, strike, etc. Thus the tiny black "flecks" on the coin in the left slab is all it takes tp differentiate it from the coin on the right.

 

Not to worry. Humans can change anything in the data base. Little known changes to coin types existing now (Trade Dollars, Barber Coins, etc) will eventually be recognized. Also remember that the computer is only an invaluable AID to human authenticators.

 

I'd guess that, with enough deep learning muscle (which might be less than we'd expect), an AI system would be able to say "this coin looks like that one with some additional toning, and it looks like someone dropped it on its edge -- as long as this image was taken later than the last one, it might be legit."

Computers see differently than humans, but they see much better than humans in most ways. AI is closing a lot of the remaining gaps.

 

You and I will have to disagree on this point. In your post you suggested the following

"A computer grading system can be done anytime since then. Professional graders grade the coin (very carefully this time), scan the coin in detail (the fingerprint) and that coin will be graded exactly the same FOREVER when ID'ed by the computer. "

My point being if a coin receives the exact same grade forever then the grading standard must remain the same forever. Else at some point in time (when the standard changes) the coin ends up getting saddled with an incorrect obsolete grade with no avenue of updating its grade based on the new standard.

 

I'm afraid you still don't get it. If I take a coin in G-4, put it in some kind of high power analyzer with never changing light type + intensity so that every single characteristic of the coin is mapped including shape, marks, metal composition,. thickness, weight, color, defects, reflectivity, etc. THE COIN WILL BE "FINGERPRINTED." In fact the AI is so good that I can put only 1/5 of the coin in the machine and it will still be properly ID'ed at the "forever grade." Now, that I have a record of that coin - I CAN CALL IT ANYTHING I WISH! Let's call it MS-65. for this exercise. So you and I can agree that the coin's grade will not change. THAT's A GOOD THING BECAUSE THAT WILL "cement" A STANDARD.

As for your disagreement pertaining to future changes in an evolving
professional grading NON-STANDARD as we have now, the AI will not allow it. When I try to tell the computer the G-4 above is to be called MS-65, it will spit it out to be regraded.

Grading is easy if all the fluff is first removed. I'll bet if I can take five of the best graders in the country and ask them to grade a twenty dollar Liberty VERY STRICTLY and ONLY by its condition of preservation (friction wear if any and non-Mint made marks marks - excluding color, strike, rariety, etc) they will get the same grade. The one that goes into the computer "fingerprint."

That way a beautiful, colorful, fantastic looking MS-64+ one-of-a-kind wonder coin will not be graded MS-68 or higher although it will be worth that money.

Grade the coin correctly, let the market put a value on it due to its "fluff."

 

I do understand. Perhaps I am not communicating very well. But if you issue a coin a forever grade and then subsequently down the road change the grading standard - at that point the coin will be forever stuck with an incorrect grade. Because the coin's grade was issued based on the old standard - not the new standard. This is why I posted the grading standard would also have to remain static/unchanged.

 

That is the bestest longest Baloney speechifying I have viewed in a very long time.

It is so good, that it ought to be placed on the CAC education forum, by you, today.

 

charley, posted: "That is the bestest longest Baloney speechifying I have viewed in a very long time. It is so good, that it ought to be placed on the CAC education forum, by you, today.

1. Your is the baloney. If you posted the truth about the "bestest" post, you would have given it a "like."

2. I guess that any know-nothing can get on the CAC forum if they would accept a post from me.

3. Your former place on this forum has been reactivated. I'm sure you will have some of the other clowns join you soon.

 

Awww, feeling left out, are we?

You are not being truthful with the audience.

You joined CAC Forums and posted some thoughts and were answered. In fact, you joined because I called on you to do so, and you answered the call. The commentary is still there. Don't try to portray yourself as Billy The Kid and proudly outlawed by all the numismatics forums because you speak truth to power. You are still a member.

Easy test: post there what you have eloquently shared on this thread. We can read in real time the thoughts of those in your profession. Why not? You have something to offer for an improvement to the future of numismatics. It is worth testing the idea, and more importantly, the future YNs read the CAC Education Forum regularly. Those YNs are important to the hobby. So....get posting .

 

WTF; why the same caustic level of replies in my posts @charley ? I truly get sick of these type of responses from you but you know that and here we go again...

 

Noted.

 

I put him back on ignore. It works wonders for what I'm able to see on the forum. I cannot recall reading anything of substance.

 

Sigh.....

 

Call it the Collective Black Cabinet.

There used to be a website but it no worky right now...

http://www.theblackcabinet.org/

 

I've got to look out my window. The pigs must be flying.

 

I have collected thousands of images on my worn out home computer and post them in forums like this as well all of my articles. If someone at ANA would be interested in this as a project please let me know!!!

 

Takes a lot of money to do what is being proposed. What business model can support that kind of capital expenditure?