The NGC Census: a modest proposal. 1 1

[Henri Charriere]

In due course, NGC will be observing its 50th anniversary, or jubilee.  To mark the occasion, I should like to suggest a modest proposal: an overhaul of its census to revise or reaffirm the accuracy of its  number of certifications. Any ideas as to how to go about insuring this monumental, prodigious undertaking can be accomplished economically, expediently and with as near accurate results possible?

[GoldFinger1969]

On 10/30/2023 at 10:55 PM, Henri Charriere said:

In due course, NGC will be observing its 50th anniversary, or jubilee.  To mark the occasion, I should like to suggest a modest proposal: an overhaul of its census to revise or reaffirm the accuracy of its  number of certifications. Any ideas as to how to go about insuring this monumental, prodigious undertaking can be accomplished economically, expediently and with as near accurate results possible?

No idea...but isn't their 50th annivesary 2037 ?  They started operations in 1987. 

[ldhair]

I don't know much about the census. Is it in need of being overhauled? I would think it gets updated each time a coin gets graded. I don't think it's possible to know how many NGC coins have been removed from the holders. Is it even possible to reaffirm the data?

[Fenntucky Mike]

All TPG census are what they are, a basic resource that's probably around 80% accurate.

Crack-outs, Crossovers, Mechanical errors, etc...... They'll never stop and a census will never be what it could or should be.

[VKurtB]

On 10/31/2023 at 6:27 AM, ldhair said:

I don't know much about the census. Is it in need of being overhauled? I would think it gets updated each time a coin gets graded. I don't think it's possible to know how many NGC coins have been removed from the holders. Is it even possible to reaffirm the data?

No, it is impossible, as you guessed. Censuses (censi?) are inherently flawed by their very nature. The best we can hope for is approximation.

Edited October 31, 2023 by VKurtB

[VKurtB]

On 10/31/2023 at 6:33 AM, Fenntucky Mike said:

All TPG census are what they are, a basic resource that's probably around 80% accurate.

Crack-outs, Crossovers, Mechanical errors, etc...... They'll never stop and a census will never be what it could or should be.

Unfortunately "mechanical errors" (I HATE that term.) are exploding in frequency.

[GoldFinger1969]

On 10/31/2023 at 7:27 AM, ldhair said:

I don't know much about the census. Is it in need of being overhauled? I would think it gets updated each time a coin gets graded. I don't think it's possible to know how many NGC coins have been removed from the holders. Is it even possible to reaffirm the data?

Roger and other researchers will apply a "fudge factor" to eliminate double-counting.  It varies by coin type...by year...by how scarce a particular series is....but I recall him posting way back in the RWB Saints-Gaudens Book Thread that a 40% haircut was not out of whack to eliminate double-counting.  That was for a coin with hundreds or thousands in the count.

For a super-rare coin with a few dozen or less, it might be a smaller haircut I would guess.

[Fenntucky Mike]

On 10/31/2023 at 12:13 PM, VKurtB said:

Unfortunately "mechanical errors" (I HATE that term.) are exploding in frequency.

It's worse than ever and they are a double hit to a census, the wrong piece is recorded and the correct piece is not. 

[VKurtB]

On 10/31/2023 at 11:44 AM, Fenntucky Mike said:

It's worse than ever and they are a double hit to a census, the wrong piece is recorded and the correct piece is not. 

One of the slabs brought to the Summer Seminar course taught by NGC people last June had one such error. Embarrassing. Object lesson: DO NOT EXPECT PRECISION AT ANY STAGE OF GRADING AND ENCAPSULATION.

 

Edited October 31, 2023 by VKurtB

[Fenntucky Mike]

On 10/31/2023 at 12:16 PM, GoldFinger1969 said:

Roger and other researchers will apply a "fudge factor" to eliminate double-counting.  It varies by coin type...by year...by how scarce a particular series is....but I recall him posting way back in the RWB Saints-Gaudens Book Thread that a 40% haircut was not out of whack to eliminate double-counting.  That was for a coin with hundreds or thousands in the count.

For a super-rare coin with a few dozen or less, it might be a smaller haircut I would guess.

40% is not an outrageous number, especially if survivors are well documented, but for most coin series I would think the margin to be less. Opposite could be true as well for series with low survival rates or mintages, who knows you could have a coin with a mintage of 1 and it was labeled wrong, or cracked out multiple times. 

[Lem E]

I treat the census like I do red book prices. 

[VKurtB]

On 10/31/2023 at 12:09 PM, Lem E said:

I treat the census like I do red book prices. 

Both quote numbers that are often too high.

[zadok]

On 10/30/2023 at 10:55 PM, Henri Charriere said:

In due course, NGC will be observing its 50th anniversary, or jubilee.  To mark the occasion, I should like to suggest a modest proposal: an overhaul of its census to revise or reaffirm the accuracy of its  number of certifications. Any ideas as to how to go about insuring this monumental, prodigious undertaking can be accomplished economically, expediently and with as near accurate results possible?

...not going to happen, not a rational expectation....

[zadok]

On 10/31/2023 at 12:10 PM, VKurtB said:

No, it is impossible, as you guessed. Censuses (censi?) are inherently flawed by their very nature. The best we can hope for is approximation.

...a flock of census is a censuses, correct first time....

[zadok]

On 10/31/2023 at 12:16 PM, GoldFinger1969 said:

Roger and other researchers will apply a "fudge factor" to eliminate double-counting.  It varies by coin type...by year...by how scarce a particular series is....but I recall him posting way back in the RWB Saints-Gaudens Book Thread that a 40% haircut was not out of whack to eliminate double-counting.  That was for a coin with hundreds or thousands in the count.

For a super-rare coin with a few dozen or less, it might be a smaller haircut I would guess.

...maybe he can write a book bout it....

[Henri Charriere]

On 10/31/2023 at 2:08 AM, GoldFinger1969 said:

No idea...but isn't their 50th annivesary 2037 ?  They started operations in 1987. 

That is correct. As failure is not an option, there is plenty of time to come up with a viable solution.  🤔 

[Henri Charriere]

On 10/31/2023 at 7:27 AM, ldhair said:

I don't know much about the census. Is it in need of being overhauled? I would think it gets updated each time a coin gets graded. I don't think it's possible to know how many NGC coins have been removed from the holders. Is it even possible to reaffirm the data?

You are scrupulously honest. I suppose not many people know much about it and quite a few are oblivious to its critical importance. The census does get updated but it is in desperate need of an overhaul. What else can a collector or dealer rely upon to determine supply and demand? Mintages, for the most part, remain static subject to the discovery of hoards and meltings, many unauthorized as revealed by counts. If the census is to have meaning, it must be accurate.

[zadok]

On 10/31/2023 at 7:00 PM, Henri Charriere said:

You are scrupulously honest. I suppose not many people know much about it and quite a few are oblivious to its critical importance. The census does get updated but it is in desperate need of an overhaul. What else can a collector or dealer rely upon to determine supply and demand? Mintages, for the most part, remain static subject to the discovery of hoards and meltings, many unauthorized as revealed by counts. If the census is to have meaning, it must be accurate.

...inherently untrue...the census is merely a reflection of what has been submitted/certified n most numismatist believe that less than 20% of existing coins have been certified...even if censuses were accurate they would not provide a true picture of what is or is not out there...in some few areas, mostly the truly rare coins, e.g. 1878-S liberty seated half dollars the existing census, Bill Bugert's real research, provides a true picture of what is n is not out there...the tpg censuses provide exactly what they were intended to provide, a report of what has been submitted/certified, no overhauls required....

 

p.s. even the 1878-S half dollar census gets previously unknown examples added....

[VKurtB]

On 10/31/2023 at 6:00 PM, Henri Charriere said:

What else can a collector or dealer rely upon to determine supply and demand?

There simply IS NOTHING that a collector or dealer can rely upon to determine supply and demand, nor will there ever be. “Yo pays yo money n’ yo takes yo chances.”

Edited November 1, 2023 by VKurtB

[zadok]

On 10/31/2023 at 8:32 PM, VKurtB said:

There simply IS NOTHING that a collector or dealer can rely upon to determine supply and demand, nor will there ever be. “Yo pays yo money n’ yo takes yo chances.”

...otherwise known as the yo-yo yo-yo effect...or as they say in paris couillon-couillon....

[World Colonial]

On 10/31/2023 at 8:32 PM, VKurtB said:

There simply IS NOTHING that a collector or dealer can rely upon to determine supply and demand, nor will there ever be. “Yo pays yo money n’ yo takes yo chances.”

There are common factors that can be used to get an idea of whether a coin is likely (noticeably) more available (and therefore, common) than personal experience. 

Too many US collectors seem to think that the TPG population data is representative of the supply because they assume most coin owners are financially motivated or prefer slabbed coins when this isn't universally true.  The other common fallacy is assuming "If I don't see it or those I know don't either, it must not exist".

With non-US coinage, the TPG counts when low almost always are because of the low value and/or cultural dislike of coins in plastic.  When I resumed collecting in 1998 and first started looking extensively at non-US coinage, I also mistakenly believed that because some coin was "old", it was "scarce" or "rare".  Most of it isn't, often or usually in "high" quality too.

With US coinage, it's because of the price and the limited or lack of improved marketability.

Here is a list of factors I once provided on the PCGS forum, presumably not exhaustive, which anyone can use to evaluate whether a coin is likely as scarce as it appears from personal observation.

*Mintage (most are not actually low, only relatively; it's a mass-produced object)

*Extent of organized local collecting.  (Extensive in all developed countries, often for a long time)

*Geographic access, extent of travel, and ease of communication.  

*Local circulation, as opposed to other geographic locations with more collecting. (Correlated to a coin's age.  Substantially applicable to trade coinage.)

*Coin age combined with length of circulation.  ("Recent" coinage could have been saved in volume within the lifetime of just those reading this thread.)

*Quality when included in prominent collections of the series.  If a coin is missing from a prominent collection or is of low quality, it’s usually at least somewhat scarce.

*FV materiality 

*Melting, where it happened.  

*Hoards (it's subjective, but it doesn't take much to make a coin common.)

*Market value which if “low”, reduces the number for sale and probability potential buyers will be aware of it.  

*Random preservation; yes, it’s possible for some centuries old coin to sit in a “change jar”, just like ancient and medieval hoards.

*Comparison to coins with the closest characteristics where the number known is better quantified and/or with higher TPG population counts.  If coins confirmed to be at least somewhat scarce meet most of these factors, the coin someone believes scarce probably isn't.