How many books are still in existence? 1 1

[Flammenwerfer]

Is it possible to estimate how many Silver/Golden age books are still in existence, based on the number of recorded graded copies?

Edited August 4, 2022 by Flammenwerfer

[theCapraAegagrus]

Not with any degree of accuracy.

[O Doyle Rules]

On 8/4/2022 at 6:52 AM, Flammenwerfer said:

Is it possible to estimate how many Silver and Golden age books are still in existence based on the number of recorded graded copies?

Curious as to why you ask? 

[O Doyle Rules]

On 8/4/2022 at 6:52 AM, Flammenwerfer said:

Is it possible to estimate how many Silver and Golden age books are still in existence based on the number of recorded graded copies?

I would also think depending on how big a key the comic is, the fewer ungraded copies that are left of that particular issue. For example, on eBay there are many graded copies of X-Men #1 up for sale and only one ungraded copy for sale. 

[shadroch]

Many, if not most SA books are not worth slabbing, so for most books no.  I think it helps give you an idea of how many mega- keys there are, both by total number and by the yearly additions, but it isn't something for amatuers to get involved in. 

[revat]

If you took one of the phd-level data analysts at Disney/Marvel, and set them on this task with all available information, and gave them a few years, I bet they could come up with a supportable model, depending on the required degree of certainty.  But there would be a A LOT of assumptions made, and lot of those would be almost pure (if seemingly reasonable) guesses.

[ExNihilo]

On 8/4/2022 at 9:28 AM, revat said:

If you took one of the phd-level data analysts at Disney/Marvel, and set them on this task with all available information, and gave them a few years, I bet they could come up with a supportable model, depending on the required degree of certainty.  But there would be a A LOT of assumptions made, and lot of those would be almost pure (if seemingly reasonable) guesses.

Honestly, even the CGC census is inaccurate.  Think about how many people have removed books from slabs without telling CGC.  Arriving at an estimated figure would have to have some wide +/- range.

[Qalyar]

On 8/4/2022 at 11:28 AM, revat said:

If you took one of the phd-level data analysts at Disney/Marvel, and set them on this task with all available information, and gave them a few years, I bet they could come up with a supportable model, depending on the required degree of certainty.  But there would be a A LOT of assumptions made, and lot of those would be almost pure (if seemingly reasonable) guesses.

Well, I'm a data analyst, although not for Disney/Marvel, haha! Let's take a look at this as something like a Fermi problem. We won't get exact answers, but we can do better than "no idea!". This isn't super short.

We'll start with two of the most important comics of all time: Action Comics 1 and Detective Comics 27. Why? Well, because money talks. Basically, AC1 and Tec27 are worth so much money that the incentive to have them professionally graded and encapsulated is unusually high. And, importantly, that the monetary incentive is still ludicrously high regardless of the condition of the book (Tec27 in 0.5 is still worth sacks of cash, and so is worth grading; that's very different from even most more "normal" keys). That doesn't mean there aren't still raw copies; indeed, we know there are raw copies. But it is reasonable to believe that the percentage of extant copies of those two books that are slabbed is higher than any other comparable comic, so that's a great place to start. CGC's census lists 75 copies of AC1 and 74 copies of Tec27. Now, before we continue, it's important to remember that one of the important caveats of the CGC Census is that books can appear on it multiple times because of cracking and resubmission. In an ideal world, the cracked label would be returned to CGC so that the original number can be decertified, but that was never going to happen consistently. For these particular books, there's been an effort on the boards to identify re-slabbed copies by identifying marks and have the original numbers removed after the fact. But it's almost a certainty that there are some physical books -- especially in the mid-to-lower grades that attract less attention -- that have snuck into the census more than once. Regardless, that means the census sets an upper bound on the number of slabbed books; there might be fewer, but there cannot be more. If you want, you can reduce the number by, say, 10% to account for slab duplication. But this is back-of-the-envelope math, so it doesn't really matter very much. Technically, there are copies in the Other Company's slabs also -- 8 of them for AC1, for example. I'm not going to worry too much about those. Several of them are extremely heavily restored at this point, and may very well be cracked out of CGC slabs. If not, we can just pretend they're still raw. Again, it doesn't really change the outcome significantly. As a final notes, census counts do not include NG books, which for Golden Age especially means "coverless copies". Maybe you count a coverless book as a "surviving" copy; maybe you don't. I don't think there are actually that many coverless copies of most books, but I know there are some issues of some titles with warehouse weirdness. It really isn't likely to change the order-of-magnitude estimates, but it's worth keeping in mind if you're into such things.

All that still doesn't tell us how many raw copies there are, and there no way to know that from the census. But it's certainly a topic that's been talked about a lot in the hobby over the years. How many Tec27s are "still out there" -- which, for this discussion, means both known copies (like the Mile High/Dentist's copy of AC1) and unknown copies (that Tec27 you would have found if you'd just looked five minutes longer in that little antique shop!)? The longstanding estimation is that, for these Greatest of All Keys, the CGC slabs represent about 50% of the total world population. I could talk a lot about trend lines and diminishing rates of encapsulation and do some statistics here in an effort to convince you that's a little too high, and that the slabbed Tec27s are more like 70% of the population... but let's not. Let's go with the 50% standby. It's a good, round number. Besides, aiming high on these estimates is good for us.

So, all told, we can reasonably say that there probably aren't more than 150 copies each of AC1 and Tec27 in the world, regardless of slabbing status or condition.

Now, we know that those books were quickly recognized as special even at the time (and, especially with AC1, we know they printed an amazing number of copies for the era). So it's reasonable to think that print runs plus survival rates mean that AC1 and Tec27 were printed in higher numbers and were also more likely to survive that other books of the same era. And, indeed, if we look at the pre-Batman Tecs, we see that their census totals run around half -- or sometimes less -- than that of Tec27. There are only 16 copies of, let's go with Tec14, on the census. Tec27 gives us an upper bound of around 150 copies; if we apply the same math to Tec14, we get an estimate of, say, 40 copies. Although still a really pricey books, the pressure to slab is probably not quite as high for Tec14 -- a random pre-Bats issue with nothing really interesting going on and a serviceable but not distinctive cover -- than for Tec27. If you don't like the 50% slabbing rate we used earlier, feel free to lower it here too. But as a sort of first order approximation, we get that there are between 40 and 150 copies of Tec14 in the world; I'd favor a number near the bottom of that range, and say that, give or take, there are about 50 of them, most of which are currently unslabbed (and most of which are consequently probably fairly-to-very low grade). For what it's worth, if you do all this math with a 70% slabbing rate instead of a 50% slabbing rate, you end up with a reasonable guess of around 30-35 copies of Tec14 extant; as I said at the start, it doesn't matter much and these are rough estimates.

The general process holds true for quite awhile. We can use the next two big books, Batman 1 and Superman 1, to benchmark the rest of the '30s, as print runs and public awareness of the medium increased markedly. There are 289 Batman 1s and 168 Superman 1s on the census; it's likely that there are not more than, say, 500 copies of either book remaining in the world today. And I think those books set the high water mark for the '30s (yes, I know, Batman 1 is a 1940 book). Whether you're looking at relatively well-known books the Fox/Holyoke Blue Beetle 1 or obscure comic strip reprints like Tip Top Comics 44, it's safe to say that there are fewer than 500 extant copies, and for many (perhaps most) such books, fewer than the 150 threshold established by AC1 and Tec27. Especially for these earlier books, the original print runs are comparatively unimportant; there was no widespread collecting hobby, and so preservation has disproportionately favored happenstance or early adopters who retained large volumes of a very diverse array of books; Edgar Church is the broadest and best example, but there's a reason why the "pedigree collection" is a dominant feature of '30s and early '40s collecting. That's not to say that books which started out with fewer copies don't have fewer copies remaining; they typically do. But the relationship between original print run and surviving population is not necessarily as clearly defined as it would become; otherwise, there would be a LOT more copies of Action Comics 1 than we see.

Once we pass in the '40s and the rest of the Golden Age, it becomes a lot harder to track. Print runs often (but not always!) increased, and preservation often (but not always!) became more likely. The overall process still holds, though. Find the biggest key within a year or two of the book you're interested in. Multiply the census total by 2, assuming that slabbed books represent only half the population for even very expensive Golden Age books, and that's your rough upper bound. For a lot of minor titles with minimal collector interest and few slabbed copies, that's a huge overestimate; some of those GA books may only have a few dozen extant copies, if that. But that's the best we can do with napkin math.

By the time you're into the '50s and the Golden Age is giving way to Silver, you probably need to increase that multiplier. The further away from the lofty prices of AC1 / Tec27 / Batman 1 / Superman 1 you get, the less incentive there is to slab, especially low grade books. There are 551 copies of Showcase 4 on the census, and 1269 copies of Showcase 22. For the latter, especially, you can see the census count drop off steeply for books below 3.0; even though a beater copy of Showcase 22 is still a valuable book, there's less urgency to get literally every copy professionally graded. We can conclude safely that there are, proportionately, a lot more 0.5 or even 2.0 copies of Showcase 22 than the census suggests. So maybe the multiplier now is 3 -- that is, only 33% of extant copies of Showcase 4 or Showcase 22 are sitting in slabs. That suggests there aren't likely to be more than about 1500 copies of the first appearance of Barry Allen, or more than around 3000 copies of the first appearance of Hal Jordon, and those numbers feel reasonable enough. Looking at a comparable book, Action Comics 252's first appearance of Supergirl, also gives us the same 3000 copy estimate. Just like with AC1 and Tec27, it's good when more than one data point suggests about the same conclusion. Likewise, it's probably safe to assume that some random '50s book with little collector interest has no more surviving copies than the big issues do, and often substantially less (because, after all, most of them had lower print runs). Most issues of the DC's long-running 50s title Flippity and Flop have all of ZERO graded copies on the CGC census, because, well, who cares about an off-brand Sylvester and Tweety knockoff? But we can be virtually certain that those issues have fewer surviving copies than the 1500 we expect from Showcase 4 (probably a lot less, but we can't really guess how much so with such minimal data).

Much past the '50s, though, the answer tends to merely be "there are a lot of copies out there". Depending on your assumptions, it's possible to believe there are as many as 10,000 copies of Amazing Fantasy 15 out there (although its rocket ascent in value means I think it's more likely to be slabbed than you'd expect, and there are "only" like 7500 extant). Books from the 60s onward are generally only "rare" in a pure mathematical sense if they started that way due to restricted print runs or else had special circumstances that made preservation unusually unlikely. The 1977 Marvel 35 cent test variants took it on both ends of that calculus, for example; or the big indie keys like Albedo 2, or various undergrounds. But that requires specific research outside of the scope of the census; the census is still a nice tool there for other reasons, but this post is quite long enough already.

 

[KirbyTown]

On 8/4/2022 at 5:52 AM, Flammenwerfer said:

Is it possible to estimate based on the number of graded copies?

Maybe, but that seems less straightforward than using print run info?

[revat]

On 8/4/2022 at 1:45 PM, KirbyTown said:

On 8/4/2022 at 3:52 AM, Flammenwerfer said:

Is it possible to estimate based on the number of graded copies?

Maybe, but that seems less straightforward than using print run info?

i think discussed in the long post above, but for golden age it seems like survival rate due to its slabbability seems like a more useful starting point than just general survival rate as a percentage of print run.

for silver age, yah it would probably be a factor, possibly a bigger factor for the vast majority of most comics.

[shadroch]

On 8/4/2022 at 1:45 PM, KirbyTown said:

Maybe, but that seems less straightforward than using print run info?

How would knowing there were 3,000,000 copies of Action 1 printed help determine how many still exist?

[Lazyboy]

On 8/4/2022 at 3:28 PM, Qalyar said:

(and, especially with AC1, we know they printed an amazing number of copies for the era)

What?

[Qalyar]

On 8/4/2022 at 4:28 PM, Lazyboy said:

What?

It's been widely reported that Action Comics 1 had a print run of 200,000 copies. This wasn't actually the highest known print run in the era (one of the Famous Funnies may have had twice that), but it's a lot higher than known or estimated values fo many of the '30s books, especially from publishers that weren't DC.

[Lazyboy]

On 8/4/2022 at 4:36 PM, Qalyar said:

It's been widely reported that Action Comics 1 had a print run of 200,000 copies. This wasn't actually the highest known print run in the era (one of the Famous Funnies may have had twice that), but it's a lot higher than known or estimated values fo many of the '30s books, especially from publishers that weren't DC.

Yeah, 200k printed is not amazingly high for any era. That's why I was wondering what number you were thinking of.

[Hudson]

On 8/4/2022 at 4:28 PM, Qalyar said:

By the time you're into the '50s and the Golden Age is giving way to Silver, you probably need to increase that multiplier. The further away from the lofty prices of AC1 / Tec27 / Batman 1 / Superman 1 you get, the less incentive there is to slab, especially low grade books. There are 551 copies of Showcase 4 on the census, and 1269 copies of Showcase 22. For the latter, especially, you can see the census count drop off steeply for books below 3.0; even though a beater copy of Showcase 22 is still a valuable book, there's less urgency to get literally every copy professionally graded. We can conclude safely that there are, proportionately, a lot more 0.5 or even 2.0 copies of Showcase 22 than the census suggests. So maybe the multiplier now is 3 -- that is, only 33% of extant copies of Showcase 4 or Showcase 22 are sitting in slabs. That suggests there aren't likely to be more than about 1500 copies of the first appearance of Barry Allen, or more than around 3000 copies of the first appearance of Hal Jordon, and those numbers feel reasonable enough. Looking at a comparable book, Action Comics 252's first appearance of Supergirl, also gives us the same 3000 copy estimate.

 

I appreciate your analysis of this and for taking the time to write out the logic behind your conclusions.

I was especially interested in your thoughts on Actions from the late 50's.  I see you estimate Action 252 at 3,000.  I think it is interesting that Action 242 has only 565 graded copies as opposed to 1,131 for Action 252.  Given these books are both "keys", simialar in value,  and were printed less than a year apart, I would think the print runs were approximately the same and the survival rate about the same.  Would you put Action 242 also at 3,000?  I am curious if you believe the fact that Supergirl being introduced and put on the cover may have changed the survival rate.  Did more females buy this book or more males keep it?  IDK.  For a long time, I have always thought that Actions from the late 50's had an upward survival rate of about 2,000, however, I appreciate your reasoning and obvioulsy I could be mistaken.  Thank you again, for your thoughts and step by step logic.  Well done.

[KirbyTown]

On 8/4/2022 at 3:54 PM, shadroch said:

How would knowing there were 3,000,000 copies of Action 1 printed help determine how many still exist?

It seemed reasonable to adjust the info we do have, like postals, into assumptions about issues that don't have such info. It was a layman's observation, I'm definitely not a data scientist 🤷‍♂️

On 8/4/2022 at 3:52 PM, revat said:

i think discussed in the long post above, but for golden age it seems like survival rate due to its slabbability seems like a more useful starting point than just general survival rate as a percentage of print run.

for silver age, yah it would probably be a factor, possibly a bigger factor for the vast majority of most comics.

Sound good to me 🥳

Edited August 4, 2022 by KirbyTown

[Qalyar]

On 8/4/2022 at 5:07 PM, Hudson said:

I appreciate your analysis of this and for taking the time to write out the logic behind your conclusions.

I was especially interested in your thoughts on Actions from the late 50's.  I see you estimate Action 252 at 3,000.  I think it is interesting that Action 242 has only 565 graded copies as opposed to 1,131 for Action 252.  Given these books are both "keys", simialar in value,  and were printed less than a year apart, I would think the print runs were approximately the same and the survival rate about the same.  Would you put Action 242 also at 3,000?  I am curious if you believe the fact that Supergirl being introduced and put on the cover may have changed the survival rate.  Did more females buy this book or more males keep it?  IDK.  For a long time, I have always thought that Actions from the late 50's had an upward survival rate of about 2,000, however, I appreciate your reasoning and obvioulsy I could be mistaken.  Thank you again, for your thoughts and step by step logic.  Well done.

Probably not possible to distinguish based solely on census data.

Based on upper bound estimation, I would confidently say that there probably are fewer than, say, 3000 copies of either book extant. Do the census numbers imply that there are significantly fewer copies of 242 vis-a-vis 252? Wellll.... The two controlling variables, neither of which we can actually know, are the actual population (which is what we want to estimate) and the slabbing rate (which we can only make semi-educated guesses about, at best):

• Market snapshots of the day aside, I'd argue that "first appearance of Supergirl" is a more distinctive book than "first appearance of Brainiac". Brianiac is a good villain, but there are lots of good villains, and he's certainly not the sort of character that has a lot of currency outside of readers of the Superpeople titles. Supergirl, on the other hand, got herself an (admittedly terrible) 1984 film plus six-ish years of an eponymously-titled CW TV series in the Arrowverse. That suggests that slabbing rates may be higher for 252. All else being equal, this might mean the slabbing multiplier for 252 would be lower than for 242.
• But we also need to consider that, by 1958-59, we're well into the era where intentional preservation via collecting is an important factor, which means print runs matter more. I don't have any data about the print runs of 1950s Action Comics issues, and its fairly likely that no such data exists in public hands. It's fairly likely that there were simply more copies of 252 to begin with, and -- all else being equal -- later books are more likely to be preserved in numbers than earlier ones.

On the balance, I think it's likely there really are fewer copies of 242 extant than 252. The census shows disparity as essentially every grade level, including the top of the chart where financial incentive leads to a higher percentage of copies getting encapsulated. I don't think we can confidently say how many fewer, especially with factors pushing our napkin math in opposite directions. If I was told to make more precise estimates with the understanding that they may not be more accurate guesses, I might think that 242 is a book that fits the normal model for its era, and so has, let's say 1500 extant copies, give or take a couple hundred, but that 252 probably has a slightly tighter multiplier, representing 2500 or so copies. But there's a lot of uncertainty there, and the error bars are nothing to sneeze at. If the Information Fairy descended from On High and revealed that there are actually exactly 2197 copies of each book in existence, I wouldn't feel that represents a failure of the overall model.

Edited August 4, 2022 by Qalyar

[Lazyboy]

On 8/4/2022 at 5:35 PM, Qalyar said:

I don't have any data about the print runs of 1950s Action Comics issues, and its fairly likely that no such data exists in public hands

Yeah, the first available numbers are a little bit later.

[Hudson]

On 8/4/2022 at 6:35 PM, Qalyar said:

Probably not possible to distinguish based solely on census data.

Based on upper bound estimation, I would confidently say that there probably are fewer than, say, 3000 copies of either book extant. Do the census numbers imply that there are significantly fewer copies of 242 vis-a-vis 252? Wellll.... The two controlling variables, neither of which we can actually know, are the actual population (which is what we want to estimate) and the slabbing rate (which we can only make semi-educated guesses about, at best):

I think one other thing we can look at is how many copies are for sale.

I just did a quick search on eBay and found the following.

Action 242

3 raw copies (2 with covers and 1 without)

16 slabbed (15 CGC and 1 other)

19 Total Books

Action 252

3 raw copies

30 slabbed (24 CGC and 6 other)

This leads me to believe that you may be correct in that there are more copies of 252 in existence than copies of 242.

After all there are almost 58% more 252s for sale than 242s (at least on eBay - which I think is a good indicator).

I wonder, however, what this does to the overall model.  In the case of 242, we can say that 16 out of 19 books were slabbed or more than 84%.

In the case of 252 the number is even higher with 30 out 33 being slabbed or almost 91%.

I agree that people selling books have more of an incentive to grade them than those not selling them, so I don't think 91% of ALL 252s are slabbed, however, I question if when looking at a key book like that the multiple should really be at 3 (suggesting only 1/3 or 33.3% of the total copies have been slabbed).  Just some "food for thought" or perhaps another piece of data that could help us narrow down what the actual numbers are.

[StillOnly25Cents]

tree fiddy