T206 Sweet Caporal Census Project

[53toppscollector]

Hello. Apologies for cluttering up the main board, but I wanted to start a new thread for this project, as I suspect it will require a lot of posts, and I want to keep it organized and in its own place, instead of lumping it into my other thread of stats and curiosities.

Judging from the thread title, you probably already know what this is going to be. I have only been on this board for a little over a year, so if something like this has been done before, you can blast this thread into the sun and I won't be offended.

I'm a curious person, and one thing that has always bothered me with the population reports for PSA and SGC is how poorly they handle the Sweet Cap factory numbers. If you are the type to care about such things, then you know that in the beginning of (graded card time), PSA did not adequately label their slabs with the requisite information. They only added "T206" and then eventually "T206 Sweet Caporal", before finally reaching an enlightened state and adding "T206 Sweet Caporal 350/25" to the labels. The result is that there are a lot of SweetCap PSA slabs out there with no identifying info on the flip to id which factory the card was from, and as a result, the population reports are not very accurate. SGC is actually worse, in that they do not catalog any of their SweetCap factory numbers. So, we have zero actual info from them. Not great!

Being the sicko that I am, I've decided I want to try and rectify this to a degree, so this project is born. I actually just did one card, it took some time, but it was actually kind of thrilling (again, I am a sicko) clicking through old scans and uncovering treasure (cardboard factory numbers). I am going to use this thread to catalog my research, devoting a new post to a new card and then cataloging the results in this main post.

Before I get into the data, I want to add some reminders and notes here so I don't have to repeat them every time I make an update. I'm big on explaining methodologies so others can crosscheck if they would like to.

Methodology

* VCP has sales data for T206s dating back to 2006, maybe even older. For each sale, they also include scans of the card. The quality of the scans varies. Some are crystal clear, others are blurry and hard to zoom in on. On the whole, I was pleasantly surprised how useful the images were.

* Older listings typically only have the front of the card image available. **Price Is Right Loser Horn Sound** For these, I did not include them in my census.

* VCP lumps in all of the SweetCap and Piedmont Sales together. I am also checking the Piedmont 350/460s (quickly) to see if they missed any Factory 42s on the labels

* If I cannot read the scan and determine the factory number, I am going to make a notation of ??? in the data. I don't want to just guess, that is the whole point of doing this census, to try and take away some of the guesswork. I will identify these cards separately.

* I am tracking the cert numbers from PSA and SGC, which helps identify cards sold more than once. In my tracking of the first card, I actually had the first instance where the scan was too blurry, but a second sale had a better quality scan, and I was able to accurately label it.

* When I show the charts with the population reports from SGC and PSA, I will include the total number graded by SGC, but obviously there is no factory number breakdown because SGC does not put that info on their flip. They group the Sweet Caps as 1909 Sweet Caporal (SC 150), 1910 Sweet Caporal (SC 350) and 1911 Sweet Caporal (SC 460) so we will only have their aggregate totals. I will include them just for reference.

* I'm going to go alphabetical through the set. I debated going by print groups, but, I'm just going to go alphabetical. Each player will get his own post with the stats.

* I reserve the right to give up on this fairly ridiculous journey at any point. Though I am genuinely curious about this, so I want to see it through. The pace of my updates will depend on other stuff that comes up.

Caveats

* This is only a look at graded cards, and only those cards in the VCP database. VCP collects data from ebay and other auction houses, so it is a good source, maybe the best source we have, especially since ebay sold listings vanish after 3 months. I have scans here on sales from like 2014. We all know that for a lot of these cards, there are a ton of raw copies floating around. This is meant to give us a decent sample that we can analyze and discuss

* Some of these cards may have been crossed over or reholdered. I can't do anything about that.

* Some cards get graded and never sold, so my census will not match up 1:1 with the population reports. Obviously.

Here is a quick visual of why the PSA pop report is tough, if you haven't ever looked at it:



This will hopefully make more sense when I show the first card in the set and how I designed my chart.

Error Log:

Atz: SC 350/25 slab has an SC 350/30 back
Baker: An SC 460/42 slab has an SC 460/42OP back
Baker: An SC 350/30 slab has an SC 350/25 back
Baker: An SC 350/25 slab has an SC 460/25 back (PSA DNA card)

[53toppscollector]

1. Abbaticchio (Blue Sleeves) [Print Group 4]



I tinkered with the layout a bit. My aim was to make it easily readable.

The first section of the graphic shows the info from the PSA and SGC pop reports online. I included the SGC row only to show the total number they have graded, since we cannot break down their numbers by factory in their pop reports. For PSA, I included the total number of Sweet Caps graded in column B (155, in this case) and then on the far right column, I included the number graded broken down by factory number in the actual pop report (80, in this case). That essentially means there are 75 copies of this card graded that don't contain the factory number on the flip, or that were not properly added to the PSA pop report. The PSA Pop Report % is the percentage for each factory based on the number that were graded and include the factory number. So for that row for this example, the denominator is 80, as that is the number that were graded and labeled with factory numbers, not 155, which is the total number of SweetCaps that PSA has graded. The 155 column is included just to show the gap between the total graded and the total that were labeled with factory numbers.

The second section of the graphic shows my work cataloging the cards using the VCP website and their scanned images of cards based on sales. The first row is the raw number I found, and the row under it shows the percentage by factory of those that I was able to catalog. It is broken down by PSA, then SGC, and then total. "Not Cataloged" are the scans where I could not definitely tell what the factory number was.

The third section of the graphic shows the "quality control" stats for PSA cards only. I wanted to keep track of how many cards were labeled incompletely or incorrectly. Correct Label with Factory is self-explanatory. The label says SC 460/25 and the back is an SC 460/25. Incorrect PSA label, also self explanatory...the label says SC 460/25 but the back is an SC 460/42OP. T206 Sweet Caporal is a previous generation label where the factory number was not listed, so it was a mystery until I clicked the back. T206 was the original label, without the SweetCap designation at all. A few of those will end up being Piedmonts as well.

I added the date on the right side so I can remember when I collected the data, which will make it easy for me to go back and update, if need be.

And we are off and under way

[53toppscollector]

2. Abbaticchio (Brown Sleeves) [Print Group 1]



Note that on this chart for the SGC population at the top, I merged together the Fac 25/30 cells because SGC does not use factory numbers. So there are 56 cards labeled as 1909 Sweet Caporal, which means they are either 25 or 30. Same applies to the 350 section, and then those 2 numbers give you the 81 total graded copies.

[53toppscollector]

3. Abbott (Print Group 2)



After going through 3 cards, I have been able to make the following ids:

Generic PSA T206 Labels > Factory numbers: 14 (with a whole bunch of Piedmonts mixed in)
Generic PSA Sweet Caporal Labels > Factory numbers: 60


When people get back into the hobby, a card collecting council should send them a pamphlet explaining the importance of flatbed scanners with CCD elements. So many blurry scans. Ugh.

[RCFire82]

James,
Quick rule of thumb I use when looking at psa's pop. report on Sweet Caporals is to use ratios. For example on the Abbaticchio (blue sleeves) above:
Let's say total Sweet Caporals graded is 160 (which is close) and 80 (or exactly half) have been identified with Factory numbers. I would say it's safe to say each Factory graded total, 460 25 (10 total), 460 30 (58 total), and 460 op (12 total) can be multiplied by 2. So one would expect to see an additional 10 460 fac 25's, 58 more 460 fac 30's, and 24 total 460 over prints when examining the 80 earlier non Factory labeled psa graded cards.

[Lobo Aullando]

That's my back-of-the-envelope rule of thumb in a nutshell. (Who doesn't love idioms?)

Also, one doesn't have to do a full census to get a good picture of the whole. A sample of 30 or so would get a pretty good estimate of what those ratios should be. (Picking every 17th card might be better than doing all the A name cards because it's less likely to encounter clustering.) It's even possible to get a good error range with that few.

Originally Posted by RCFire82

James,
Quick rule of thumb I use when looking at psa's pop. report on Sweet Caporals is to use ratios. For example on the Abbaticchio (blue sleeves) above:
Let's say total Sweet Caporals graded is 160 (which is close) and 80 (or exactly half) have been identified with Factory numbers. I would say it's safe to say each Factory graded total, 460 25 (10 total), 460 30 (58 total), and 460 op (12 total) can be multiplied by 2. So one would expect to see an additional 10 460 fac 25's, 58 more 460 fac 30's, and 24 total 460 over prints when examining the 80 earlier non Factory labeled psa graded cards.

Edited to add:


Just following the rules, even though I love the pretty tables.

[53toppscollector]

Originally Posted by RCFire82

James,
Quick rule of thumb I use when looking at psa's pop. report on Sweet Caporals is to use ratios. For example on the Abbaticchio (blue sleeves) above:
Let's say total Sweet Caporals graded is 160 (which is close) and 80 (or exactly half) have been identified with Factory numbers. I would say it's safe to say each Factory graded total, 460 25 (10 total), 460 30 (58 total), and 460 op (12 total) can be multiplied by 2. So one would expect to see an additional 10 460 fac 25's, 58 more 460 fac 30's, and 24 total 460 over prints when examining the 80 earlier non Factory labeled psa graded cards.

Yeah, this is kind of how I think about it in my head, but I think the reason I wanted to do this project was just to see if my assumptions generally track across the board, or if there are weird outliers that may not really be known just be using the ratios/estimates. Plus, looking through old scans is actually kind of fun, to be honest.

Originally Posted by Lobo Aullando

Also, one doesn't have to do a full census to get a good picture of the whole. A sample of 30 or so would get a pretty good estimate of what those ratios should be. (Picking every 17th card might be better than doing all the A name cards because it's less likely to encounter clustering.) It's even possible to get a good error range with that few.

Sure, that is fine for estimates. And ballpark estimates are fine for most. I just really want to dig in here and see what kind of things I can uncover by methodically going through everything available. I do understand it is a bit insane