Two C.J. Stroud rookie parallels were on my watchlist this week, both PSA 10, both from 4 Sharp Corners, both cards I wanted.
The 2023 Panini Prizm Purple #339, numbered to 125, was at $522.00 with 12 bids and a day and a half left. The Green Scope #339, numbered to 75, was at $700.00 with 18 bids.
I did not bid on either one. The reason was the PSA pop report line sitting a few lines below the price.
This is the PSA pop report vs print run problem in its cleanest form. Total PSA pop on the Purple: 118, against a print run of 125. Total PSA pop on the Green Scope: 68, against a print run of 75. Both figures as of August 24, 2026.
Read that fast, and it says 94% of every Purple ever printed has already been slabbed by PSA, and 91% of the Green Scopes. My first reaction was that the number had to be broken. It is not broken. I was reading it as a headcount. A pop report has never been one.
The short version
-
A PSA pop report counts certification records, not physical cards. One card graded three times is three records.
-
Graded totals can exceed a print run. The hobby blog SlabStox found a Mike Trout Blue Refractor auto numbered to 150 showing 163 total graded copies back in 2019, with no breakdown by grader.
-
The number also runs low. Raw copies, copies still in sealed product, and copies graded only by BGS, SGC, or CGC are invisible to it.
-
The pop figure inside an eBay listing comes from PSA's own database, matched on the certification number and refreshed daily.
-
Gem rate and population growth are the reads worth having on a modern numbered card, and both carry their own caveats.
What does a PSA pop report actually count?
A PSA population report counts grading records. Every time PSA issues a grade, a new certification number enters the database and the count on that card line goes up by one.
Nothing in that process publicly confirms that the physical card behind cert number two is different cardboard than the card behind cert number one. On a base card with a pop of 40,000, that distinction does not matter. On a card numbered to 125, it decides whether you bid.
The scale is worth holding onto. GemRate's year-end data showed 26.8 million grading records across the major authenticators in 2025, with PSA handling 19.26 million of them at a 72% market share. Worth noting for a football card discussion: sports cards were only 10.0 million of that total and were actually down 12% year over year, with the growth coming from TCG, which was up 95%. Even at 10 million a year, that is a firehose of records moving through a system that has no concept of an individual copy.
Can a PSA pop report be higher than the print run?
Yes. Pop reports count grading records rather than copies, so one card that gets cracked out and regraded can occupy two lines on PSA's own count. A hobby blog documented combined graded totals running past the print run back in 2019. A PSA pop report vs print run gap is not automatically a reporting error.
The hobby blog SlabStox ran the exercise in March 2019 across Bowman Chrome autographs. In 2016 Bowman Chrome, a Juan Soto Orange Refractor auto numbered to 25 showed 36 total graded copies across BGS and PSA, 32 of them at BGS and 4 at PSA. A Vladimir Guerrero Jr. Orange Refractor auto, also numbered to 25, showed 37, split 31 BGS and 6 PSA. A Soto Red Shimmer numbered to 10 showed 15. Outside that set, SlabStox also checked a Mike Trout Blue Refractor auto numbered to 150, which showed 163 total graded copies, with no grader split given. Those counts are seven years old, and by the logic in the next section they have almost certainly gone up since.
Two honest caveats. SlabStox was hunting for cards that broke their print runs, and even inside that hunt it found one that did not. A 2016 Bowman Chrome Vladimir Guerrero Jr. Blue Refractor auto numbered to 150 sat at 149 total graded, just under the run. The bigger caveat is that every one of those figures is a combined total. On the two cards where SlabStox published the split, PSA's own count was 4 and 6 against a print run of 25, nowhere near the run on its own. Combined totals also carry the double-count problem covered in the crossover section below. That makes this evidence the phenomenon exists at the hobby-wide level, not a documented case of one grader's report passing the run on its own. Exceeding the print run is possible with nothing actually broken, and it is not the default.
A 2023 Prizm parallel of the 2023 Offensive Rookie of the Year is exactly the card people grade the day they pull it. A pop of 118 could be 118 distinct serial numbers. It could be 95 distinct cards with 23 of them going through the machine twice. Nothing in the pop report will tell you which. That is the whole problem.
Why do pop reports run higher than the number of cards?
One mechanic adds records to PSA's own count without adding cards, a second is why those extra records never leave, and a third inflates the hobby-wide total: crack and resubmit, the old record that never gets pulled, and crossovers from other graders.
Crack and resubmit is the familiar one. Somebody buys a PSA 9, breaks it out of the holder, and sends the raw card back in hoping for a 10. If it hits, PSA issues a new certification number and the pop report carries two records for one card. On a card where the 10 is worth multiples of the 9, that is a rational bet plenty of people take repeatedly.
The old record is the quieter problem. There is no published PSA policy on removing a record when a card gets cracked out. In a 2013 Collectors Universe thread the answers conflict, with some collectors saying PSA pulled the record after they mailed the certification label back and others saying it never happened. Whatever the practice, almost nobody mails the label back. It costs postage and effort, and the payoff cuts both ways. The person holding the old 9 label is the same person who just got the 10 back. Pulling that 9 shaves one off the total, which is the number most buyers glance at first, and it does not change the pop of 10s at all. It only shrinks the denominator, which pushes the gem rate up and makes the 10 look more common. A small win on the number people misread, a small loss on the one that actually describes the card.
Crossovers work differently, and the distinction matters for the 118. Cross a BGS 9.5 into a PSA 10, and PSA's count goes up by one for a copy it genuinely had not graded before, so PSA's own number is not inflated by it, unless that card carried a PSA record earlier in its life, which happens. Beckett's count does not come down on its own, though, which means that card lives in two companies' censuses permanently and any combined multi-grader total counts it twice. The hobby blog Cardhound Vintage has written that PSA has card fingerprinting technology capable of catching a PSA-to-PSA resubmission. I have not found public confirmation from PSA that anything like that runs in the grading line, so treat it as an unconfirmed claim from one source. Even if it does run, it would do nothing about a card arriving from a different company.
Does reholdering a card inflate the pop report?
No. A PSA reholder keeps the original certification number, so the population count does not change.
The service swaps the case and updates the label style while the grade and the cert number stay put. Same record, same count. This is the part of the system that works exactly the way you would assume it does.
The pattern underneath it is narrower than it first looks. A reholder stays honest because no new record is created at all. The cert number rides along with the card into the new case, so PSA never has to decide whether the thing in front of it is something it has seen before. Every other path back into the system makes PSA judge whether it has seen the card before, and short of the unconfirmed fingerprinting claim above, it has no reliable way to. A raw card off somebody's kitchen table might be one it has already counted. A card arriving in a Beckett holder is usually one it has not. Either way a new record goes in, and in practice records almost never come back out.
Why do PSA pop reports also run lower than reality?
Because a pop report can only see cards PSA has graded. Raw copies, copies still sitting in sealed 2023 Prizm product, and copies that only ever went to BGS, SGC, or CGC are all invisible to it.
Those other graders are not a rounding error. CGC alone graded 4.92 million cards in 2025 across all categories, and BGS and SGC sit on top of that. How much of any of it is football is not something the public totals break out. A copy can also get logged under a slightly different card line, which is a live risk in a Prizm set carrying a dozen near-identically named parallels.
The pop number is therefore a log of PSA's grading activity on that card line. It is close to a complete log of that activity, allowing for the misfiled-line risk above, and a badly incomplete count of how many copies exist in the hobby. Treat it as the number of times PSA has logged work on that line, and you will be close to right. Treat it as an inventory, and you will be wrong every time.
Where does the pop number on an eBay listing come from?
It comes from PSA. eBay matches the listing's certification number against PSA's database and displays the population spread across grades inside the listing, refreshed daily.
That removes the easy explanation. The 118 is not eBay lumping parallels together or running a stale scrape. It is PSA's own figure for that exact card line, updated within the last day. Anyone who wants to argue with it is arguing with what a pop report is.
Why doesn't PSA record the serial number?
Because the population report is organized by card line and grade, a structure built to count how many copies of a card sit in each grade, never to track which copies they are. PSA photographs both sides of every card it grades today, serial number included, and publishes the image, but it never exposes that number as data anywhere in the pop report or the cert lookup.
Since January 2023, PSA has scanned the front and back of every card it grades under a program called SecureScan, and those images are posted publicly on the certification verification page. The product did not exist before 2023, so every one of those 118 Stroud Purple certs falls inside the program and has a PSA-made image of the card sitting on PSA's own website. The serial number is printed on the card. It is right there in the scan.
PSA has the image, and the image is public. The serial number inside it has never surfaced as a searchable field, and nothing in the public setup connects a certification record to a specific copy as data, even while PSA is looking directly at it.
I am not going to guess at motive or pretend I know PSA's internal roadmap. I do think this is the most fixable problem in the whole mess. Log the serial number on numbered cards, expose it in the pop report, and most of the guesswork on the high side of this, the crack-and-resubmit records and the stale ones, goes away. The low side, the raw and sealed and rival-slabbed copies, stays exactly where it is.
There is a manual workaround if you care enough. Cert scans are public, but the pop report will not hand you the 118 cert numbers, so you work backward from listings. Sold listings show the slab, and on a card with a pop under 150, a patient person can start collecting distinct serial numbers and build a partial census by hand. It is tedious, and it will be incomplete. It is also more than the pop report is going to give you.
What should you actually read on a numbered card?
Read the gem rate and the population growth, and stop reading the total against the print run.
Gem rate is the share of graded records that came back a 10. On the Stroud Purple, that is 64 of 118, or 54.2%, as of August 24, 2026. On the Green Scope, it is 35 of 68, or 51.5%.
Gem rate is not clean either, and it is worth understanding why. Gem rate is built out of grading events. Copies never enter the math. When a 9 gets cracked and comes back a 10, the old 9 record almost always stays in the denominator, so the reported rate understates the share of PSA-graded copies actually sitting in 10 holders. If PSA cleared those stale records off the Stroud Purple, 64 of 118 would become 64 of a smaller number and the percentage would climb. Read it as the odds a submission comes back a 10, and as a conservative read on what share of the PSA-graded copies are in 10 holders today. It says nothing about the raw, sealed, and rival-slabbed copies from the section above, which sit outside the math entirely. It is still the better read, because it describes the grade distribution rather than posing as an inventory.
Context matters more than the raw percentage. GemRate data reported by cllct put the sports card gem rate at 34% through the first half of 2025, down from 40% across 2024, with the all-category figure including TCG sitting at 43%. That 34% covers every sports card graded, vintage included, and my read is that modern chromium parallels run well above it. A 54% on a 2023 Prizm is not the outlier it looks like next to the industry average. What it does tell you is that this is not a condition-sensitive card, and you should not be paying a condition-scarcity premium for one.
The second read is velocity. A pop of 118 that has not moved in eight months means something entirely different from a pop of 118 that was 90 in the spring. GemRate tracks population growth over time and pulls PSA, BGS, SGC, and CGC into one view, which is the widest angle available, with the caveat from the crossover section above: a combined total counts every crossed card twice.
How I handle pop before I bid
I run comps on anything I am serious about, and pop is one input into that.
The order is boring on purpose. Sold comps come first, because what people actually paid beats every theory about scarcity, and I comp against the all-in number rather than the hammer price, which is its own trap on some platforms. Gem rate comes before total pop, since that is the number telling me what I am really buying. Then I check whether the population has been growing, because a modern card can quietly reprice itself as its pop climbs. Then I treat the total as a record of grading activity and refuse to turn it into a tally of surviving copies, no matter how much the layout of the listing wants me to.
This is the same trap we hit at BattlBox with paid acquisition. Our best-performing ads, the ones with the lowest CAC, also pull in the most bad actors, and that wrecks the cohort behind them. We have killed wildly successful low-CAC campaigns because the LTV underneath was so much worse. The CAC number was accurate every single day it sat on the dashboard. What we assumed it meant was wrong.
Frequently Asked Questions
Can a PSA pop report be higher than the card's print run?
Yes. Pop reports count certification records, not physical cards, so a card that was cracked out and regraded appears more than once. The hobby blog SlabStox documented a Mike Trout Blue Refractor auto numbered to 150 showing 163 total graded copies in 2019, though that figure combines every grader rather than PSA alone.
Does PSA remove a card from the pop report when it is cracked out?
There is no published PSA policy on this. In a 2013 Collectors Universe thread, collectors reported mixed results, with several saying PSA pulled the record after they mailed the certification label back and others saying it never happened. In practice, almost nobody mails the label back.
Does a PSA reholder create a duplicate in the pop report?
No. A reholder keeps the original certification number and only replaces the case and label, so the population count does not change.
Where does the PSA pop number on an eBay listing come from?
From PSA. eBay matches the listing's certification number against PSA's database and displays the population spread inside the listing, with the data refreshed daily.
Does PSA record the serial number on numbered cards?
Not as data. PSA has scanned the front and back of every card graded since January 2023 under SecureScan and publishes those images on its public certification verification page, so the serial number is visible in the scan. It is not captured as searchable data, and it is not connected to the population report.
Is gem rate a more reliable read than total pop?
It is better, though not clean. Gem rate measures grading events rather than copies, and a card's old lower-grade record usually stays in the denominator after a crack and resubmit, so the figure understates what share of the PSA-graded copies are actually sitting in 10 holders. It still describes the grade distribution, which is more useful than a total that reads like a headcount.
Final Thoughts
I still want both Stroud cards. What changed is what I think the pop number was telling me.
It was telling me how many times PSA has graded that card line, which is genuinely useful information for a different question than the one I was asking. Once you stop asking a receipt log to behave like a census, it becomes a good tool again.
The part that sticks with me is how fixable this is. PSA is photographing the serial number of every numbered card that moves through the building and publishing the image, and the population report still cannot tell you whether it has seen that copy before. The hobby has spent years pricing modern cards on scarcity, using a number that was built to log grading work.
Until that changes, do what you would do with any number somebody hands you. Ask what it counts before you decide what it means.
