What 32,000+ Graded Cards Reveal About the Card Market
We analyzed 32,000+ cards through AI grading to uncover the real card grading statistics for 2026. The data challenges common assumptions about gem rates, grading ROI, and which cards are actually worth submitting.
Jamie Budesky·Founder, CardGrade·Published Mar 17, 2026 · Updated Jun 5, 2026·8 min read
In short: CardGrade's analysis of 32,000+ cards found that most cards fall below a PSA 8 threshold, centering is the single biggest grade-killer, and the widely-cited 43% gem rate is distorted by selection bias. See why AI pre-screening pays for itself before your next submission.
We Analyzed 32,000+ Cards. Here Is What We Found.
The card grading industry processed an estimated 12.5 million cards in 2025, generating somewhere between $187 million and $312 million in revenue at $15 to $25 per card. That is a lot of money changing hands -- and a lot of collectors making individual decisions about whether their cards are worth grading.
But those decisions are largely based on gut feeling, forum advice, and a handful of widely cited statistics that turn out to be misleading.
For the State of Card Grading 2026 study, we analyzed over 32,000 cards through CardGrade's AI grading system. Unlike industry datasets that only capture cards people chose to submit, our data includes everything -- the cards that would have graded beautifully and the cards that never should have left the penny sleeve. The results paint a very different picture than the one most collectors have in their heads.
The Harsh Truth: Most Cards Are Not Worth Grading
This is the finding that will sting, but it is the one that can save you the most money. The majority of cards scanned through our system fall below the PSA 8 threshold -- the grade at which most modern cards start to carry a meaningful premium over raw value.
That means most cards, if submitted to a grading company, would come back in a slab that is worth less than the cost of grading. At $25 per card through PSA's Collectors Club tier, every card that grades below an 8 represents money lost. Not just "less profit than expected" -- actual negative ROI.
The math is straightforward. If you submit 100 cards at $25 each, that is $2,500 in grading fees alone before shipping, insurance, and supplies. If 60 of those cards come back at PSA 7 or lower, you just spent $1,500 on slabs that added no value. That is money you could have spent on better cards, or simply kept in your pocket.
Centering: The Number One Grade Killer
Across all 32,000+ cards in our dataset, centering emerged as the single most common reason cards miss top grades. Not corner dings. Not surface scratches. Centering.
This makes sense when you think about it. Corners, edges, and surface condition are partially within your control -- you can handle cards carefully, store them properly, and inspect them before purchase. But centering is baked in at the factory. A card that came off the press at 65/35 left-to-right is never getting a PSA 10, no matter how pristine the rest of it is.
The centering bottleneck is especially frustrating because it is invisible to casual inspection. Most collectors can spot a dinged corner, but few can eyeball the difference between 55/45 centering (passable) and 62/38 centering (grade-limiting) without a tool. That is why we built the free centering tool -- it takes the guesswork out of the most impactful factor in your card's grade.
The Company Spread: Same Card, Different Grade
Our analysis confirmed what experienced collectors have long suspected: the same physical card can receive meaningfully different grades depending on which company evaluates it. PSA, BGS, and CGC each apply their own standards, tolerances, and weighting to the four subgrade categories (centering, corners, edges, surface).
This is not a criticism of any company -- it is a reality of the grading landscape. PSA holds roughly 70% market share, with SGC at around 15%, BGS at 10%, and CGC at 5%. Each company has earned its market position through a specific approach to grading. But for collectors, the practical implication is clear: choosing the right grading company for a specific card is a strategic decision, not a default one.
A card with slightly off centering but pristine everything else might fare better at one company than another. Understanding those differences before you submit can be the difference between a 9 and a 10.
The Selection Bias Finding: 43% Gem Rate Is Misleading
Here is the statistic that gets thrown around constantly: roughly 43% of cards submitted to PSA receive a Gem Mint 10 grade, according to data from GemRate and CLLCT. On the surface, that sounds encouraging. Almost half of all submissions get a 10? Your cards must have decent odds, right?
Not so fast. That 43% figure suffers from massive selection bias. The cards being submitted to PSA are not a random sample of all cards in existence. They are the best of the best -- cards that collectors have already hand-picked, inspected, and pre-selected as their strongest candidates.
Our unfiltered dataset tells a different story. When you look at all cards -- not just the ones someone already decided were worth grading -- the gem rate drops dramatically. The 43% figure only holds for the highly curated subset of cards that makes it through collectors' personal screening process and into a submission envelope.
This is one of the most important card grading statistics for 2026: your cards do not have a 43% chance of getting a 10 unless you are pre-screening effectively. For a deeper dive into what this means for your submissions, read our full analysis of PSA gem rate selection bias.
Category Performance Varies Widely
Not all card categories perform the same when it comes to grading outcomes. Our data showed meaningful differences across sports cards, Pokemon, and other trading card game categories.
Modern sports cards from major manufacturers tend to have relatively consistent print quality, but centering issues remain common. Pokemon cards, particularly modern sets, show higher variance in surface quality. Vintage cards are a category unto themselves, where the grading calculus shifts entirely because even lower grades carry significant premiums.
For collectors who focus on a specific category, understanding your category's typical bottlenecks helps you target your inspection efforts. A Pokemon collector should pay extra attention to surface quality. A sports card collector should measure centering on every card. For category-specific grading data, see our grading results by category breakdown.
The Pre-Screening ROI Is Overwhelming
If there is one takeaway from the entire study, it is this: pre-screening cards before submission is the single highest-ROI activity in the card grading hobby.
Here is the math. Say you have 50 cards you are considering submitting at $25 each. Without pre-screening, you are spending $1,250 and hoping for the best. If our data is representative, a significant portion of those cards will grade below the threshold where the slab adds value.
Now say you pre-screen those 50 cards and identify the 20 that are genuine high-grade candidates. You spend $500 on grading fees instead of $1,250. Your hit rate on PSA 9s and 10s goes up dramatically because you filtered out the dogs. The 30 cards you held back cost you nothing, and you avoided $750 in wasted grading fees.
Over a year of submissions, that adds up to thousands of dollars saved. And that is a conservative estimate.
What This Means for You
The State of Card Grading 2026 study is not just an academic exercise. It is a practical tool for making better grading decisions. Here is how to use the findings:
Stop submitting blind. The data is clear -- unscreened submissions have a much lower hit rate than most collectors assume. Pre-screen everything.
Check centering first. It is the number one bottleneck and the easiest to evaluate with the right tools.
Do not assume 43% gem rate odds. That number applies to pre-selected cards, not your raw stack. Read the selection bias analysis to understand why.
Choose your grading company strategically. Different companies handle different card characteristics differently. Match the card to the company.
Run the ROI math before every submission. Factor in grading cost, expected grade, and the graded vs. raw price difference.
Jamie Budesky is the founder of CardGrade and the engineer behind its AI vision grading pipeline. An Army veteran and IT specialist (DoD, since 2017), he writes about card grading, AI/ML grading technology, and collecting strategy — grounded in CardGrade's own grading data.
