CGC 10s with more whitening than PSA 9s: Real difference or are CGC scans harsher?

I’m considering CGC 9–10s as alternatives to PSA 9s, but some CGC certification scans make me hesitate. Even on CGC 10s, I’m seeing corner/edge whitening and white spots that look worse than on an average PSA 9 I own (and I own quite a lot of PSA 9s).

I have a handful of CGC 9s, and some look relatively clean. Still, almost every CGC 10 scan I’ve looked at seems to show at least some edge or corner wear. That makes me hesitant to buy anything across the CGC 9–10 range: if that’s what I’m seeing on 10s, I’d expect 9.5s and 9s to have more wear on average.

I used AI to informally compare PSA and CGC scans of roughly 50–70 cards that had been cross-graded from PSA to CGC. It suggested that CGC scans make whitening and white spots more visible than PSA scans. But that’s still just comparing scans. I know what a PSA 9 looks like on average in hand, whereas my experience with CGC is much more limited.

For those who own both, how do CGC 9s, 9.5s, and 10s actually compare with PSA 9s in hand? Do CGC scans exaggerate these flaws, or does the visible wear generally reflect what you receive?

EDIT:

Here’s my AI-assisted comparison of visible whitening on graded Pokémon cards, scoring corners and edges from scans.

In this small sample, PSA 9s and 10s looked similar, while the CGC grade groups overlapped substantially. Scans of the same cards also received different whitening scores in PSA and CGC holders, suggesting that imaging and holder differences may affect what we see.

PSA uses at least two scanners (possibly more given how many locations they have). A cheaper one used for most tiers and a better quality scanner for higher tiers. CGC scans tend to be in between in terms of how clear it is to find defects like scratching or whitening imo.

In terms of grades / cross grading, keep in mind both companies are not consistent as it comes down to the individual grader at the end of the day. And both companies use different scales entirely. CGC has admitted that certain subgrades have a higher weight on overall grade whereas PSA its probably similar. At the end of the day, I’ve crossed anywhere from cgc 8.5 through normal cgc 10 to psa 10. Its all a dice roll.

Most CGC 10s I’ve seen are PSA 8/9 under current grading standards, even pristine CGC 10s are often surprisingly off centered. CGC went from being too harsh 8.5 PSA 9/10 cross over darlings to realizing people wanted 10s and then becoming way too forgiving in my opinion. I can’t bother with them anymore except to get decent crack out binder copy value.

As far as images go,
all images will be different. Even in person, lighting will influence what you do and don’t see. The plastic tomb itself obscures flaws. After cracking open enough of them I have figured that out the hard way. Just use your discretion, its the best anyone can do. I chuck pictures into an editing program, and mess with images to try and reveal flaws if it is expensive enough. Inverting the image colors does wonders to show flaws.

As far as grades go-

Psa 9s from 100x certs and forward are definitely on average stronger cards than current cgc 9s. Before that I found the quality to be extremely comparable and cgc was typically the stronger card but the scales changed. A lot of new cert PSA 9 are cgc 10 potential crosses

Seems like you’re looking at older certs.

Most CGC 10s (not pristine) with newer certs are extremely clean cards, especially for vintage.

They are basically 9.5/9.5/10/10 grades.

Pristine can have one 9.5 grade.

Remember CGC went through several scale changes (but actually said they were doing it, unlike psa who don’t all the time but in the shadows).

3 cert blue labels - clean

4 certs blue labels - often loose grades

Earlier cert CGC 10s (black labels) - can be iffy as they were the 9.5 graded to be upgraded to CGC 10 and they can have a 9 subgrade

Current cert 10s - very clean

Remember, psa 8-9s can be often under or over graded, as can PSA 10s. Just look at the card on an individual basis…and remember, psa scans can hide surface and edge issues!

Hope this helps

I’ve wondered if the contrast from the frosted slab edge with psa scans obscures or does not contrast whitening as much as the clearer cgc case against black background for cgc scans, and if the sharper contrast with the cgc scans due to that difference makes rough cuts (not necessarily whitening post manufacturing) more visible. But I don’t know, I’ve just had this thought often when I examine scans when I’m considering buying a card.

I’ve always wondered the same thing. Anecdotally, my BGS/CGC cards don’t display as well as my PSA cards simply because they have a lot more glare. I assume that’s from light coming through the edges. If so, it seems plausible that scans of cards in non-frosted slabs would appear harsher.

I would recommend looking at the card and seeing if it matches your subjective expectations for its grade. There are clean copies and stinkers in every grade at every company. Best to take it card-by-card rather than assuming.

The AI analysis offers the illusion of objectivity. It looks like you just fed the image to an LLM and asked it to output a subjective score?

I think it would be more interesting to get the ai to produce a script that uses computer vision to get real quantitative metrics. It could be a fun project to scale this up to thousands of scans and actually nail down real patterns

Thanks for your feedback. This was an LLM-based visual assessment rather than an objective measurement. I used a predefined scoring rubric, enlarged corner and edge crops, partial blinding, and uncertainty checks. I’d describe it as a structured, exploratory analysis with considerable uncertainty. It may help identify possible patterns or suggest a direction for further investigation.

Computer vision would be a useful next step. One challenge is that scans from different grading eras can differ in scanner hardware, resolution, contrast, and exposure. Those differences would need to be accounted for so that the analysis measures whitening rather than changes in how the cards were imaged.

I followed up on the computer-vision suggestion and built a script to compare 46 pairs of PSA and CGC scans of the same cards. It uses fixed numerical rules to measure whitening-like edge and corner features, rather than asking an LLM to assign scores. I also tested three different detection methods to see how much the findings depend on the algorithm.

Two methods found higher overall coverage in CGC scans; the third gave mixed results. All three found higher average corner coverage in CGC scans at their main settings, but excluding potentially ambiguous boundary features weakened the overall differences.

Since these are paired scans of the same cards, the question remains whether PSA and CGC imaging makes whitening appear different. The results suggest that it can, although the strength of that difference depends on the detection method. Glare, reflections and image processing may contribute, and this analysis cannot separate those effects.

I’ve attached the report with the methods, example images and results.

paired-PSA-CGC-methodology-and-analysis.pdf (4.1 MB)