PSA grading Pokémon cards: what actually decides the grade

Grading a Pokémon card is a bet. You pay a fee and post the card away, and weeks later a number comes back that can multiply what the card is worth — or leave you slightly poorer than when you started. The useful question is not should I grade my cards in the abstract. It is which of my cards can still reach the grade that makes the fee worth paying, and how much of that I can work out before the card leaves the house.

A surprising amount, as it turns out. Not all of it — but the single factor that most often costs a card its 10 is also the only one that is pure geometry, published as a number, and measurable on your kitchen table.

The four things PSA is assessing

PSA grades on centering, corners, edges and surface. A card is capped by its worst attribute: razor-sharp corners do not rescue a card whose borders are visibly lopsided, and perfect centering does not rescue a card with a crease.

Three of those four are condition — they are about what has happened to the card since it left the printer. The fourth, centering, is about how the card was cut at the factory. It was decided before anyone opened the pack, it will never change, and unlike the other three it is not a judgement call about wear. It is a ratio.

That distinction matters more than it sounds, because it means centering is the one factor you can settle in advance, at home, for nothing.

Centering is a ratio, and PSA publishes the thresholds

Centering is the relationship between opposite border widths — left against right, top against bottom. A card cut dead centre measures 50/50. The convention is to quote the worst side, so "60/40" means neither border may take more than 60% of the pair.

PSA publishes the tolerance for every grade, front and reverse:

Grade	Front	Reverse
10 GEM-MT	55/45	75/25
9 MINT	60/40	90/10
8 NM-MT	65/35	90/10
7 NM	70/30	90/10
6 EX-MT	80/20	90/10
5 EX	85/15	90/10
4 VG-EX	85/15	90/10
3 VG	90/10	90/10
2 GOOD	90/10	90/10
1.5 FR	90/10	90/10

Read that as: the widest the borders may differ and still reach that grade. Everything else about the card has to clear its own bar too — this is a ceiling, not a promise.

The 5% rule, and why the hobby cannot agree on 55/45 versus 60/40

Ask what centering a PSA 10 needs and you will get two answers, confidently, from people who both collect seriously. One says 55/45. The other says 60/40. They are both right, and the reason is a rule that rarely gets explained properly.

Every grade from NM 7 upward has two thresholds, not one. The number in the table is the floor a card always clears. PSA then allows a further 5% on front centering as eye-appeal leeway, and whether a card sitting inside that band actually gets the grade comes down to a grader's judgement of how the card looks.

So for a 10:

• A 55/45 front clears the published floor. Centering is not what stops this card.
• A 60/40 front is inside the 5% band. It is possible, and it depends on the grader.
• A 61/39 front is outside the band, and no amount of eye appeal brings it back.

This is why an honest pre-grade read is two answers rather than one — what a card is guaranteed, and what it might reach. Anything that hands you a single confident grade before the card has been graded is telling you something it cannot know. There is more on the specifics of the 10 in what makes a card a PSA 10.

The back is almost unfailable — with one exception

Look down the reverse column and something jumps out: it reads 90/10 for every grade from 9 all the way down to 1.5. A back has to be dramatically miscut to fail that. Reverse centering is effectively a non-event.

Except at the top. The 10 is the only grade whose reverse tolerance is genuinely failable, at 75/25. That gives you a clean rule for how to spend your time: measure the front first, and only bother measuring the back if the front came in at 55/45 or better. If the front cannot reach a 10, the back cannot cost you one.

The OC qualifier

When centering falls below the minimum for a grade, PSA does not quietly drop the card a rung. It adds an "OC" (off-centre) qualifier, and the card comes back as, say, "PSA 9 OC" — a 9 in every respect except that its centering did not meet the 9's bar.

You can ask for an unqualified grade instead. In that case the same card drops to the next grade whose centering tolerance it actually does meet. Qualified cards typically trade at a discount to their clean equivalents, which is worth knowing before you choose.

What a photograph genuinely cannot tell you

Here is where we will argue against our own product for a moment, because the alternative is pretending.

Centering can be measured from a photo, because it is geometry: the borders are either wider on one side or they are not. Corners, edges and surface cannot be. A soft corner, a whitened edge, a light crease, a print line, a dent — these are three-dimensional defects and specular effects, and a single flat photograph carries almost no information about them. A dent is a highlight that moves when you tilt the card under a lamp. A still image does not move.

So no app can look at a photo and tell you a grade. What an app can honestly do is measure the one factor that is measurable, and then ask you about the three that are not — by recognition rather than by self-assessment, because "are your corners sharp?" gets a yes from everybody, whereas "which of these three reference corners looks like yours?" gets an answer you can use.

Any tool that returns "your card will grade PSA 10" from one picture is guessing, and the parts it is guessing about are three quarters of the decision.

Japanese cards

Two things collectors ask constantly, and the answers are unrelated to each other.

PSA grades Japanese Pokémon cards, and grades them on the same four factors. Separately, most identification tools are far worse at Japanese cards than English ones, because the catalogue behind them is English-first and a Japanese set number does not resolve against an English set list. Cardient carries 221 Japanese sets and roughly 20,900 Japanese cards alongside the English catalogue, so a Japanese card resolves to its exact printing rather than to whatever English card shares its number.

Japanese print runs also have a reputation for tighter cuts than their English counterparts of the same era — which is a claim about averages, not about the specific card in your hand. Measure it.

What to do before you pay a submission fee

The sequence that wastes the least money:

1. Measure the front centering. It is free, it takes under a minute, and it is the factor most likely to rule out the grade you were hoping for.
2. If the front cannot reach the grade you need, stop. Nothing further about the card will rescue it.
3. If it can, look hard at corners, edges and surface — under a lamp, tilting the card, which is the only way those three reveal themselves.
4. Then compare what the card is worth raw against what it is worth at the grade it can realistically reach, and weigh that gap against PSA's current fee for the tier you would submit under. Both PSA's fees and their service tiers change, so read them from PSA directly rather than from an article.

Cardient does step 1 for free and unlimited — it straightens the card for camera perspective first, then measures both axes on both faces and shows you the guaranteed and the possible grade side by side. How to measure centering walks through the arithmetic if you would rather do it by hand.

---

Cardient is an independent app and is not affiliated with, endorsed by, or sponsored by Professional Sports Authenticator (PSA), Beckett (BGS), CGC, or The Pokémon Company International. PSA centering tolerances quoted here are PSA-published; BGS and CGC tolerances are community-cited, not official. A centering measurement is an estimate, not a grade.