Most graded cards arrive with a single number on the label — a 9, a 10 — and that’s the whole story you get. But that one figure is the combination of several separate measurements, and the way they combine can hide as much as it reveals. This is what subgrades are, why they matter, and what they tell you that an overall grade can’t.

Every grade is really four measurements

A card’s condition isn’t one thing. It’s assessed across four independent factors — centering, corners, edges and surface — each scored on its own. A subgrade is simply the score for one of those factors. The overall grade is what you get when you combine them. (For how the four are measured, see how card grading works.)

Why the overall number can mislead

Two cards can carry the same overall grade and be in noticeably different condition. One might be strong everywhere — good centering, sharp corners, clean edges, flawless surface. The other might be near-perfect on three factors but let down by one: a slightly off-centre print, or a single soft corner. Same headline number, very different cards. The overall grade tells you the result; the subgrades tell you why.

That “why” matters, because the weakest factor usually sets the grade. If you know a card is a 9 because of centering — with everything else at gem-mint level — you understand the card far better than the number alone allows.

What subgrades are good for

• Buying with your eyes open. Subgrades let you see exactly what you’re paying for, instead of trusting one figure. A 9 held back only by centering is a different proposition to a 9 with edge wear.
• Knowing what can’t improve. Centering is fixed at manufacture — it will never get better. Corners and edges can degrade with handling. Subgrades show you which is which.
• Spotting genuine quality. A card that scores high across every factor is rarer, and more desirable, than one that scrapes the same overall grade on the back of a single strong factor.

Why measurement makes subgrades trustworthy

Subgrades are only useful if each one is reliable on its own. That’s difficult by eye — it’s hard to put a consistent number on “how sharp” a corner is across thousands of cards. Measuring each factor physically — the actual border ratios, the actual corner geometry, the actual edge and surface condition — means every subgrade is a number you can stand behind and reproduce, not a gut call. It’s the same reason two measured grades agree where two opinions might not.

Reading them on your report

Your CALIBRE grading report shows each factor’s measurement, not just the final grade — so you can see precisely where a card stands and what, if anything, kept it off the top. Nothing is buried in a single digit.

Want to see your own cards broken down factor by factor? Submit a card and read its measurements in full.

Holo, reverse-holo and the new wave of textured special-art cards are the ones collectors chase — and the ones that are hardest to grade consistently. The very shine that makes them desirable is also what hides wear, throws off the eye, and confuses photo-based grading. Here’s why foils are such a problem, and how measuring a card sees through the glare.

Shine is the enemy of a consistent read

A foil surface doesn’t reflect light evenly. Tilt a holo a few degrees and the whole card changes — bright bands sweep across it, the artwork flares, and a hairline scratch that was obvious a second ago disappears into the glare. That sensitivity to angle is exactly what makes grading by eye unreliable on holos: what you see depends on how you happen to be holding the card and where the light is.

It cuts both ways. Glare can hide genuine scratches, so a worn card looks cleaner than it is. And the foil’s own pattern — the lines and sparkle baked in at manufacture — can be mistaken for scratches, so a mint card looks damaged. Two careful graders under two different desk lamps can reasonably disagree.

Texture adds a second problem

Modern special-art, alternate-art and “textured” cards take this further. The surface is deliberately embossed with a raised pattern you can feel. Under raking light that texture casts tiny shadows — and now the grader has to separate three things at once: the intended texture, the foil’s shine, and any actual surface damage. By eye, that’s genuinely hard. It’s why surface is the factor that trips people up most on premium cards. (For the four factors generally, see corners, edges and surface.)

Why AI grading struggles here too

Photo-based and AI grading inherit the same problem — and arguably make it worse. A model trained on images learns what a grade tends to look like, but a holo photographed at a slightly different angle looks like a different card. The shine it learned to ignore on one scan becomes a “scratch” on the next. It’s predicting an appearance, not measuring a surface. We go into that distinction in AI grading vs physical measurement.

How measurement sees through the glare

The fix isn’t a sharper eye or a bigger training set — it’s changing what you capture. Instead of one photo under one light, the card is measured under controlled lighting from multiple, known directions. Because every light position is fixed and repeatable, the foil’s shine becomes information rather than a distraction: a true scratch behaves one way as the light moves, while the foil’s manufactured pattern behaves another. The surface is reconstructed from the physics of how light actually reflects off it, not guessed from a single flattering or unflattering frame.

That’s the whole point of measuring rather than judging: the result doesn’t depend on the angle you happened to catch. A holo graded this way scores the same whether it’s shimmering or dull in your hand. See how card grading works for the full method.

What it means for your holos

If your best cards are foils — and for most collectors, they are — this is exactly where a measured grade earns its keep. The cards most likely to be mis-judged by eye or by a photo are the ones a controlled, physical read handles most cleanly. You don’t have to worry that a grader caught your card in bad light, or that a scan flattened its texture into a flaw.

Got a textured chase card you’re unsure about? That’s exactly the kind of card worth measuring properly — submit it for grading and see precisely what its surface is doing.

Centering is set when a card is cut, but corners, edges and surface are where everyday handling shows up — and where many cards lose their top grade. Here’s what each one means.

Corners

Corners are the first thing to go. Even careful handling can soften a sharp corner or bring tiny whitening to the tip. Graders look for crisp, sharp corners with no rounding or fraying — and because there are four of them, the weakest corner tends to set the standard.

Edges

Edges pick up chipping and whitening, especially on dark-bordered cards where any wear shows instantly. Clean, unbroken edges are the goal; nicks and whitening along an edge pull the grade down.

Surface

Surface covers scratches, print lines, dimples, indentations and any loss of gloss. Holo and foil cards can show surface issues more readily. It’s the factor most affected by how a card has been stored and handled.

How CALIBRE assesses them

Rather than a quick visual call, each factor is assessed from precise imaging and recorded in your report, alongside centering. The lowest factor tends to limit the overall grade — which is why protecting corners, edges and surface matters. See how to store your cards.

Start your order →

Not all “tech-driven” grading is the same. There’s a real difference between an AI predicting a grade and a system measuring the card. Here’s what’s actually happening.

What AI grading does

AI grading trains a model on lots of previously-graded cards, then predicts a grade for a new card from its photos. It’s fast, but it’s a prediction — an estimate of what a grade “should” be based on patterns, not a direct measurement of your specific card. Predictions can be confidently wrong, and they’re hard to explain.

What physical measurement does

CALIBRE measures the card itself: border widths for centering, precise imaging for corners, edges and surface. The grade comes from those measurements, so it’s repeatable — the same card produces the same grade every time — and every figure is recorded in your report. See how it works.

Why the difference matters

Measurement is transparent and consistent. You can see the numbers behind your grade, and you won’t get a different answer on a resubmission because the model had a bad day. For a decision that affects a card’s value, measured beats predicted.

Start your order →

When a card is graded it’s sealed in a slab — but the slab does more than protect the card. Here’s why “tamper-evident” matters.

Protection

The slab shields your card from handling, bending, moisture and light, holding it in the exact condition it was graded in. A graded card you bought five years ago should look the same today.

Tamper-evidence

Just as importantly, the slab is designed so it can’t be opened and resealed without leaving signs. That’s what stops someone swapping the card inside or altering the label. A grade you can’t trust the seal on is a grade you can’t trust at all.

The certificate link

Each slab carries a unique certificate number that ties the physical card to its grade and report in our registry. Together, the tamper-evident seal and the verifiable certificate are what make a grade meaningful — see how to verify one.

Why it matters when buying

A photo of a slab proves nothing on its own. A tamper-evident slab plus a certificate you can check is the difference between trusting a grade and hoping it’s real.

Start your order →

“Grading” can mean very different things depending on who’s doing it. This guide explains how card grading actually works — what’s assessed, how the four factors combine into a single grade, and why measuring a card produces a more consistent result than judging it by eye or predicting it with AI.

What grading involves

At its core, grading answers one question: what condition is this card in? To answer it consistently, a grader has to assess four separate things — centering, corners, edges and surface — and combine them into a single number on a 1–10 scale. The detail is in how each of those is judged, and how reliably. New to the scale itself? Start with card grades explained.

Centering

Centering is how evenly the printed design sits within the card’s borders — equal margins left and right, top and bottom, on both faces. Cards are cut from large sheets, and small shifts in that cut leave one border wider than the one opposite. Because it’s set at the moment of cutting, centering can’t be improved, which makes it the factor that most often keeps a card off the top grade. We cover it in depth in what is centering.

Corners

Corners are usually the first thing to wear. Even careful handling can soften a sharp corner or bring faint whitening to the tip. A grader looks for crisp, sharp corners with no rounding or fraying — and since there are four, the weakest one tends to set the standard.

Edges

Edges pick up chipping and whitening, which shows instantly on dark-bordered cards. Clean, unbroken edges are the goal; nicks and whitening pull the grade down.

Surface

Surface covers scratches, print lines, dimples, indentations and any loss of gloss. Holo and foil finishes can show surface issues more readily. It’s the factor most affected by how a card has been stored and handled.

How four factors become one grade

The four factors are assessed separately, then combined — and crucially, the lowest factor tends to limit the overall grade. A card that’s a 10 on three factors but a 7 on corners isn’t a 10. That’s why a single number can hide a lot, and why a good grading report shows the subscores, not just the headline. A flawless card with one soft corner grades very differently from one that’s marginal everywhere.

Why human grading varies

Traditional grading asks a person to make those judgements and combine them. People are remarkably good at it — but they’re still people. Lighting, fatigue, the order cards are seen in, and ordinary human variation all nudge a borderline card one way or the other. A card on the line between a 9 and a 10 is exactly where two graders are most likely to disagree — and that inconsistency has real cost, which is why some collectors resubmit the same card hoping for a better number. We unpack this in why two graders can disagree.

Why AI prediction isn’t measurement

Some services use AI to predict a grade from photos. It’s fast, but a prediction is an estimate of what a grade “should” be based on patterns in past cards — not a direct measurement of your specific card. Predictions can be confidently wrong and are hard to explain. Measurement is different: it produces numbers you can see and check. Our market comparison sets the approaches side by side.

How CALIBRE measures

CALIBRE doesn’t ask a person to judge the card, and it doesn’t predict from a model — it measures. Centering comes from the actual border widths; corners, edges and surface are assessed from precise imaging. Because the inputs are measurements rather than impressions, the grade is repeatable: the same card produces the same grade every time. See the full process in how it works.

Your grading report

Every grade comes with a report showing the overall grade, the four subscores, and the measurements behind them — a transparent record rather than a verdict you have to take on trust. Each report ties to a unique certificate published in our registry, so any grade can be verified. Here’s how to verify a certificate.

Want to see it on your own cards? Start your order →

Send the same card to two human graders and you can get two different grades. That’s not a scandal — it’s a known limitation of grading by eye. Here’s why it happens, and how measurement removes it.

Why opinions diverge

Traditional grading asks a person to judge centering, corners, edges and surface and combine them into one number. People are remarkably good at this — but they’re still people. Lighting, fatigue, the order they see cards in, and simple human variation all nudge a borderline card one way or the other. A card on the line between a 9 and a 10 is exactly where two graders are most likely to disagree.

The cost of inconsistency

Inconsistency is expensive. The gap between a 9 and a 10 can be a large gap in value, so a grade that could have gone either way introduces real uncertainty — and it’s why some collectors resubmit the same card hoping for a better number.

How measurement fixes it

CALIBRE doesn’t ask a person to judge the card — it measures it. Centering comes from the border widths; corners, edges and surface are assessed from precise imaging. Because the inputs are measurements rather than impressions, the same card produces the same grade every time, and every measurement is recorded in your report. See exactly how it works, or read how we measure centering.

Proof you can check

Every grade we issue is recorded with its measurements and published with a verifiable certificate. Browse the full population of cards we’ve graded in the registry.

Start your order →

Of all the things that decide a card’s grade, centering is the one collectors underestimate most. Here’s what it is, how it’s measured, and why it can be the difference between a 9 and a 10.

What centering means

Centering is how evenly the printed design sits within the card’s borders. In a perfectly centred card, the border is the same width on the left and right, and the same top and bottom — on both the front and the back. Cards are cut from large printed sheets, and small shifts in that cut leave one border wider than the one opposite it.

Why it matters so much

You can have flawless corners, clean edges and a perfect surface, but if the design is noticeably off to one side, the card won’t reach the top grade. Because centering is fixed at the moment the card is cut, it’s the one factor you can’t improve — which is exactly why high grades on poorly centred cards are rare and valuable.

Front and back both count

A card can look well-centred on the front and be badly off on the back. Grading considers both faces, so it’s worth checking the reverse before you assume a card is a top-grade candidate.

How CALIBRE measures it

Rather than eyeballing the borders, CALIBRE measures them. The card is imaged and the border widths are measured precisely on both faces, so centering is a number, not an opinion — and the same card produces the same result every time. It’s recorded in your report alongside corners, edges and surface. See how it works.

Think you’ve got a well-centred gem? Start your order →

Every card passes through CALIBRE in roughly sixty seconds. In that time, over one hundred physical measurements are captured and analysed. Here’s what each one does.

The principle

Every physical object has measurable properties. A trading card is no exception. Its thickness, weight, surface texture, material composition, and response to different wavelengths of light are all measurable quantities. Take enough of these measurements with enough precision, and you can describe the card completely — its authenticity, its condition, its history.

CALIBRE’s grading approach is built around this idea. Rather than looking at a card and forming a judgement about it, CALIBRE measures the card and reports the results. The grade emerges from the data, not from interpretation. The same measurements also create a card fingerprint — a physical signature that can be matched on future submissions to confirm the card’s identity over time.

Transmitted light — the black core test

Genuine Pokémon cards, Magic cards, sports cards and most trading cards are manufactured with a thin black core layer sandwiched between the front and back printing. In genuine stock this is carbon-black paper — paper manufactured with carbon black pigment integrated into the pulp, not printed on the surface. Carbon black is an extremely efficient absorber across visible light and into UV and IR, which is why a genuine card is genuinely opaque under transmitted light.

CALIBRE automates this test with precision. A calibrated LED panel below the card shines uniform white light upward through the card. A high-resolution camera above captures the transmitted light pattern. Genuine cards appear as solid black shapes. Fakes glow visibly.

Counterfeit cards are typically printed on standard white or off-white stock with a black layer printed on the surface rather than integrated into the paper. Surface printing does not achieve the same opacity: light leaks through.

The same pass also enables card fingerprinting. The carbon-black fibres inside the core are randomly distributed during manufacture. That distribution is effectively unique per card — like a snowflake — and it does not change when the face of the card is scratched, cleaned, or cosmetically repaired. That is why a fingerprint can still match the same physical card even if the visible surface changes.

Photometric stereo — surface reconstruction

If you photograph a card under flat overhead lighting, the result is a picture of the card’s artwork and text. If you photograph the same card under light coming from one side, the shadows cast by every tiny scratch, dent, and print defect become visible.

Photometric stereo takes this principle further. CALIBRE photographs each card four times, with light coming from four different directions — left, right, front, and back. By comparing how the shadows change between these four images, an algorithm reconstructs the actual three-dimensional surface of the card at microscopic scale.

The output is not a photograph but a depth map. Every scratch is measured by its depth and length. Every dent by its volume. Every print defect by its extent. At CALIBRE’s imaging resolution of eighty pixels per millimetre, surface features as small as a tenth of a millimetre are captured precisely.

Structured light — 3D shape measurement

Cards are supposed to be flat. In reality, they warp. Heat, humidity, storage conditions and manufacturing variation all cause cards to develop subtle bows or twists. A bowed card is graded lower than a flat one, and structured light is the measurement that detects this.

A digital light projector casts a series of precisely known stripe patterns onto the card. A camera records how those stripes bend as they hit the card’s surface. The mathematics of how a known projected pattern distorts when it lands on a curved surface allows the card’s three-dimensional shape to be reconstructed with sub-millimetre precision.

The result is a contour map of the card. Perfectly flat cards produce flat maps. Bowed cards produce curved maps. The precise shape and magnitude of any warp is quantified and included in the grade.

Ultraviolet fluorescence — invisible signatures

Most people have seen the effect of UV light in clubs and at airports. Certain materials — banknote security features, white shirts, bank security strips — glow under UV light because they contain chemicals that absorb ultraviolet wavelengths and re-emit them as visible light. This is called fluorescence.

Modern trading cards contain security features that behave the same way. Printing inks have specific fluorescence responses. Coatings and varnishes fluoresce distinctively. Holographic elements contain signatures only visible under UV. CALIBRE captures two independent UV images of every card — one at 365nm (deep ultraviolet) and one at 395nm (near ultraviolet) — because different security features respond to different wavelengths.

A genuine card produces a specific, predictable UV signature. A counterfeit card produces a different signature, regardless of how convincing its visible appearance might be. The UV passes are one of CALIBRE’s strongest authentication tools.

Infrared transmission — seeing beneath the surface

Infrared light has a longer wavelength than visible light. This means it penetrates thin layers of material that are opaque to our eyes. CALIBRE uses 850nm infrared illumination to see through the top layer of printing and inspect the structure of the card underneath.

This is how CALIBRE detects layer substitution, hidden tampering, and damage that has been cosmetically repaired. A card that has been re-backed — where the back has been replaced to hide damage — looks perfect to the eye but reveals its history under infrared.

Blue light — coating and surface treatment

Blue light at 450nm wavelength sits at the edge of the visible spectrum and reveals surface features that are hard to see under normal lighting. This includes adhesive residue from sleeve damage, coating inconsistencies, trimmed edges that have been disguised, and surface treatments applied post-manufacture.

Weight — material consistency

Every genuine card of a given type weighs a specific amount. A standard Pokémon card weighs between 1.7 and 1.8 grams. Deviations from this range indicate one of several possibilities: counterfeit card stock (usually lighter or heavier), moisture damage (heavier), or thickness variation (potentially either).

CALIBRE weighs every card to sub-gram precision. This alone is not a grading factor, but combined with the other measurements it provides an important confirmation signal.

Putting it all together

The sixty-second measurement process captures all of the above in sequence: transmitted light for the authentication test, then weight, then the full spectral sequence of UV, IR and blue, then photometric stereo from four angles, then structured light for shape analysis — first on the front face of the card, then after it is flipped, again on the back face.

Thirty-three measurements in total. From those measurements, the grade is calculated deterministically — the same inputs always produce the same output. No interpretation. No judgement. No opinion.