Centering vs Corners vs Edges vs Surface: What Actually Costs You the 10

Any one of the four criteria can cost you the 10. Beckett states plainly that "cards are graded on four key categories: centering, corners, edges and surface," and its published examples show that a single weak category drags the final grade down with it: centering 10, corners 6, edges 10, surface 10 comes back as a 7. PSA's Gem Mint 10 is written the same way, with a centering tolerance of approximately 55/45 front and 75/25 back sitting alongside its requirements for corners, edges, surface and focus. A flaw in one place is a flaw in the grade.

What makes centering different is not that it matters more, but that it is the only one of the four you can turn into a number at home. Left border 2.6 mm, right border 2.2 mm, and you know your card is 54/46 and inside PSA's line before you spend a cent. Corners, edges and surface are judged by a grader under magnification, and no company publishes a numeric threshold for any of them. That asymmetry should shape how you pre-screen: measure centering first because it is cheap and objective, then spend your attention on surface, corners and edges in that rough order.

The rest of this article explains what each criterion actually is, how PSA, BGS, CGC and TAG weigh them against each other, and what the gem-rate data says about how often the 10 slips away.

The four criteria and what graders are looking at

Beckett's definition of centering is the clearest one-liner: "Centering refers to the alignment of the card within its borders. A card with perfect centering has even borders on all sides." SGC's 10 Gem description is a useful checklist for the other three, asking for "sharp focus, four sharp corners, free of stains, no breaks in surface gloss, no print or refractor lines, and no visible wear." CGC's Pristine 10 adds "flawless color and registration," a reminder that surface includes the print itself, not only damage after printing.

Centering also has a property the others do not: it is fixed at the factory cut and cannot change. Corners and edges degrade with handling, and surface picks up scratches in a sleeve, but your card's centering on the day it was cut is its centering forever. The only way to "improve" it is trimming, which PSA treats as evidence of alteration (no-grade N1).

How each company combines the four

PSA: one number, centering written into every grade

PSA does not publish subgrades. Each grade description lists a centering tolerance next to the corner, edge and surface language, and the final number reflects the whole card. The centering clause is explicit all the way down: 55/45 at 10, 60/40 at 9, 65/35 at 8, 70/30 at 7, 80/20 at 6, 85/15 at 5 and 4, 90/10 at 3 and below, with 90/10 backs from 9 to 4 and 75/25 at 10.

Two PSA rules tie centering to the other criteria. Half-point grades between 2 and 9 are available to cards that "exhibit high-end qualities within each particular grade," and PSA says there "will be a clear focus on centering" in that decision. And the OC qualifier exists for the card that is perfect except for centering: PSA's glossary example is a card "which exhibits all the qualities of a NM-MT 8, but is 90/10 centered left to right," which "will receive a grade of NM-MT 8oc." Since June 2021 PSA assigns OC at the grader's discretion and usually lowers the number instead.

BGS: four subgrades, and the lowest one rules

Beckett prints all four subgrades and has published how they combine. From Beckett's own process article: "The overall numerical grade is not a simple average of the four subgrades… the lowest grade is the most heavily weighted." The examples are worth memorizing because they show exactly how much a single weak category costs:

Beckett's phrasing on the second row: "The most this card could receive was .5 above the lowest sub-grade." And on the general rule: "the final grade rarely, if ever, exceeds two levels above the lowest of the four characteristic grades."

So at Beckett, centering is exactly as important as the other three, no more and no less: whichever subgrade is lowest sets the ceiling. A card measuring 58/42 on the front takes a centering subgrade of 8 (Beckett's 8 line is 60/40 both ways) and, by the examples above, cannot finish higher than about 8.5 no matter how perfect the corners are. A 50/50 card with a corner 9 finishes around 9.5, not 10. The BGS centering standards article has the full ladder and the Gold vs Black distinction.

CGC: subgrades only on Pristine 10

CGC's sub-grade categories are the same four, "Centering, Surface, Corners and Edges," but CGC discontinued optional sub-grades in 2022, noting that "fewer than 20% of submitters were requesting sub-grades." Today a CGC Pristine 10 is automatically a card with all four at 10, and CGC's Gem Mint 10 is defined as a card at 10 overall where "one of the grading criteria does not meet the requirements of a Pristine 10," with centering "not to exceed approximately 55/45" front and 75/25 back. CGC describes its process as multiple graders examining every card "aided by advanced technology," reaching a consensus on the final grade.

TAG: eight subgrades, measured by machine, reviewed by humans

TAG breaks each criterion into front and back, producing "eight subgrades: front and back centering, front and back corners, front and back surface, and front and back edges," and scores the card on a 1000-point scale that maps to a 10-point grade. Its DIG report lists every defect it found (TAG calls them DINGS, "Defects Identified of Notable Grade Significance"), imaged with what TAG describes as photometric stereoscopic imaging and viewable at up to 800x magnification. TAG is explicit that this is machine-assisted rather than machine-only: "All defects identified by TAG's grading system are thoroughly reviewed by human grading experts before Grade Finalization."

For centering, TAG's Gem Mint 10 band (950–989 points) is ~55/45 front with ~70/30 (sports) or ~65/35 (TCG) back; Pristine (990–1000) is ~51/49 front. TAG's structure is the closest any company comes to telling you which of the four criteria cost you the points, which is why the TAG centering guide is worth reading even if you submit elsewhere.

Which criterion costs the 10 most often?

No grading company publishes a breakdown of why cards miss the 10, so anyone who tells you "it is always centering" or "surface is 70% of failures" is guessing. What is published is the overall gem rate, and it frames the problem:

• GemRate data reported by cllct for the first half of 2025: PSA gem rate 43% across 8.9 million cards, down from 45% industry-wide in 2024. TCG cards gemmed at 50%, with 2020s cards at 59%; sports cards at 34%. CGC 53% overall and 60% on TCG; Beckett 53% overall and 55% on TCG; SGC 21%, reflecting a vintage-heavy mix.
• PokeInvest's PSA pop-report tally as of August 21, 2026: an overall PSA 10 rate of 43.9% on more than 15 million graded Pokémon cards, 60.3% for modern (2023 and later) and 9.9% for vintage (pre-2003).

Read those numbers with the four criteria in mind. A modern pack-pulled card almost always has four sharp corners and clean edges out of the pack, so when 40% of them fail the 10, it is centering or surface doing it. Collectors consistently report that on modern holofoil, surface (print lines, holo scratches, roller marks) and centering are the two recurring culprits, with centering the easier of the two to see in a scan. On vintage, the picture inverts: corner wear and surface aging dominate, and centering is often the thing that is fine relative to the grade the rest of the card supports. Neither is an official statistic; both are consistent with which criteria change with age and which do not.

The PSA 9 vs PSA 10 centering article goes further into why the single flaw that turns a 10 into a 9 is so often a 57/43 front.

Why centering is the one you can settle before submitting

Centering is two lengths and a division. Measure the left border and the right border from the card edge to the outer edge of the printed frame, add them, and divide each by the total:

• Left 2.6 mm, right 2.2 mm → 2.6 / 4.8 = 54.2% → 54/46. Inside PSA's 55/45, BGS 9.5's 55/45 both ways, CGC Gem Mint 10's 55/45, TAG Gem Mint's ~55/45.
• Top 3.0 mm, bottom 2.3 mm → 3.0 / 5.3 = 56.6% → 57/43. Outside all of those lines by a hair, on the 9 tier at PSA, CGC, SGC and TAG, and a centering subgrade of 8 at BGS.

That second measurement is the entire conversation about this card. The worst axis governs (PSA: "the percent of difference at the most off-center part of the card"), so a 54/46 left/right does not rescue a 57/43 top/bottom. You can establish this with a caliper, a flatbed scan and a pixel ruler, or the free card centering calculator, which measures from a photo using manual guide lines and maps the result to PSA, BGS, CGC, SGC and TAG standards. Precision matters: on a 2 mm border, a 0.25 mm error shifts the ratio by about 6 points, so shoot straight down, out of the sleeve, and measure each side at two points to catch a tilted cut.

Now try to do the same for a corner. There is no published "corner tolerance" in millimeters of whitening at any company. You can look under a 10x loupe and decide the tip is sharp, but you cannot know the grader's threshold the way you know 55/45. That is the whole reason centering should be step one: it is the only criterion where a home measurement is directly comparable to the written standard. The pre-grading checklist is built around this order.

A practical priority order when choosing cards to submit

This is collector guidance, not a grading-company rule, but it follows from what is measurable and what is common:

1. Centering first, with numbers. Measure both axes, front and back. If the worst front axis is outside 55/45, the card cannot be a PSA 10, BGS 9.5, CGC Gem Mint 10, SGC 10 or TAG Gem Mint 10 on published standards, regardless of how perfect the rest is. That eliminates cards for free. Keep 56/44 to 60/40 cards in a "9 candidate" pile rather than discarding them; a PSA 9 can still be worth grading.
2. Surface second, under light. Tilt the card under a bright lamp and look across the face at a low angle for print lines, scratches, dimples and factory roller marks, especially on holofoil and chrome stock. Check the back too; surface on the reverse counts.
3. Corners third, under magnification. All four, front and back. Look for the tiny white dot or fuzz at the tip that is invisible at arm's length.
4. Edges fourth. Run your eye along each edge for chipping, rough cuts and silvering on dark-bordered cards, which show edge wear far more than white borders.
5. Then decide the company. If the card is 50/50 to 51/49 with nothing else wrong, the Pristine tiers (BGS 10, CGC Pristine, SGC PRI, TAG Pristine) are in reach. If it is 52/48 to 55/45, the gem tier is open everywhere. If the back is 61/39 to 75/25, PSA and CGC are the gem-tier options because BGS 9.5 caps backs at 60/40 and TAG at ~70/30 sports or ~65/35 TCG.

A worked decision: a modern holo measures 53/47 and 52/48 on the front, 66/34 on the back, and under the lamp shows one faint print line through the holofoil. Centering says gem everywhere except BGS 9.5 and TAG TCG on the back. The print line is a surface question you cannot quantify, but SGC's own 10 GM wording ("no print or refractor lines") and CGC's "flawless color and registration" for Pristine tell you it is a real risk. The centering did not cost this card the 10; the surface might. Submit knowing that, or pick a copy without the line.

Bottom line

• All four criteria are required for a 10; Beckett's published examples show the lowest subgrade sets the ceiling (C10/Co6/E10/S10 → 7).
• Centering is the only criterion with a published number you can measure at home: 55/45 front for the gem tier at PSA, BGS 9.5, CGC, SGC and TAG; 50/50 for the pristine tiers.
• Corners, edges and surface have no numeric thresholds; you screen them by inspection, and the grader's threshold is not something you can verify in advance.
• Gem rates run roughly 43% at PSA overall and around 60% on modern Pokémon, so even pack-fresh cards fail the 10 often, mostly on centering and surface, collectors report.
• Measure centering first because it is free and objective, then spend your time on surface, corners and edges, then choose the company whose back tolerance your card actually meets.

Frequently asked questions

Does centering matter for a PSA 10?+

Yes. PSA's Gem Mint 10 standard states the image must be centered within a tolerance not to exceed approximately 55/45 on the front and 75/25 on the reverse. A card that is flawless everywhere else but measures 60/40 on its worst front axis sits on PSA's 9 line. PSA notes a small variance may be permitted for eye appeal, but centering is a written requirement, not a tiebreaker.

What are the four grading criteria?+

Centering, corners, edges and surface. Beckett states that cards are graded on those four key categories and prints a subgrade for each. CGC uses the same four names for its Pristine 10 subgrades. TAG splits each into front and back for eight subgrades on its DIG report. PSA does not publish subgrades but describes each of these attributes in its grade definitions.

Which is worse for the grade, a soft corner or off-centering?+

It depends on how far off each is. Beckett's published examples show the final grade is pulled toward the lowest subgrade: a card with centering 10, corners 6, edges 10 and surface 10 receives a 7. A card with centering 8 and three 9.5s lands around 8.5. A minor corner touch on an otherwise gem card and a 58/42 front are both typically a one-tier drop; a badly dinged corner costs far more than mild off-centering.

Can I measure corners, edges and surface at home like centering?+

Not to a number. Centering is a ratio of two border widths, so a ruler, caliper, scan or app produces a figure you can compare to a published tolerance. Corners, edges and surface are judged by inspection under magnification and light, and no grader publishes a numeric threshold for them. You can screen them well with a 10x loupe and raking light, but only centering gives you a pass or fail before you pay a fee.

Why do modern cards fail the 10 so often if they come straight from the pack?+

Any one of the four criteria can block the 10. Collectors report that on pack-fresh modern cards, corners and edges are usually fine, so the two most common reasons are centering outside 55/45 and surface flaws such as print lines, scratches on holofoil or factory roller marks. GemRate data reported by cllct showed a 43 percent PSA gem rate in the first half of 2025, with 2020s TCG cards at 59 percent, so even modern product fails more than a third of the time.

What does a BGS Black Label require beyond centering?+

All four subgrades must be 10. Beckett's centering line for Pristine 10 is 50/50 all around on the front and 55/45 or better on the back, and the gold label version allows one 9.5 subgrade with three 10s. So a 50/50 card with a microscopic edge flaw can be a gold Pristine 10 but not a Black Label. Beckett also limits Black Label to modern cards, not BVG cards from 1980 and older.

Sources & further reading

• PSA Grading Standards
• Beckett Grading Scale
• Beckett: Walking through the BGS process
• CGC Cards: Perfect 10 and sub-grades
• TAG Grading: What is the TAG DIG report
• cllct: How many cards grade as 10s? Gem rate at 43% so far in 2025