A Centering Check studyJuly 202610,000 graded cards

The first public card-by-card audit of PSA's centering standard.

1 in 11PSA 10 Pokémon cards graded under today's standard exceed PSA's own centering limit

We measured 10,000 randomly sampled PSA-graded cards and compared the 4,840 fronts with a datable standard against the limit PSA published when each was graded. Among gem mints graded under the stricter 2025 standard, 9.0% of fronts are past the 55/45 line - about 1 in 11. Under the older 60/40 standard it is 0.3%, and 4.0% across every datable PSA 10.

4.3% clear-cut4.7% near the line
01The main distribution

The whole PSA 10 pool, measured against its own limit

Each dot below is one PSA 10 front - compliant cards included, colored from harmless to confirmed. Grey dots measured inside PSA's limit; white dots measured a hair past it, within the measurement margin we refuse to call a violation; amber and orange dots are the violations we confirmed by hand. Recent cards are held to the stricter 55/45 limit they were graded under, cards from before the 2025 tightening to the older 60/40.

measured compliantpast the line · within margin, not countedconfirmed · near the lineconfirmed · clear-cuteach dot is one PSA 10 front
Graded under the 2025 standard (limit 55/45)1,159 cards · 104 confirmed past the limit (9.0%)
past the limit →PSA's limit · 55/45our margin50/5055/4560/4065/35worse centering →

106 cards measure a hair past the line but inside the one-point measurement margin - not counted as violations.

Graded under the older standard (limit 60/40)1,570 cards · 4 confirmed past the limit (0.3%)
past the limit →PSA's limit · 60/40our margintoday's limit · 55/4550/5055/4560/4065/35worse centering →

28 cards measure a hair past the line but inside the one-point measurement margin - not counted as violations.

Two things stand out. Violations concentrate in the 2025 era: the pool graded under the stricter limit misses it 9.0% of the time, versus 0.3% for the older era. And the older pool spreads far past today's line: 588 of its 1,570 cards (37%) measure past 55/45 - centering that would not meet a 2025 gem-mint label, sitting under an older one.

4.0%
of PSA 10 fronts exceed PSA's limit
108 of 2,729 with a datable standard
2.4%
across all grades, 6 to 10
115 of 4,840 fronts
0 of 4,840
backs are over their limit
the published back tolerance is loose enough that nothing fails it
02The standard

What "past the limit" means - on a real card

A centering ratio splits the border between opposite sides: 55/45 means the wider side takes at most 55% of the total margin. PSA publishes a maximum ratio for every grade, and tightened the front numbers in early 2025. On the scans below, the solid cream line is the card's cut edge and the solid dark line is the printed frame. Centering is how evenly the space between those two lines is split - left against right, top against bottom - and every card is captioned with its measured ratio next to the limit it was graded against. Click either card to inspect it full-size.

Inside the limit
52/48limit 55/45

PSA 10 · 2021 POKEMON CELEBRATIONS CLASSIC COLLECTION #4 CHARIZARD-HOLO CLASSIC COLL-BASE SET

Verify on PSA →
Confirmed violation
63/37limit 60/40

PSA 10 · 2023 POKEMON JAPANESE SV1S-SCARLET ex #098 TEAM STAR GRUNT SUPER RARE

Verify on PSA →
GradeFront · before 2025Front · 2025Back
PSA 1060/4055/4575/25
PSA 965/3560/4090/10
PSA 870/3065/3590/10
PSA 775/2570/3090/10
PSA 680/2080/2090/10

Because a printed grade reflects the standard in force at grading time, every card this study rates is tested against the front limit published when it was graded - the older numbers for older certificates, the 2025 numbers for recent ones.

03What it means for you

Reading this as a buyer, a submitter, or a vintage holder

Buying a graded card

A gem-mint label does not guarantee gem-mint centering: 1 in 11 of the PSA 10 fronts graded under the current standard is past PSA's own limit. On a high-value purchase, measure the centering from the listing photo before you pay the premium - it takes seconds and the label will not do it for you.

Submitting raw cards

Centering is the one grading factor you can measure exactly before paying a grading fee. Check your card against the current 55/45 front standard first: if it measures past the limit, you know the centering risk going in - before the fee, not after.

Holding older slabs

An older PSA 10 certifies the older, looser standard. 37% of pre-2025 gem mints we measured sit past today's 55/45 line (20% even allowing two points of measurement error). When old and new slabs of the same card trade at the same price, the label is not promising the same centering.

04Method

From 10,000 random cards to 115 confirmed cases

  1. 1

    Sample

    10,000 PSA-graded Pokémon cards drawn uniformly at random by certificate number, grades 6 to 10. About 54% are PSA 10 and 46% are Japanese - no cherry-picking, no curated lists.

  2. 2

    Measure

    Each PSA scan runs through the same Centering Check pipeline anyone can use: the card is detected, flattened to a straight-on view, and the gaps between its cut edge and printed frame are measured into front and back centering ratios.

  3. 3

    Date the standard

    PSA tightened its front standard in early 2025, so each card is compared against the limit in force when it was graded. Cards graded too close to the change to date reliably - about half the sample - are excluded from every rate rather than guessed at, leaving 4,840 fronts with a datable standard.

  4. 4

    Verify by hand

    Every card exceeding its limit was inspected manually. Measurements that misplaced a border were rejected, ambiguous cases were set aside rather than counted, and only confirmed violations are counted and shown. As a robustness check, the entire sample was also measured by two independently trained detector versions - the headline rate held.

4,840
fronts with a datable standard · of the 10,000 sampled - the rest were graded too close to the 2025 change to date, or were not standard cards
389
measured past their limit · every one inspected by hand
115
confirmed violations · a violation must clear its limit by at least a full point - the benefit of the doubt we give PSA and our own pipeline - and survive manual review

Primary sources: PSA's published grading standards, and every card's public certificate on psacard.com - linked on each card below.

05The receipts

Every confirmed case, verifiable on PSA's own site

All 115 front violations confirmed by direct measurement. On each scan, the solid cream line is the card's cut edge and the solid dark line is the printed frame; every caption shows the measured ratio next to the limit the card was graded against. 61 are clear-cut; 54 sit close enough to the limit that measurement error could matter, and are labelled accordingly. Click any card to inspect it full-size and step through them one by one; each also links to its public certificate on psacard.com, so you can confirm the grade yourself.

Sort
showing 16 of 115 cards
65/35T/B
limit 60/40
PSA 9JP·2025 standard

2023 POKEMON JAPANESE SV3a-RAGING SURF #089 PARASOL LADY SPECIAL ART RARE

Verify on PSA →
65/35T/B
limit 60/40
PSA 10·Older standard

2021 POKEMON SWORD & SHIELD FUSION STRIKE #270 FA/ESPEON VMAX FUSION STRIKE-SECRET

Verify on PSA →
63/37T/B
limit 60/40
PSA 10JP·Older standard

2023 POKEMON JAPANESE SV1S-SCARLET ex #098 TEAM STAR GRUNT SUPER RARE

Verify on PSA →
63/37L/R
limit 60/40
PSA 9·2025 standard

2024 POKEMON SVP EN-SV BLACK STAR PROMO #132 GRENINJA ex SFA SPECIAL ILLUSTRATION COLL

Verify on PSA →
62/38T/B
limit 60/40
PSA 10JP·Older standard

2021 POKEMON JAPANESE SWORD & SHIELD START DECK 100 #419 FA/MARNIE'S PRIDE START DECK 100

Verify on PSA →
62/38T/B
limit 60/40
PSA 9JP·2025 standard

2025 POKEMON JAPANESE M1S-MEGA SYMPHONIA #090 ACEROLA'S MISCHIEF SPECIAL ART RARE

Verify on PSA →
62/38T/B
limit 60/40
PSA 9·2025 standard

2025 POKEMON SVP EN-SV BLACK STAR PROMO #174 EEVEE ex PRE SUPER-PREMIUM COLLECTION

Verify on PSA →
62/38T/B
limit 60/40
PSA 9·2025 standard

2023 POKEMON MEW EN-151 #198 VENUSAUR ex SPECIAL ILLUSTRATION RARE

Verify on PSA →
62/38T/B
limit 60/40
PSA 10·Older standard

2022 POKEMON GO #031 MEWTWO VSTAR

Verify on PSA →
62/38T/B
limit 60/40
PSA 9·2025 standard

2025 POKEMON JTG EN-JOURNEY TOGETHER #167 N'S RESHIRAM ILLUSTRATION RARE

Verify on PSA →
62/38L/R
limit 60/40
PSA 9·2025 standard

2025 POKEMON MEG EN-MEGA EVOLUTION #184 LILLIE'S DETERMINATION SPECIAL ILLUSTRATION RARE

Verify on PSA →
61/39L/R
limit 55/45
PSA 10JP·2025 standard

2008 POKEMON JAPANESE GIRATINA VS. DIALGA #007 DIALGA-HOLO G/D-DIALGA-1ST EDITION

Verify on PSA →
60/40L/R
limit 55/45
PSA 10JP·2025 standard

2025 POKEMON JAPANESE SV-P PROMO #232 IONO'S WATTREL SV9 BOX PURCHASE CAMPAIGN

Verify on PSA →
60/40T/B
limit 55/45
PSA 10·2025 standard

2025 POKEMON WHT EN-WHITE FLARE #171 HILDA SPECIAL ILLUSTRATION RARE

Verify on PSA →
59/41T/B
limit 55/45
PSA 10JP·2025 standard

2023 POKEMON JAPANESE SV2a-POKEMON 151 #181 SNORLAX ART RARE

Verify on PSA →
59/41T/B
limit 55/45
PSA 10JP·2025 standard

2025 POKEMON JAPANESE M2-INFERNO X #013 MEGA CHARIZARD X ex

Verify on PSA →
06Caveats

What this study does and does not show

  • It is our measurement, not PSA's

    We measure from PSA's published scan; PSA grades the card in hand and weighs centering alongside corners, edges, and surface. A card past the centering line can still earn its grade on overall eye appeal - so this measures compliance with the published centering rule, not whether any individual grade was "wrong."

  • Cards near the line are uncertain

    The 55/45 standard sits close to dead-center, and a card measured a hair past it can be within our measurement error. Those cases are labelled "near the line" and kept separate from the clear-cut ones everywhere on this page - including the headline split.

  • Not every card was hand-checked

    Manual review covered every card measured past its limit. Cards measured well inside the limit were not individually verified, so the grey dots carry ordinary measurement noise in both directions - a few more true violations may hide just below the line, which makes the headline conservative in that respect.

  • Half the sample has no datable standard

    Cards graded too close to the early-2025 standard change cannot be assigned a limit honestly, so they are excluded from every rate rather than guessed at. An excluded card can hide a violation from our count, but it can never add one - the error from this exclusion runs in PSA's favor, not against it.

07FAQ

Common questions

Can a PSA 10 Pokémon card be off-center?
Yes. In a random sample of 10,000 PSA-graded Pokémon cards, 9.0% of PSA 10 fronts graded under the current 55/45 standard (104 of 1,159) exceed PSA's own published centering limit - about 1 in 11. Across every PSA 10 with a datable standard it is 4.0%. Every case was confirmed by hand and can be checked against its public PSA certificate.
Did PSA change its centering standard?
Yes. In early 2025 PSA tightened its published front centering tolerances - a PSA 10 front went from 60/40 to 55/45. Back tolerances were unchanged. Cards in this study are tested against the standard in force when they were graded.
Are older PSA 10s more off-center than new ones?
Older gem mints were graded against a looser standard, so the same red label certifies less centering: 37% of the pre-2025 PSA 10 fronts we measured sit past today's 55/45 limit (20% even after allowing two points of measurement error), versus 0.3% past the standard they were actually graded under.
Does back centering matter for PSA grades?
In practice, barely. PSA's published back tolerance is loose (75/25 for a PSA 10), and 0 of the 4,840 backs we measured exceeded it. The back standard is effectively unenforced because it is nearly impossible to fail.
How do I check a card's centering before buying or submitting?
Measure it instead of eyeballing it. The free Centering Check tool - the same measurement pipeline used for this study - reads the centering ratios from a photo or scan in seconds and compares them against every major grading standard.

The same measurement, on your card

Every number on this page came from a pipeline you can use for free. Upload a photo, get an objective centering read in seconds, and compare it against every grading company's standard - before you buy, and before you submit.