Valdar / Free tools / Card Centering Calculator
Card centering calculator: measure the borders, get the ratio and the grade cap
Four border measurements in any unit, and you get the ratio a grader would write down — plus the highest grade PSA, BGS and SGC publish for that ratio. It is the one grading input you can settle yourself, before you pay anyone a fee.
Measure at the middle of each side, to the edge of the printed border — not to the edge of a photograph sitting inside the design. Leave a pair blank to judge one axis only.
Worse axis: left–right · 3.5 points off centre · wider on the left
| Company | Highest grade this centering allows | Its front tolerance |
|---|
Front tolerance only, from each company’s published scale read on 2026-09-11. Centering is a ceiling, never a grade: corners, edges and surface can bring a perfectly centred card down from here, and the reverse has its own, looser tolerance that is checked separately.
What the number means
Centering is the only part of a card grade that is a measurement rather than an opinion. Corners are judged under a loupe, surface under angled light, edges against a standard that lives in a grader’s hands — but centering is two divisions, and the answer is the same whoever does the arithmetic. That is why it is the right place to start when you are deciding whether to send a card away.
The ratio is written larger-side-first. If the left border measures 3.2 mm and the right measures 2.4 mm, the two together are 5.6 mm, the left is 57% of that and the card is 57/43 on the left–right axis. The same sum on the top and bottom gives the second axis. A grader judges the worse of the two, so a card that is beautifully centred side to side and 70/30 top to bottom is a 70/30 card.
Because it is a ratio, the unit does not matter. Millimetres from a steel rule, sixty-fourths of an inch, or pixels counted off a 600 dpi flatbed scan all produce the same answer, and the scan is the most accurate of the three for anyone without a machinist’s eye. What does matter is measuring the same thing on both sides: to the edge of the printed border, at the middle of the side, with the card flat.
The published tolerances, company by company
One of these moved recently and was not announced. PSA tightened the top of its ladder in early 2025: Gem Mint 10 went from 60/40 to 55/45 on the front, and every grade below it shifted with it, so 60/40 is now the Mint 9 line. A great deal of published advice — and a great many collectors’ memories — still carries the older numbers and is one step too generous at the top four grades. The table below is what the page says today.
These are the front and back tolerances each company states, with a link to the page it states them on. They are guidelines rather than a machine threshold — PSA words every one of them as “approximately”, and a card sitting on a boundary is decided by a person holding it — but they are the boundary of the argument, and a card well outside one is not going to be talked back inside it.
| Grade | Front | Back |
|---|---|---|
| Gem Mint 10 | 55/45 | 75/25 |
| Mint 9 | 60/40 | 90/10 |
| NM-MT 8 | 65/35 | 90/10 |
| NM 7 | 70/30 | 90/10 |
| EX-MT 6 | 80/20 | 90/10 |
| EX 5 | 85/15 | 90/10 |
| VG-EX 4 | 85/15 | 90/10 |
| VG 3 | 90/10 | 90/10 |
| Good 2 | 90/10 | 90/10 |
| Poor–Fair 1 | no stated limit | no stated limit |
| Grade | Front | Back |
|---|---|---|
| Pristine 10 | 50/50 | 60/40 |
| Gem Mint 9.5 | 55/45 | 60/40 |
| Mint 9 | 60/40 | 70/30 |
| NM-MT+ 8.5 | 65/35 | 80/20 |
| NM-MT 8 | 70/30 | 80/20 |
| NM 7 | 75/25 | 90/10 |
| EX-MT 6 | 80/20 | 90/10 |
| EX 5 | 85/15 | 90/10 |
| Grade | Front | Back |
|---|---|---|
| Pristine 10 | 50/50 | 60/40 |
| Gem 10 / 9.5 band | 55/45 | 70/30 |
| Mint 9 | 60/40 | 90/10 |
| NM-MT 8 | 70/30 | 90/10 |
Source: PSA grading standards, Beckett card grading scale, SGC grading scale. Companies revise these; the links go to the live pages rather than to a copy.
The back is where submissions actually die
Read the back column of those tables again. PSA allows 75/25 on the reverse of a Gem Mint 10 and 90/10 from Mint 9 downward; BGS and SGC are looser still. Those numbers look generous, and that is exactly why they catch people: a collector measures a front at 56/44, decides the card is a 10, and never turns it over. On cards printed with a thin border on one side of the back — which is most of the 1970s and a great deal of the 1990s — the reverse is the binding constraint, and nothing about the front will rescue it.
Measure both sides. It costs thirty seconds and it is the cheapest grading information available anywhere, because it is free and it is certain. The calculator above takes front measurements by default; run it twice, once for each face, and read the worse of the two answers as the card’s ceiling.
Why the same card measures differently on different sets
A border is not a border. On a 1989 Upper Deck the white frame is a wide, unambiguous block and measuring it is trivial. On a modern full-bleed Prizm there is no border at all in the ordinary sense, and centering is judged on the position of the design elements inside the trim. On a 1971 Topps, where the border is black, a millimetre of drift is visible from across the room and the same millimetre on a white border is not — which is why black-bordered sets have a reputation for being brutal to grade and why their gem populations are so small.
The physical size of the card matters too, for the same reason. Standard modern cards are 63.5 mm wide; a tobacco-era card is 36.5 mm. The same half-millimetre of drift is a much larger share of the narrower card’s border, so vintage measurements need a finer instrument than a school ruler and are the strongest case for measuring off a scan instead of by eye.
| Format | Inches | Millimetres | Why it matters here |
|---|---|---|---|
| Standard modern card | 2.5 × 3.5 | 63.5 × 88.9 | The size every card has been since 1957 Topps, and the size every sleeve, toploader and slab is built around. |
| Tobacco card (T206 era) | 1.4375 × 2.625 | 36.5 × 66.7 | Roughly 1 7/16 by 2 5/8 inches. Borders are narrow, so a millimetre of drift is a much larger share of the border than on a modern card. |
| 1952–1956 Topps | 2.625 × 3.75 | 66.7 × 95.3 | Larger than the modern standard, which is why these need their own holders and why the 1952 set does not fit a normal binder page. |
| Oversize / box-bottom | 3.5 × 5 | 88.9 × 127 | Box-bottom panels, Topps Super and similar. Most graders slab these in a larger holder at a higher fee tier. |
How to measure, in order
- Put the card on a flat surface in even light. A card held in the hand bows slightly, and a bowed card reads wider on the side nearest the eye. If it is in a sleeve, take it out or measure from a scan.
- Scan it at 600 dpi if you can. A flatbed scan removes parallax entirely and lets you count pixels, which is finer than any rule you own. Open the file, zoom to the border, and read the pixel coordinates.
- Measure to the printed border, at the middle of the side. Not to the edge of a photograph inside the design, and not at a corner, where a rounded cut distorts the reading.
- Do all four sides in the same unit. Mixing millimetres and sixteenths is the one arithmetic error this tool cannot detect, because both numbers are plausible.
- Check for a diamond cut before you trust the ratio. A card whose borders are wider at the top than at the bottom on one side and the reverse on the other was cut at an angle. That is a cutting fault, not centering, and it is graded as one.
- Turn the card over and repeat. The back has its own tolerance and it is frequently the one that binds.
What this changes about a submission
The practical value of measuring is that it removes one of the four grading dimensions from the guesswork before you spend anything. If the card measures 68/32 on the front, the published PSA scale caps it at 8 on centering alone, and the entire question of whether the corners are sharp enough for a 9 has become irrelevant. That is a decision you can make in a minute with a ruler rather than in ten weeks with a receipt.
It also tells you which company to send it to, which is a different question from which company is best. The three ladders above are close but they are not the same: a card at 56/44 is outside PSA’s Gem Mint band and inside BGS’s Gem Mint 9.5 band, so the same measurement fails one top grade and clears the other. Where the premium on a card sits entirely in one company’s top grade — and on modern cards it usually does — that difference is the whole submission.
Finally, it changes what you buy. A raw card offered as “sharp corners, looks mint” can be measured from the seller’s own photograph before you bid, provided the photograph is square to the card. A listing that shows the card at an angle, or crops the borders, is a listing that has decided not to let you measure it.
Bottom line
Centering is the grading input you can establish yourself, for free, in under a minute, and it caps the grade no matter how good the rest of the card is. Measure the border at the middle of each side in any consistent unit, take the worse of the two axes, and compare it with the published tolerance: PSA allows about 55/45 on the front for a Gem Mint 10 since it tightened the scale in early 2025, BGS wants 55/45 for a 9.5 and 50/50 for a Pristine 10, and every company’s reverse tolerance is looser but still binding. A card outside the band for the grade you were hoping for will not be argued back inside it, and knowing that before you pay the fee is worth more than any other single piece of grading advice.
Questions this tool gets asked
How do I measure card centering accurately?
Measure the border between the edge of the card and the edge of the printed image, at the middle of each of the four sides, not at a corner. Use the same unit on all four measurements — millimetres on a steel rule, or pixels on a flatbed scan at 600 dpi — because the result is a ratio and the unit cancels out. The most common mistake is measuring to the edge of a photograph inside a design rather than to the edge of the printed colour block that defines the border.
What centering do you need for a PSA 10?
PSA publishes approximately 55/45 on the front and 75/25 on the reverse for Gem Mint 10. That means the wider border may be up to about 55% of the combined width of the two opposing borders, on the worse of the two axes. A card at 58/42 is outside the published band for a 10 on centering alone, whatever its corners look like. Note that this tightened in early 2025 — it used to be 60/40, and a great deal of published guidance still carries the older number, which is one step too generous.
Does the back of the card count for centering?
Yes, and it is the half people forget. Every company states a separate and looser tolerance for the reverse, and every company applies it. A card that measures 55/45 on the front and 95/5 on the back is capped by the back, which is why a submission that looked certain comes back two grades lower with no visible damage anywhere.
Is a 60/40 card off-centre?
Since PSA tightened its scale in early 2025, 60/40 is the Mint 9 line rather than the Gem Mint 10 line — so on PSA it now caps the card at 9, and on BGS it sits at Mint 9 too. Before 2025 the same card was inside PSA’s Gem Mint band, which is why older advice and this answer disagree. On a card whose premium sits entirely in the top grade, that single step is the difference between two very different prices.
What is a diamond cut?
A card cut at an angle, so the border is wide at the top on one side and wide at the bottom on the other. It is a cutting fault rather than a centering fault and graders treat it as one, which means this calculator will report a plausible ratio for a card that has a different problem entirely. If the two measurements on the same axis disagree between the top and the bottom of the card, measure again at both ends before trusting the number.
Can centering alone stop a card getting a 10?
Yes, and it is the single most common reason a clean-looking card does not. Corners, edges and surface are judgements made under magnification; centering is a measurement anyone can take at the kitchen table before paying a fee. That asymmetry is the whole argument for measuring first.
Do older cards get judged more leniently on centering?
In practice the market judges them more leniently, and SGC in particular grades vintage against what the set actually produced rather than against a fixed ratio, because the cutting equipment of the 1950s did not hold registration. The published tolerances do not change by era; what changes is how much a 70/30 copy is penalised in price when 70/30 is the best anyone has seen of that card.
Does this tool send my measurements anywhere?
No. The whole calculation is a few divisions, and it runs in the page. There is no request to any server when you type, nothing is stored, and the tool works with the network switched off once the page has loaded.
The card is in your hand. Point the camera at it.
These calculators work on numbers you already have. Valdar gets you the numbers: it identifies the card from one photo, pulls what that exact card has actually sold for, and estimates the four grading sub-dimensions before you pay a submission fee. Free to try on iPhone and Android.