Magic: The Gathering Card Centering: Borders, Tolerances and Grading
MTG card centering explained: where to measure on the M15 black border, old-border and borderless frames, how to see a black cut edge, and every grader's tolerance.

Magic: The Gathering cards are graded against the same published centering tolerances as every other trading card. For a PSA 10 that means approximately 55/45 on the front and 75/25 on the back, with the worse of the two axes deciding. A PSA 9 allows approximately 60/40 and 90/10. BGS wants 55/45 both ways on the front for a 9.5 and 50/50 for a Pristine 10. CGC's Gem Mint 10 matches PSA at approximately 55/45 and 75/25. There is no Magic-specific rule anywhere.
What is Magic-specific is the black border. On a modern M15-frame card the measurable border is a black band running from the cut edge inward to the card's interior frame, and black-on-black is the hardest edge in the hobby to photograph. Many of the bad MTG measurements posted in collector groups are not measurement mistakes at all; they are cases where the camera could not tell where the card stopped and the desk began.
This guide covers the frame anatomy for every Magic era, which lines to measure, how to make a black cut edge visible, the tolerance table for all five major graders, and a worked example with real arithmetic.
Anatomy of a Magic card for centering purposes
Magic cards are standard size, 2.5 x 3.5 in (63.5 x 88.9 mm), so every grader's tolerance applies without adjustment.
The modern M15 frame (introduced with Magic 2015 in 2014 and still the default) is the frame most submissions use. A title bar with the card name and mana cost sits at the top, an art window below it, a type line with the set symbol under the art, then a text box for rules and flavor text and a line of collector information and artist credit along the bottom. All of that sits inside a card frame, and around the whole thing runs a plain black border out to the cut edge. That outer black band is your measurement target.
The 8th Edition through M14 frame (2003 to 2014) has the same geography: black border, then an interior frame with a title bar, art window, type line and text box. The border-to-frame boundary is just as crisp, so the method does not change.
Old-border cards (1993 to 2003) use a beveled tan or brown interior frame with a rounded art window, and came in both black-bordered and white-bordered printings. White-bordered cards are, for measurement purposes, the easiest Magic cards in existence: a white band against a colored interior is unmistakable, and you can measure them on a dark mat like any Pokémon card. They are also the least forgiving to the eye, because a high-contrast border makes a small shift obvious, which is exactly the situation PSA describes when it says graders may make a judgment call based on the strength or weakness of a card's eye appeal.
Borderless, showcase, extended-art and full-art cards. Showcase frames usually keep a black border and change only the interior, so nothing about the method changes. Extended-art treatments push the artwork across the side borders, leaving a usable black band only at the top and bottom. Borderless cards remove the band entirely and run art to all four cut edges. Those last two need the fallback method below. Foils of any treatment are printed and cut on the same sheets as non-foils, so what changes is glare, not geometry.
The back. Every deck-legal Magic card shares one brown back with a symmetrical central design inside an inner frame line. Because that design is identical on every card, the back is the most reliable surface on a Magic card for a clean four-sided measurement, and it carries its own published tolerance at most graders. PSA, BGS, CGC and TAG all state a back figure; SGC publishes none at any grade.
Where to measure on a modern black-bordered card
Measure the black band, from the physical cut edge of the card to the line where black meets the interior frame.
- Left/right: cut edge to the frame line, on both sides, at the same height on each side.
- Top/bottom: cut edge to the top of the title bar's frame line, and cut edge to the bottom of the interior frame beneath the collector line.
Two cautions specific to this frame. First, do not measure to the art window: the art, title bar and text box are printed in a fixed relationship to the interior frame, so they add nothing the frame line has not already told you, and the art window is not vertically centered in the card anyway. Second, the top and bottom black bands are not designed to be identical widths on every frame generation, so measure like to like, then sanity-check against the back and against a known well-centered copy of the same printing.
The formula is the standard one: larger border divided by the two borders combined, written larger first. Take a card whose combined left-plus-right black border is 6.0 mm:
- 55/45 means 3.30 mm and 2.70 mm, a difference of 0.60 mm.
- 60/40 means 3.60 mm and 2.40 mm, a difference of 1.20 mm.
Six tenths of a millimeter separates the published PSA 10 line from the published PSA 9 line on a typical Magic border. That is smaller than the width of a ruler's printed mark, which is why a ruler is the wrong tool here. More ratio-to-millimeter conversions are in what centering ratios mean in millimeters.
The black-border problem, and how to solve it
Every other centering guide tells you to put the card on a dark matte surface. For Magic that advice is backwards and it will cost you several points of measured ratio.
- Use a white or light background. A black border against black felt has no detectable edge; against white paper it has a hard one. This one change fixes most bad MTG readings.
- Scan rather than photograph. A flatbed with the lid closed surrounds the card with an even white field, has zero perspective error, and has a known scale: at 600 dpi one millimeter is roughly 24 pixels, so a pixel measurement resolves well under a tenth of a millimeter.
- If you backlight, do it gently. Holding a card against a window makes the outline pop, but strong backlight bleeds around the edge and makes the card read slightly smaller than it is. Backlighting is a good way to see a problem and a poor way to measure one.
- Kill the glare on foils. Two diffuse sources at low angles from opposite sides, or open shade, or a scanner. A single overhead lamp puts a hot spot on one side of the frame and erases the line you are measuring to.
- Take the card out of the sleeve. A penny sleeve adds a bright seam exactly where the black cut edge is, and on a black-bordered card that seam is often the brightest thing in the photograph.
- Measure each border at two points. A skewed cut reads as two different numbers along the same side. One number hides it.
Borderless, extended-art and full-art Magic cards
No grading company publishes a method for cards without a border, so what follows is collector practice rather than official policy.
- Use fixed printed elements, not the artwork. On a borderless card the title bar and the text box sit consistently across every copy of that printing. Measure from each side's cut edge to the corresponding end of the text box for a left/right reading. The artwork is not designed to be symmetrical and is useless as a reference.
- On extended-art cards, you still have a top and bottom band. Measure that axis normally and treat left/right by the element method.
- Compare against a known good copy, and measure the back. High-resolution images of the same card in a top-grade slab expose a shift that is invisible in isolation, and the brown back is where an objective four-sided measurement still lives.
The reference-point logic, including the reconstruction method for cards with only two or three usable borders, is worked through in how to measure centering on full-art and borderless cards.
Magic centering tolerances at every grader
All figures are from the companies' official standards pages, verified August 2026. Magic is a TCG, so where TAG splits sports and TCG the TCG figure is shown. CGC publishes numeric ratios below Gem Mint 10 only for sports and non-sports cards and states none for TCG at 9 and below.
| Grader / grade | Front | Back |
|---|---|---|
| PSA 10 Gem Mint | ~55/45 | 75/25 |
| PSA 9 Mint | ~60/40 | 90/10 |
| PSA 8 NM-MT | ~65/35 | 90/10 |
| BGS 10 Pristine (Gold or Black Label) | 50/50 | 55/45 |
| BGS 9.5 Gem Mint | 55/45 both ways | 60/40 |
| BGS 9 Mint | 55/45 both ways | 70/30 |
| BGS 8 NM-MT | 60/40 both ways | 80/20 |
| CGC Pristine 10 | 50/50 | not stated |
| CGC Gem Mint 10 | ~55/45 | 75/25 |
| SGC 10 Pristine / 10 Gem | 50/50 / 55/45 | not stated |
| TAG 10 Pristine (990-1000) | ~51/49 | ~52/48 (TCG) |
| TAG 10 Gem Mint (950-989) | ~55/45 | ~65/35 (TCG) |
| TAG 9 Mint (900-949) | ~60/40 | ~75/25 (TCG) |
Three things follow for a Magic submitter. PSA and CGC are the most forgiving on the back at the top grade, allowing 75/25. BGS is the strictest at the top on both faces, and its 9 keeps the front at 55/45 while only loosening the back, so BGS 9 front centering is as demanding as BGS 9.5. And BGS figures describe the centering subgrade, not the final number, because Beckett states that the final grade is not a simple average and that the lowest subgrade is the most heavily weighted. The full ladder is in the centering tolerance by grade table, and the same thresholds are collected on our grading standards guide.
On which company to use, there is no published data on grading volume by game, so treat anything you read about "the MTG grader" as convention rather than fact. Collectors do report that BGS subgrades are popular for high-value modern Magic because centering is broken out and visible on the label. Pick on the published tolerance your card actually meets, not on reputation.
Worked example: reading a real measurement
An M15-frame mythic, out of the sleeve, scanned at 600 dpi on a flatbed.
- Front left border 3.4 mm, right border 2.6 mm. Total 6.0 mm. 3.4 / 6.0 = 56.7%, so 57/43 left to right.
- Front top border 3.1 mm, bottom border 3.3 mm. Total 6.4 mm. 3.3 / 6.4 = 51.6%, so 52/48 top to bottom.
- Back top 3.6 mm, bottom 2.4 mm. Total 6.0 mm. 3.6 / 6.0 = 60%, so 60/40 on the back.
The worse front axis is 57/43 and the worse back axis is 60/40. Nobody averages the two axes; PSA states that centering is designated as the percent of difference at the most off-center part of the card, and BGS writes its figures as applying "both ways." That worst-axis rule is the single most misunderstood part of centering, and it gets its own article: both axes off-center and the worst-axis rule.
Placing 57/43 front and 60/40 back on the table:
- PSA and CGC: two points past the approximate 55/45 line, so a 9 on paper. PSA's published 5% front leeway applies to cards grading 7 or better and its standards page allows a small variance at the grader's discretion based on eye appeal, so a 10 is possible but is not the published outcome. The 60/40 back is comfortably inside 75/25.
- BGS: 57/43 fails 55/45 both ways, the line for both 9.5 and 9; 60/40 both ways is the BGS 8 line. The back at 60/40 passes 9.5's back figure exactly. Centering subgrade lands around 8.
- SGC and TAG: past the 55/45 Gem line, so 60/40 at SGC 9 and ~60/40 at TAG Mint 9. The 60/40 back is inside TAG's ~75/25 TCG Mint 9 figure.
One card, several ceilings, none of them depending on how good the corners are. That is the point of measuring before you pay a fee.
Miscuts, crimps and skewed cuts
Magic produces a few defects that people file under "centering" and that graders do not.
Miscut. PSA defines its MC qualifier as covering any card that exhibits an atypical cut for the issue, which may result in portions of the card being cut off or more than one card being visible, and it lists an N8 Miscut no-grade for cuts abnormal enough that PSA declines to grade at all, with no grading fee charged. A Magic card showing a strip of its sheet neighbor along one edge is a miscut, not a 95/5 card, and the ratio stops being the useful description. What each label does to value is in off-center vs miscut: OC and MC qualifiers.
Crimps. A crimp is a pinch or crease pressed into the card by the pack-sealing machine. It is an edge and surface defect with nothing to do with where the sheet was cut, and no amount of good centering rescues it.
Skewed or diamond cuts. When a card is cut slightly off square, opposite borders taper: a side measuring 2.6 mm at the top can measure 3.1 mm at the bottom. Beckett allows very slight diamond cutting at BGS 7, slight at 6 and 5, moderate at 4 and 3, noticeable at 2, and heavy at 1; SGC's 7 description notes a card may exhibit a slightly skewed cut; CGC states an extremely noticeable diamond cut is allowed at 4.5. This is why you take two readings per side. A single reading averages a good end with a bad end, and the grader sees the bad end.
A repeatable routine before you submit
Card out of the sleeve, on white paper or straight onto a flatbed. Scan at 600 dpi, or shoot straight down with the camera parallel to and centered over the card. Take two readings per side: front left/right cut edge to the interior frame line, front top/bottom frame line to frame line, then all four sides of the brown back. Write down the worst front axis and the worst back axis, find the highest row in the table where both pass, and treat that row as your centering ceiling at that company. If either number sits within two or three points of a line, re-measure before deciding; that band is where "approximately" and grader discretion decide the outcome.
The free card centering calculator does the measurement and the mapping from a photo in the browser, with guide lines you drag onto the frame line yourself, which is what you want on a black border where automatic edge detection has the least contrast to work with. The general hand method, including the two-readings-per-side technique, is in how to measure card centering. If you also collect other games, the same logic applied to a different frame is in the Lorcana centering guide.
Bottom line
- Magic is graded on the standard tolerances: PSA 10 is ~55/45 front and 75/25 back, BGS 9.5 is 55/45 both ways and 60/40 back, CGC Gem Mint 10 matches PSA, TAG Gem Mint is ~55/45 front and ~65/35 back for TCG. No grader publishes a Magic-specific rule.
- Measure the black band from the cut edge to the interior frame line, never to the art window, with two readings per side to catch a skewed cut.
- Reverse the usual advice on background: black-bordered cards need a white surround, and a flatbed scan is the most reliable home method.
- Borderless and extended-art treatments have no band; measure to fixed printed elements and lean on the symmetrical brown back.
- On a typical 6.0 mm combined border, the PSA 10 and PSA 9 lines are 0.60 mm apart. A ruler cannot resolve that; a scan or a caliper can.
- Crimps and diamond cuts are not centering faults. Miscuts leave the ratio behind entirely and are labeled MC or returned as an N8 no-grade.
Frequently asked questions
What centering does a Magic card need for a PSA 10?
The same as any other card. PSA's published Gem Mint 10 standard is centering within approximately 55/45 on the front and 75/25 on the reverse, with the worse of the left/right and top/bottom axes governing. PSA 9 allows approximately 60/40 front and 90/10 back. No grading company publishes a Magic-specific centering tolerance.
Where do you measure centering on an MTG card?
On a modern M15-frame card, measure the black border itself: from the cut edge of the card to the line where the black stops and the interior card frame begins, on all four sides. Left and right form one ratio, top and bottom the other. Do not measure to the art window, because the art is printed in a fixed relationship to the frame and tells you nothing extra about where the cut landed.
How do you see the edge of a black-bordered card to measure it?
Put the card on a white or light background instead of the dark mat you would use for a Pokémon card. Black border against a dark surface is the single most common cause of a wrong MTG measurement. A flatbed scan with the lid closed gives an even white surround and no perspective error, which is why collectors reach for a scanner on black-bordered cards.
Are old-border Magic cards harder to grade for centering?
Collectors widely report that pre-2003 Magic printings are less consistently cut than modern ones, which matches the general pattern across the hobby, but no grading company publishes set-level data and we will not invent any. What is objectively true is that white-bordered printings show the same ratio far more obviously than black-bordered ones, and that PSA explicitly allows graders a small variance based on eye appeal.
How do you measure centering on a borderless or showcase Magic card?
There is no black band to measure, so no grading company publishes a method. Collectors measure from the cut edge to a fixed printed element that runs parallel to it, such as the ends of the text box or the title bar, compare against a known well-centered copy of the same card, and lean on the brown Magic back, which is symmetrical and measurable on all four sides.
Does a crimp or a diamond cut count as bad centering?
No. A crimp is packaging damage from the pack-sealing machine and is a surface and edge defect. A diamond cut is a skewed cut, so opposite borders taper along their length rather than being uniformly unequal. Centering describes a parallel shift. Measure each border at two points along its length; if the two readings differ noticeably, you have a skew, and a single ratio understates the problem.


