Tilted Print and Diamond Cut vs Off-Center: What Graders Call It
Diamond cut vs off-center: a shift and a rotation are different faults. The geometry, why you measure each border twice, and BGS, CGC, SGC and PSA's rules.

Two cards can both be described as "off-center" and have completely different faults. On the first, the print sits to one side of the cut: the left border is 2.9 mm and the right border is 2.1 mm, and each of them is that width all the way along. That is a parallel shift, the classic off-center card, and it is what a centering ratio was designed to describe.
On the second, the print is rotated by a fraction of a degree relative to the cut. Now the left border is 2.29 mm at the top of the card and 2.71 mm at the bottom, and the right border does the mirror image. The card looks subtly like a rhombus in the holder. Collectors call it a diamond cut or a tilted print, Beckett writes allowances for it into its published scale, and — this is the part that matters — it can average out to a perfect 50/50 while still being visibly crooked.
That is why the single most important measuring habit in this whole subject is taking two readings per side instead of one. A single reading per side turns a rotation into a fake off-center number; averaging two readings without recording the spread hides the rotation entirely. You need both the average and the spread.
Shift versus rotation, geometrically
Centering is a property of where the cut fell relative to the printed design, and there are only two ways for a rectangle to be misplaced relative to another rectangle: it can be translated and it can be rotated.
- Translation (off-center). Every border on an axis differs from its opposite by a constant amount. Left(top) = Left(bottom). One ratio describes the card completely.
- Rotation (diamond cut, tilted print). Each border tapers linearly along its length, and the two opposite borders taper in opposite directions. Their sum is constant everywhere: at any height, left + right still equals the design's combined border width. So the ratio depends entirely on where you put the ruler.
Real cards have some of both, and they have separate causes. A shift comes from the sheet or the blade being in the wrong place. A rotation comes from the sheet being loaded out of square, the image being printed at a slight angle on the sheet, or a cutting angle drifting over a run — the mechanics are covered in why are trading cards off-center.
There is also a third, rarer fault worth separating out: a non-perpendicular cut, where the blades themselves are not at right angles, so the cardboard is a parallelogram rather than a rectangle. In that case the card's own corners are not 90°.
The test that separates them: measure the card's two diagonals. A standard card is 63.5 × 88.9 mm, so each diagonal should be √(63.5² + 88.9²) ≈ 109.25 mm. If both diagonals match, the cardboard is a genuine rectangle and any taper you are seeing is the print being rotated. If they differ, the cut itself was out of square.
How much a fraction of a degree costs you
This is where the intuition fails. Assume a card with 2.5 mm nominal side borders (5.0 mm combined), a print rotated by angle θ about the card's center, and no shift at all — so the card averages exactly 50/50.
The taper along a side border read between two points 15 mm in from each corner (a span of about 59 mm on the card's 88.9 mm height) is 59 × tan θ. The ratio at any height y is (2.5 + Δ) / 5.0, where Δ is the offset at that height.
| Rotation | Taper along a side border (over 59 mm) | Ratio at the card's midpoint | Ratio 15 mm from a corner | Ratio at the corner itself |
|---|---|---|---|---|
| 0.1° | 0.10 mm | 50/50 | 51/49 | 51.6/48.4 |
| 0.25° | 0.26 mm | 50/50 | 52.6/47.4 | 53.9/46.1 |
| 0.5° | 0.52 mm | 50/50 | 55.1/44.9 | 57.8/42.2 |
| 1.0° | 1.03 mm | 50/50 | 60.3/39.7 | 65.5/34.5 |
| 2.0° | 2.06 mm | 50/50 | 70.6/29.4 | 81/19 |
Half a degree is enough to reach PSA's published 10 line at a point a grader will actually look at, on a card whose average centering is flawless. One degree reaches the 9 line. And this matters legally, so to speak, because of how PSA words its standard: centering "is designated as the percent of difference at the most off-center part of the card." A grader reading the worst point on a rotated card is reading the corner, not the middle.
Thin borders are punished harder, since the same physical taper is a bigger fraction of a smaller number. On a 2.0 mm sports white border (4.0 mm combined), the same 0.5° rotation reads 56.4/43.6 at the 15 mm point and 59.7/40.3 at the corner.
One more geometric consequence: the taper is largest on the left and right borders, because those are read along the card's long 88.9 mm dimension. Read along the 63.5 mm width instead — the top and bottom borders, sampled 15 mm in from each corner, a span of about 33.5 mm — the same 0.5° rotation produces only 0.29 mm of taper. If you only have time to check one axis for skew, check the sides.
Why one reading per side gives you the wrong story
Take a card with a 0.4° rotation and no shift. Over the 59 mm span the taper is 0.41 mm, so the left border runs from about 2.29 mm at the top to 2.71 mm at the bottom, and the right border mirrors it.
Now watch what three different measuring habits report about the same card:
- One reading per side, both taken near the top. Left 2.29, right 2.71 → 2.71 ÷ 5.00 = 54.1%, reported as 54/46. You conclude the card is mildly off-center to the right.
- One reading per side, both taken near the bottom. Left 2.71, right 2.29 → 54/46 the other way. You conclude it is mildly off-center to the left. Same card, opposite diagnosis.
- Two readings per side, averaged. Left (2.29 + 2.71) ÷ 2 = 2.50, right 2.50 → 50/50. You conclude the card is perfectly centered, and you are right about the average and wrong about the card.
None of those three is a useful report on its own. The useful report is: average 50/50, spread 0.42 mm, skewed. Record both.
The practical threshold that shows up across measuring guides is about 0.2 mm: if your two readings on the same side differ by more than that, the card is skewed rather than merely off-center — provided your image is square. A tilted photo produces the same signature, so on a phone shot always rule out perspective before you diagnose the card, which is one reason a flatbed scan is the cleaner instrument for this particular question.
What each grading company says
Beckett is the only major company that writes diamond-cut allowances into its published centering text, grade by grade.
| Grade | Beckett's published diamond cut wording |
|---|---|
| BGS 8 and above | no allowance listed |
| BGS 7 Near Mint | "Very slight diamond cutting is allowed." |
| BGS 6 and 5 | "Slight diamond cutting is allowed." |
| BGS 4 and 3 | "Moderate diamond cutting is allowed." |
| BGS 2 Good | "Noticeable diamond cutting is allowed." |
| BGS 1 Poor | "Heavy diamond cutting is allowed." |
The absence at 8 and above is the informative part. On the published scale, a visible diamond cut caps the BGS centering subgrade at 7, and because Beckett's overall grade is not an average and "the lowest grade is the most heavily weighted," a 7 centering subgrade pulls the whole card down hard. Beckett's own worked example of a 10/6/10/10 card producing a final 7 shows how steeply. The full ladder is in BGS centering standards.
The other companies say much less, and where they do speak, it is at the bottom of the scale:
- SGC notes in its grade 7 description that a "card may exhibit a slightly skewed (diamond) cut." Nothing at 8 and above.
- CGC mentions diamond cutting once, at 4.5, where "an extremely noticeable diamond cut is allowed."
- PSA publishes no per-grade diamond cut allowance at all. It handles atypical cuts through two other mechanisms: the MC (Miscut) qualifier, defined for "any and all cards that exhibit an atypical cut for the issue, which may result in portions of the subject card being cut off or more than one card being visible," and the N8 Miscut no-grade for a cut abnormal enough that PSA declines to grade, for which PSA states you are not charged the grading fee. Between "normal" and "atypical" sits PSA's eye-appeal clause, which reserves a judgment call to the grader. How the qualifiers work in practice is covered in off-center vs miscut: OC and MC qualifiers.
- ACE publishes no diamond cut rule either. It reports a percentage for left, right, top and bottom centering on both faces, from measurements it describes as accurate to a thousandth of a millimetre, and applies its published thresholds to whatever those percentages say.
Read together, the pattern is consistent: no major grader makes room for a visible skew above the Near Mint band. Nobody publishes a "slight diamond cutting allowed" line at 9 or 10, and where the companies do publish one it starts at 7 or lower.
So can a tilted card be a 10?
It depends entirely on whether the tilt is visible, and "visible" has a rough numeric meaning you can test for.
- Under about 0.1° of rotation (a taper of roughly 0.1 mm along a side border): your two readings per side will agree within a tenth of a millimeter, the worst point reads about 51/49, and nothing about the skew will be visible or measurable at grading resolution. This is not what is blocking your 10.
- Around 0.25° (taper about 0.26 mm): detectable in a careful measurement, worst point around 53/47 on a 2.5 mm border. Probably still fine on the ratio; whether a grader reads it as "skew" is a judgment call and PSA's eye-appeal language is exactly the discretion that decides it.
- At 0.5° and above (taper 0.5 mm or more): the worst point crosses the top-grade ratio line on its own, the taper is visible against the slab frame, and on Beckett's published scale it is a centering subgrade problem regardless of the average.
The blunt version: a card with an average of 50/50 and a half-degree rotation is not a better submission than a card sitting at a clean, flat 55/45. The flat card presents one number that lands inside a published line. The rotated card presents a number that changes depending on where you look and a shape a grader can see from across the desk.
How to measure and report it properly
- Take the card out of the sleeve and lay it flat on a matte contrasting surface. Scan it at 600 dpi or higher, or shoot straight down; perspective mimics skew perfectly, so the image has to be square before any of this is meaningful.
- Measure each border at two points, about 15 mm in from each corner, from the cardboard edge to the outer printed border.
- Record four numbers per axis, not two: left-top, left-bottom, right-top, right-bottom.
- Compute the average ratio (average left ÷ (average left + average right)) and the spread on each side (the difference between the two readings).
- If any spread exceeds about 0.2 mm, note the card as skewed and also compute the worst-point ratio, using the two readings taken at the same end of the card. That worst-point number is what PSA's "most off-center part of the card" language points at.
- Check the diagonals if the taper is large, to see whether the cardboard itself is out of square.
- Repeat on the back. Front and back are printed in separate passes, so a card can be skewed on one face and square on the other.
- When you sell or ask for an opinion, quote all three figures: average ratio, worst-point ratio, and spread. "52/48 average, 57/43 at the top, 0.4 mm taper on the left" is a complete description. "52/48" alone is not.
The free card centering calculator handles the ratio arithmetic in the browser with manual guide lines, which is what you want here because you will be running the numbers twice per axis — once at each end of the card — rather than accepting a single automatic reading. The full home-measurement procedure across rulers, calipers, scans and apps is in how to measure card centering at home, and every company's published tolerances sit together in the grading standards guide.
Bottom line
- Off-center is a translation and diamond cut is a rotation; the ratio formula describes the first properly and can completely misdescribe the second.
- A rotation's opposite borders always sum to the same total, so the ratio depends on where you measure — the average can be a flawless 50/50 on a visibly crooked card.
- Half a degree of rotation gives about 0.5 mm of taper along a side border, which reads 55/45 fifteen millimeters from the corner and 57.8/42.2 at the corner itself on a 2.5 mm border; one degree reads 60/40 and 65.5/34.5.
- Beckett is the only company publishing per-grade allowances: very slight at BGS 7 down to heavy at BGS 1, with nothing listed at 8 and above, so a visible skew caps the centering subgrade at 7. SGC allows a slightly skewed cut at 7, CGC mentions it at 4.5, and PSA publishes none, using the MC qualifier and N8 Miscut no-grade instead.
- Detect it by measuring each border at two points about 15 mm from each corner; a spread over roughly 0.2 mm means skew, provided the image itself is square.
- Check the diagonals (about 109.25 mm on a 63.5 × 88.9 mm card) to tell a rotated print from a cut that was not perpendicular, and report average ratio, worst-point ratio and spread together.
Frequently asked questions
What is the difference between a diamond cut and an off-center card?
An off-center card is a parallel shift: the print sits to one side of the cut, so one border is wider than the opposite one, but each border keeps the same width along its whole length. A diamond cut is a rotation: the print is turned by a fraction of a degree relative to the cut, so each border tapers, wide at one end and narrow at the other. A card can have either fault, both at once, or a rotation that averages out to a perfect ratio while still looking crooked.
Does BGS allow diamond cuts?
Only below the Mint level. Beckett's published scale allows very slight diamond cutting at BGS 7, slight diamond cutting at 6 and 5, moderate diamond cutting at 4 and 3, noticeable diamond cutting at 2 and heavy diamond cutting at 1. No diamond cut allowance is listed at BGS 8 or above, so on the published scale a visible skew caps the centering subgrade at 7.
Can a diamond cut card still get a 10?
A visible one, no. Beckett lists no diamond cut allowance at 8 and above, CGC only mentions diamond cutting at 4.5, and SGC only at 7, so none of them make room for a visible skew at the top of the scale. PSA publishes no per-grade diamond cut allowance at all and instead treats an atypical cut through its MC qualifier and its N8 Miscut no-grade, with eye appeal at the grader's discretion. A skew small enough that both readings on a side agree within a tenth of a millimeter is invisible and is not the thing blocking your 10.
How do I detect a diamond cut when measuring?
Measure each border at two points, roughly 15 mm in from each corner, and record both readings rather than only the average. On a pure parallel shift the two readings match. If they differ by more than about 0.2 mm the print is rotated relative to the cut. Averaging is what hides it: a rotation about the card's center can average to a perfect 50/50 while the card is visibly crooked at both ends.
Is a diamond cut the same as a non-perpendicular cut?
No, and the test separates them. If the print is rotated but the cut is square, the card is still a true rectangle and only the borders taper. If the blades were not perpendicular, the card itself is a parallelogram, so its corners are not right angles. Measure both diagonals: on a standard 63.5 by 88.9 mm card each diagonal should be about 109.25 mm. Equal diagonals mean the cardboard is a rectangle and the fault is in the print alignment.
How much does a small rotation change my centering ratio?
More than you would guess, because graders read the worst point. On a card with 2.5 mm borders, a rotation of only half a degree produces about 0.5 mm of taper along a side border, so 15 mm in from a corner the ratio already reads 55/45 even though the card averages 50/50. At one degree the same point reads 60/40 and the corner itself reads 65/35. Thinner borders are punished harder.


