Left/Right vs Top/Bottom Centering: How Graders Combine Both Axes

Graders do not add, average or otherwise blend your left/right and top/bottom centering into one number. Each axis is measured on its own, and the worse of the two is the one compared against the grade tolerance. PSA's published standard says it outright: centering "is designated as the percent of difference at the most off-center part of the card." Beckett writes its front tolerances as "55/45 both ways," and TAG's Pristine 10 requires that front centering not exceed "a single measurement of 51/49."

So a card that is 50/50 left to right but 62/38 top to bottom is, for grading purposes, a 62/38 card. It does not get credit for the perfect axis. And a card that is 58/42 on both axes is a 58/42 card, which on paper actually measures better than the first one, even though it is off in two directions.

That single rule answers most of the "top/bottom vs left/right" questions collectors ask, but the details matter: how close each axis can get to the line, why top/bottom is the axis that usually fails, where grader discretion comes in, and how to measure both axes so you know what you are submitting. This article walks through all of it with real millimeter examples.

What the graders actually say about combining axes

None of the major companies publish a formula that merges the two axes. What they publish is a tolerance, plus wording that makes it clear the tolerance applies to each direction separately.

Two things follow from this. First, a card cannot "make up" for a bad top/bottom with a great left/right. Second, a card does not get penalized extra on the written scale for being off on both axes, as long as each axis is within tolerance. Where both-axes cards do get punished is in eye appeal, which we cover below.

Worked examples: one axis off vs both axes off

Here are four cards measured from the card edge to the outer printed border, with the standard 2.5 x 3.5 in (63.5 x 88.9 mm) card and a front border that is nominally 2.5 mm all round (so 5.0 mm total per axis). The ratio on each axis is the larger border divided by the sum of the two opposite borders.

Card A is the classic "one axis ruins it" case. Left/right is nearly perfect, but a 1.0 mm difference top to bottom lands exactly on PSA's 9 line, and that is the number PSA compares.

Card B is the "both axes off" case. A 0.8 mm difference on each axis gives 58/42 twice. Under the worst-axis rule, B measures better than A (58/42 beats 60/40), and both land in PSA 9 territory on paper. Neither meets the written 10 standard.

Card C shows how close you can get and still be inside: one axis at 54/46 and the other right at 55/45. Both are within the approximately-55/45 Gem Mint tolerance, so centering is not what would stop a 10 here.

Card D is the one collectors find most frustrating: a perfectly balanced left/right, but a 1.5 mm top/bottom difference that reads 65/35. That is the PSA 8 centering line. PSA's 5% leeway clause says a card that "meets all of the other requirements for PSA 9 and measures 65/35 off-center on the front... may be deemed to meet the PSA front centering standards for PSA 9 if the eye appeal of the card is good," so D could still land at 9 on eye appeal, but it is not a 10 candidate on centering.

The lesson for submissions: when you measure, write down both axes, circle the worse one, and compare only that one against the grade table. You can do this in a minute with the free card centering calculator, which reports left/right and top/bottom separately and maps the worse axis to each company's standard, or by hand following the method in how to measure card centering at home.

Why "both axes off" still looks worse than the math says

The written standard is indifferent to whether one or two axes are off, but your eyes are not, and PSA explicitly reserves room for that. Its standards page notes that "a small variance may be permitted on occasion based on the card's overall eye appeal," and its eye-appeal section says graders "reserve the right, based on the strength or weakness of the eye appeal, to make a judgment call" for cards "just within or just outside the printed centering standards."

A card that is off on one axis shifts the image straight up, down, left or right. The opposite pair of borders stays balanced, which gives the eye a reference line and makes the card read as "a little high" or "a little left." A card that is off on both axes shifts the image diagonally toward one corner: one corner of the border is fat, the diagonally opposite corner is thin, and nothing looks parallel. Collectors generally find that harder to ignore, and it is the kind of thing a grader's eye-appeal judgment can react to even when each axis is technically inside the line.

This is also why the forum claim that "60/40 can only be on one axis, not both" keeps circulating. It is not in any published standard. What collectors report is that a card with one borderline axis and everything else perfect sometimes gets the benefit of the doubt, while a card that is borderline in both directions rarely does. Treat that as an observation about discretion, not a rule. For the official line on 60/40 specifically, see is 60/40 centering a PSA 10?.

Why top/bottom is usually the axis that fails

Ask a room of collectors which axis cost them the 10 and most will say top/bottom. That is consistent with what collectors report across modern Pokémon and sports cards, and there are process reasons behind it.

Each cutting direction has its own error. Cards are printed on large sheets and cut in two passes, one direction at a time. Traditional descriptions of Topps-era cutting have the sheet cut into rows first and then into individual cards column by column. Any misalignment of the sheet on the cutter, any drift in the printed image relative to the sheet edge, and any dulling of the blade adds error to that pass. Left/right and top/bottom errors are independent, so it is normal for one to be fine and the other not. Why a given press run leans toward top/bottom error is set-specific, but the two axes are never guaranteed to track each other.

The percentage punishes thin reference borders. A 0.5 mm shift on a 2.5 mm border moves the ratio from 50/50 to 60/40. The same 0.5 mm shift on a 4 mm border is only 56/44. On many designs the top and bottom reference edges are thinner or more cramped than the sides (a text box or name bar sits close to the bottom edge; the top border is interrupted by a set logo or card name), so the same physical error turns into a worse number.

The long axis tilts more. A card is 88.9 mm tall and 63.5 mm wide. A tiny angular misalignment of the sheet produces a larger border difference along the longer dimension. That is also why a diamond-cut card (borders that taper from one end to the other) usually shows its worst taper top to bottom.

Practically, this means you should measure top/bottom with extra care. Measure the top and bottom borders at two points each (near the left edge and near the right edge) and use the worse reading; PSA's "most off-center part of the card" wording is broad enough to cover a tapering border, and you want to know the worst case before a grader does.

How the worst-axis rule plays out at each grader

The rule is the same everywhere; what changes is the number your worst axis has to beat. These are the official front tolerances for the top two grades at each company (verified against the published standards as of August 2026; see the PSA centering standards article and the grading standards guide for the full ladders).

Notice BGS: a 58/42 worst axis is an 8 centering subgrade territory there (60/40 both ways), while the same card is a PSA 9 on centering. Since Beckett's final grade is weighted toward the lowest subgrade, one weak axis drags the whole card. TAG goes the other direction in precision: it measures each axis by machine and reports front and back centering as separate subgrades on its DIG report, so there is no discretion to hope for. Both are worth knowing before you choose a grader for a one-axis-off card; the TAG centering explainer covers the machine side.

The back follows the same logic with looser numbers: worst back axis against 75/25 for a PSA 10, 55/45 for a BGS Pristine 10, 60/40 for a BGS 9.5, 75/25 for a CGC Gem Mint 10, and ~70/30 (sports) or ~65/35 (TCG) for a TAG Gem Mint 10.

How to measure both axes yourself (and report them properly)

1. Take the card out of the sleeve and photograph or scan it flat, straight down, with even light. Perspective tilt exaggerates one axis and hides the other, which is exactly the error you are trying to avoid.
2. Pick your reference edges. Measure from the physical card edge to the outer printed border line, the same line on all four sides. Do not measure to the artwork.
3. Measure left and right at the vertical middle of the card, then again near the top and bottom. Use the worst pair.
4. Measure top and bottom at the horizontal middle, then near the left and right edges. Use the worst pair.
5. Calculate each axis: larger border ÷ (larger + smaller) × 100. Write both ratios, larger number first, e.g. "L/R 53/47, T/B 59/41."
6. Circle the worse axis. That is your centering number for grading. Compare it, not the better axis and not an average, to the tolerance table.
7. Repeat on the back and compare the worse back axis to the back tolerance.

If you are using the free card centering calculator, it does steps 3 to 6 for you: you drag guide lines onto the border edges, it returns left/right and top/bottom separately, and it flags which axis is the limiting one for PSA, BGS, CGC, SGC and TAG. The app does the same from a photo. Either way, the number to pay attention to is the worse one.

When you describe a card in a listing or a forum post, give both axes and the back. "58/42 T/B, 52/48 L/R, back 60/40" tells a buyer everything; "about 55/45" tells them nothing about which direction, and it invites the assumption that you quoted the good axis.

Bottom line

• Graders test left/right and top/bottom separately and the worst axis is your centering number. PSA's own wording is "the most off-center part of the card"; BGS says "both ways"; TAG says "a single measurement."
• Nobody averages. 50/50 + 62/38 is a 62/38 card. 58/42 + 58/42 is a 58/42 card and measures better on paper.
• Being off on both axes does not cost extra under the written standard, but a diagonal shift looks worse, and PSA's eye-appeal clause gives graders room to react to that.
• Top/bottom is the axis collectors report failing most often: independent cutting error, thinner reference borders, and a longer dimension that amplifies tilt.
• Measure each axis at two points, use the worst reading, report both ratios plus the back, and compare only the worst one to the grade you are hoping for.

Frequently asked questions

Does PSA average left/right and top/bottom centering?+

No. PSA's published standard says centering is designated as the percent of difference at the most off-center part of the card, so the worse of the two axes is the number that gets compared to the grade tolerance. A card that is 50/50 left to right and 62/38 top to bottom is treated as a 62/38 card, not a 56/44 card.

Can a card be 60/40 on one axis and still get a PSA 10?+

On paper, no. PSA's current Gem Mint standard is approximately 55/45 on the front, and a 60/40 measurement on either axis exceeds that. PSA's page does allow a small variance at the grader's discretion based on eye appeal, and collectors report occasional 10s with one borderline axis, but you should plan on 60/40 being the PSA 9 line.

Is it worse to be off on both axes than badly off on one?+

For the written standard, only the worst single measurement counts, so 58/42 on both axes measures better than 52/48 plus 62/38. Visually, though, a card shifted on both axes drifts diagonally toward a corner, which many collectors find more noticeable than a clean shift along one axis, and PSA's eye-appeal clause lets graders react to that.

Why is top/bottom centering so often the axis that fails?+

Collectors report that top/bottom is the axis that most often costs a grade on modern cards. Each cutting direction carries its own error, and on many designs the top and bottom reference borders are thinner or harder to read than the sides, so the same absolute shift in millimeters turns into a larger percentage. The card is also taller than it is wide, so it can tilt slightly more along its long axis.

Does the worst-axis rule apply to the back of the card too?+

Yes. You measure the back's left/right and top/bottom the same way, take the worse axis, and compare it to the back tolerance, which is looser: 75/25 for PSA 10, 55/45 for a BGS Pristine 10, 60/40 for BGS 9.5, and roughly 70/30 sports or 65/35 TCG for a TAG Gem Mint 10. Front and back are judged separately.

What does 58/42 mean if it doesn't say which direction?+

By convention the larger percentage is written first, so 58/42 simply means the wider border holds 58 percent of the combined width and the narrower one holds 42 percent. Direction does not change the grade. Some collectors write 42/58 or add 'left-heavy' or 'high' to describe which way the image sits, but the grading comparison uses the same number either way.

Sources & further reading

• PSA Grading Standards (centering tolerances, OC qualifier text, eye appeal note)
• Beckett Grading Scale (BGS centering 'both ways' wording)
• TAG Grading Rubric (front and back centering per score band)
• TAG Grading help: How do I get a Pristine 10 grade? ('single measurement of 51/49')
• ACE Grading Scale (centering 'for any opposite pair')