E-Reader and Partial-Border Cards: Centering With Missing Borders
How to measure centering when a dot-code strip or a full-bleed edge removes a border: ACE's three-border reconstruction, its error bars, and the fallbacks.

The centering formula assumes something that most cards give you and some cards do not: two opposite borders of the same intended width. Take that away — a dot-code strip running along an edge, art bleeding off one side, a design element crossing the border — and the arithmetic has nothing to divide.
The Pokémon-e sets of 2002 and 2003 (Expedition, Aquapolis, Skyridge) are the case collectors run into most, because the e-Reader dot-code strip occupies edge space that would otherwise be plain border. But the problem is much broader than one era: modern full-bleed alt arts, cards where a nameplate or wing runs off one edge, and any design whose top and bottom borders were never drawn the same width as its sides all land in the same place.
There is exactly one published method for it. ACE Grading describes reconstructing a missing border from the pair you can see. No other grading company publishes any rule for borderless or partial-border measurement at all. Everything else in this article is collector practice, and it is labelled as such.
First, count the borders you can actually measure
Before you touch a ruler, look at the card and classify it. Do not assume the strip or the bleed sits where it does on some other card in the set — check yours.
| What you can see | Can you form a ratio? | What to do |
|---|---|---|
| Four borders, same design element on each opposite pair | Both axes | Measure normally |
| Three borders (one edge occupied or full-bleed) | One axis directly, one by reconstruction | ACE's three-border method, below |
| Two borders on the same axis | That axis only | Measure it, report the other axis as not measurable |
| Two borders on adjacent edges (an L shape) | Neither axis | Reference copy, the back, or millimeters only |
| One or no measurable borders | Neither | Reference copy, the back, or millimeters only |
Two traps to avoid while classifying:
A strip is a printed element, so its outer edge is a perfectly valid reference line. It runs parallel to the cut and sits at a fixed place in the design, exactly like a border does. What it is not is the same nominal width as the border on the opposite edge. So you can measure the gap from the cardboard edge to the strip with complete confidence — you just cannot put that gap and the opposite border into a ratio, because the ratio formula assumes the two were designed to be equal.
Never reconstruct from the artwork. The art on these designs is composed, not centered. It carries no information about where the cut fell.
ACE's three-border reconstruction, worked
ACE Grading says its centering score is calculated using measurements accurate to 1/1000th of a millimetre, and that it reports a percentage for left, right, top and bottom centering on both the front and the back. Its centering blog states that on cards where only three borders are visible, it calculates "the average border size with the known pair (left and right)" and uses that to "figure out what the size of the border should be."
In other words: the known pair tells you the design's nominal border width, because a cut shifts one border out by exactly as much as it shifts the other in. Their average is invariant to the cut. You then assume the missing border was drawn to that same nominal width.
Worked, with a card whose bottom edge is occupied by a strip:
- Measure the known pair. Left border 2.62 mm, right border 2.18 mm.
- That axis's ratio, directly. 2.62 ÷ (2.62 + 2.18) = 2.62 ÷ 4.80 = 54.6% → 55/45 left to right. The physical shift is (2.62 − 2.18) ÷ 2 = 0.22 mm.
- Derive the nominal. (2.62 + 2.18) ÷ 2 = 2.40 mm. This is what an uncut, perfectly placed border on this design measures.
- Measure the visible border on the broken axis. Top border 2.86 mm.
- Infer the missing one. Treat the hidden bottom as 2.40 mm.
- Compute. 2.86 ÷ (2.86 + 2.40) = 2.86 ÷ 5.26 = 54.4% → 54/46 top to bottom. Equivalently: the top border is 2.86 − 2.40 = 0.46 mm wider than nominal, so the cut fell 0.46 mm low.
- Take the worst axis. 55/45 left to right is the card's centering, since graders read the worse of the two.
Do the same on the back, which on most Pokémon cards has a complete symmetric border and needs no reconstruction at all.
The assumption, and exactly what it costs you
Step 5 is the whole method and it is an assumption: that the design intended the same border width on all four sides. That is true of most standard frames and untrue of plenty of others, including many layouts where a nameplate or a set strip makes the top and bottom deliberately different from the sides.
Here is what being wrong costs, holding the visible top border at 2.86 mm and varying only what you assume about the hidden bottom:
| Assumed nominal for the hidden border | Resulting top/bottom ratio | Error vs the 2.40 mm assumption |
|---|---|---|
| 2.20 mm | 56.5/43.5 | +2.1 points |
| 2.30 mm | 55.4/44.6 | +1.0 point |
| 2.40 mm (ACE's method) | 54.4/45.6 | — |
| 2.50 mm | 53.4/46.6 | −1.0 point |
| 2.60 mm | 52.4/47.6 | −2.0 points |
| 3.00 mm | 48.8/51.2 | the wide side flips to the bottom |
So the rule of thumb: every 0.2 mm you are wrong about the nominal costs you about 2 ratio points, and a large enough error reverses which side is wider. That is why the reconstruction is only trustworthy when the visible borders on both axes are clearly the same design element at the same nominal width — the same colored frame, the same thickness, the same printed line. Where the design is obviously asymmetric, the method's assumption is broken and it should not be used at all.
ACE's own honesty about this is the useful part: it publishes a method with a stated assumption instead of an unstated judgment. The full ACE ladder and how the company reports per-side percentages is in ACE Grading centering standards.
When you cannot reconstruct: three fallbacks
If the card leaves you with two adjacent borders, or none, reconstruction has nothing to work from. In rough order of preference:
1. Calibrate against reference copies. The nominal border width is a property of the design, not of your card, so you can recover it from other copies. Measure the edge-to-strip gap (or edge-to-element gap) on several copies of the same card or the same design. Individual cuts scatter around the intended position, so the average across a handful of copies approximates the nominal, and your card's deviation from that average is your shift in millimeters. Three copies is a weak estimate, ten is a decent one. A high-resolution scan of a graded example is the best single reference, because its grade puts a published bound on how far off-center it can be.
2. Use the back. Most Pokémon backs carry a full symmetric printed border that measures normally. Collectors report using the back border when the front is ambiguous, and it is a reasonable hint — but front and back are printed in separate passes and registered separately, which is precisely why the companies that publish a back figure at all set it looser than their front figure (PSA allows 75/25 on the back of a 10 against approximately 55/45 on the front; BGS wants 55/45 on the back of a Pristine 10). Treat the back as a strong indication of where the cut fell, and as its own separate measurement against its own line, never as a stand-in for the front number.
3. Report millimeters, not a ratio. If the axis has no valid denominator, do not invent one. "Top gap to the frame line 2.86 mm, bottom edge occupied by the dot strip at 1.40 mm from the cut, measured from a 1200 dpi scan" is a complete, honest, checkable statement. A percentage derived from two gaps that were never designed to be equal is worse than no percentage, because it looks authoritative and is not.
The general problem of choosing a reference element when the design fights you — thin frame lines, double borders, chrome inner borders — is worked through in how to measure centering on full-art and borderless cards.
What graders actually do with these cards
Be clear about the state of the evidence:
- ACE publishes the three-border method quoted above, and publishes per-side percentages on its report. That is the entire published record.
- PSA describes comparing "the measurements of the borders from left to right and top to bottom" and designates centering "as the percent of difference at the most off-center part of the card." It publishes nothing specific to full-bleed or partial-border designs, and its standards reserve a small variance at the grader's sole discretion based on overall eye appeal, plus a 5% leeway on front centering for cards grading PSA 7 or better.
- Beckett publishes a centering ratio for every grade and states that cards are graded on four key categories — centering, corners, edges and surface — but publishes no partial-border rule.
- CGC, SGC and TAG publish tolerances but no partial-border method.
Collectors report that on cards with a missing or occupied border, graders appear to work from whatever consistent printed element the design offers and lean more heavily on eye appeal than they would on a plain-bordered card. That is a reasonable inference from how the standards are written — PSA's eye-appeal clause has more room to operate when there is less to measure — but it is a report, not a published policy, and you should not plan a submission around it.
The practical consequence is a submission rule rather than a measurement rule: on a partial-border card, the measurable axis is the one you can plan around, and the reconstructed axis is the one that can surprise you. If the measurable axis is comfortably inside the line you are aiming at, the card is worth submitting. If the measurable axis is itself borderline, the reconstruction will not rescue it.
Documenting a partial-border card when you sell it
Centering disputes on these cards are almost always disagreements about the reference, not about the ruler. Two people measure the same card, one uses the frame line and one uses the text box, they get numbers eight points apart, and both are arithmetically correct. Pre-empt it by publishing your working:
- The images. A flat 600 or 1200 dpi scan of the front and the back, out of the sleeve, on a contrasting matte backing. A scan removes the perspective argument entirely, which is the second most common dispute.
- The reference, named per axis. "Left/right measured to the outer edge of the colored frame; top measured to the same frame; bottom edge occupied by the dot strip."
- The raw millimeters. All the gaps, two readings per side. A buyer who disagrees with your reference can re-derive their own ratio from your millimeters; they cannot do anything with a bare percentage.
- The method for anything inferred. "Hidden bottom border reconstructed at 2.40 mm as the average of the left and right borders, following ACE's published three-border method."
- The ratios, worst axis flagged. Front and back, both axes, and say which one governs.
- The uncertainty. If two readings on a side differ by more than about 0.2 mm, say so and note the taper rather than quietly averaging it away — see tilted print and diamond cut vs off-center.
The free card centering calculator is useful for exactly this class of card because it works from manual guide lines rather than automatic border detection: you place a line on the cardboard edge and a line on whatever element you have chosen, on each side, and it does the ratio arithmetic and maps the result to the PSA, BGS, CGC, SGC and TAG standards. On a reconstructed axis you feed it your inferred number and label the output accordingly. The underlying procedure for every method is in how to measure card centering at home.
Bottom line
- Count your measurable borders before you measure anything. A dot-code strip or a full-bleed edge does not remove the reference line, it removes the matching pair, which is what the ratio formula needs.
- ACE Grading publishes the only method: average the known pair to get the design's nominal border, then treat the missing border as that nominal. Left 2.62 and right 2.18 give a 2.40 mm nominal; a visible 2.86 mm top then reads 2.86 ÷ 5.26 = 54/46.
- The method's assumption is that all four borders share a nominal width. Being wrong by 0.2 mm costs about 2 ratio points, and being wrong by 0.6 mm can flip which side reads wider — so only use it where the visible borders are clearly the same design element.
- No grading company publishes an official rule for borderless or partial-border measurement; ACE's three-border note is the whole published record, and everything else is collector practice.
- Fallbacks in order: calibrate the nominal from several reference copies, use the back as a hint against its own looser tolerance, or report millimeters and refuse to invent a denominator.
- When selling, publish the scan, the named reference per axis, the raw millimeters, the reconstruction method and the worst axis. Reference disagreements, not ruler errors, are what these disputes are actually about.
Frequently asked questions
How do you measure centering on an e-Reader card?
Count the borders you can actually measure first. Wherever a dot-code strip occupies an edge, that edge no longer has a border of the same nominal width as its opposite, so the normal ratio formula does not apply to that axis. Measure the axis that still has a matching pair normally, then reconstruct the missing border using ACE Grading's published three-border method: average the known pair and treat that average as what the missing border was designed to be. Record the millimeter shift as well as the ratio, because the reconstruction carries an assumption.
What is ACE Grading's three-border method?
ACE's centering blog explains that on cards where only three borders are visible it calculates the average border size from the known pair, for example left and right, and then uses that to work out what the size of the missing border should be. So if left is 2.62 mm and right is 2.18 mm, the average is 2.40 mm, and a visible top border of 2.86 mm is compared against an inferred 2.40 mm bottom, giving 2.86 divided by 5.26, or about 54/46. It is the only published method any grading company offers for a missing border.
Do grading companies have an official rule for partial-border cards?
No. ACE Grading is the only company that has published anything, and its method covers three-border cards specifically. PSA, BGS, CGC, SGC and TAG publish centering tolerances but no separate rule for borderless, full-bleed or partial-border designs. PSA's text describes comparing the borders left to right and top to bottom and designates centering at the most off-center part of the card, which leaves the reference choice to the grader. Treat everything beyond ACE's method as collector practice, not policy.
How wrong can the reconstructed border be?
About two ratio points for every 0.2 mm you are wrong about the missing border's nominal width. Working from a visible 2.86 mm top border, assuming the hidden bottom is 2.40 mm gives 54/46, assuming 2.20 mm gives 56/44, and assuming 2.60 mm gives 52/48. If the design's top and bottom borders were never the same width as the sides, the reconstruction can even flip which side reads as wider, so only use it where the visible borders are clearly the same design element.
Should I use the back of the card when the front has no usable border?
As a strong hint, not as proof. Most Pokemon backs carry a full symmetric printed border you can measure normally. But the front and back are printed in separate passes and registered separately, so the back tells you where the cut fell relative to the back print, which is usually but not always the same story as the front. Every grader that publishes a back figure at all sets it looser than its front figure, so a back measurement is a separate reading with its own line, not a substitute for the front.
How should I describe a partial-border card's centering when selling it?
State four things: which printed element you measured to on each axis, the method you used for any missing border, the millimeter gaps themselves and not only the percentage, and whether the reading came from a scan or a photo. A buyer can re-derive your ratio from raw millimeters even if they disagree with your reference choice, and disagreements over reference choice are the single most common cause of centering disputes on full-art and partial-border cards.


