Measured study · 320 measurements
Under half of raw cards clear PSA 10 centering.
Every centering calculator online tells you what the 55/45 rule is. None of them tell you how often real cards pass it. We measured 302 raw Pokémon cards from eBay listing photos — 320 measurements in all — and 48.1% cleared PSA's front-centering window. On centering alone. Before corners, edges or surface are even looked at.
Measured March 2026 to July 2026 with our own review tool. “Worst side” means the larger share on each axis, then the larger of the two axes — the way the hobby quotes it.
The headline number, and what it costs you
154 of 320 measurements sit inside the PSA 10 front-centering window. That is roughly one in two. So if you buy raw cards hoping for a 10, centering by itself is coin-flip odds — and it is the one defect you can fully judge from a listing photo before you spend a cent. Corners, edges and surface you are guessing at. Centering you can read.
That is the whole argument for training your eye on it: it is the cheapest grade-killer to screen out, and it is measurable from the photo the seller already gave you.
Top/bottom is what actually kills the card
Split the two axes and they are not close. The left/right split clears the PSA 10 window 76.9% of the time; top/bottom clears it only 61.3% of the time. When a card has a limiting axis, it is the vertical one 57.5% of the time.
| Axis | Clears 55% | Average worst side | Is the limiting axis |
|---|---|---|---|
| Left / right | 76.9% | 53.9% | 42.5% |
| Top / bottom | 61.3% | 55.1% | 57.5% |
Practical read: people eyeball the left and right borders because a side-shifted card is obvious. The vertical shift is the one that slips past, and it is the one more likely to cost you the 10. Check top-to-bottom first.
How far off are the misses?
Most failures are near misses, not disasters. 41.6% of measurements land between 55% and 60% on their worst side — outside the 10 window, inside the 9 window. Only 10.3% are worse than 60/40. The distance between a 10 and a 9 is usually two or three percentage points, which is exactly why calling it by eye is a real skill and not a party trick.
| Worst side | Measurements | Share | Verdict |
|---|---|---|---|
| 50.0-52.5% | 47 | 14.7% | Dead centre |
| 52.6-55.0% | 107 | 33.4% | Inside the 10 window |
| 55.1-60.0% | 133 | 41.6% | Misses 10, clears 9 |
| 60.1-65.0% | 15 | 4.7% | Misses 9 |
| over 65% | 18 | 5.6% | Badly off |
Worst measured in the sample: 88.1%.
How the sample was chosen — read this before quoting the number
This is not a random sample of eBay. These are cards that came up during real buying research between March 2026 to July 2026 — cards someone pulled up and looked at closely because they already looked like they might be worth buying. Pre-filtered by a human eye, in other words.
Which makes 48.1% an optimistic ceiling, not an average. The pass rate across all raw Pokémon cards listed on eBay is very probably worse than this, because this sample already had the obviously-crooked ones thrown out. If you want the honest version of the headline: even among cards that had already survived a sniper's first glance, over half missed the PSA 10 centering window.
Other limits worth stating: 320 measurements across 302 cards, because some cards were measured from two photos. Measurements are taken from the seller's own listing photo with perspective correction, not from the card in hand — a photo shot at an angle measures differently than a flatbed scan, and we only keep measurements where the tool actually found the card's four outer corners. And centering is one input among several: a perfectly centred card still fails on a soft corner. Nothing here predicts a grade.
Use the same tool on a real listing
Every card above was measured with the same centering + zoom tool you can use right now on a card's current eBay listings. Pick a card, pull up its photos, and check the centering yourself before you buy.