The method
Do the math before you slab.
Every verdict is the same calculation, shown in full. Here's what goes in, what comes out, and where it can be wrong.
The question we answer
Not “what will this card grade?” but: are you likely to end up with more money by grading this card than by selling it raw? That's a question about fees and odds, and it's one a phone can help with.
1. Centering
We find the card's outer edge and the inner frame of the artwork, then measure the four borders. Centering is reported the way graders talk about it: the bigger side's share, like 55/45 left/right and 60/40 top/bottom. You can drag the guides if the auto-detect is off, and a manual adjustment always wins.
We compare that to each grader's published centering tolerances to find the highest grade your centering still allows. A card that measures 63/37 on the front won't get a 10 no matter how clean the corners are, so the 10 comes out of the odds before anything else happens.
Measurement runs in your browser. The photo never leaves your device.
2. Grade odds
We start from a gem-rate prior: roughly how often cards of that era, finish and rarity come back a 10 when they look clean. Then we adjust for your centering caps and your own condition check (corners, edges, surface, on a 1–5 scale). The result is a probability for each grade, shown as a chart like this:
| Grade | Probability |
|---|---|
| PSA 5 or lower | <1% |
| PSA 6 | 2% |
| PSA 7 | 6% |
| PSA 8 | 22% |
| PSA 9 | 52% |
| PSA 10 | 18% |
The prior is labelled with its basis, and you can override it. If you know your set gems at 40% for clean pulls, put in 40% and watch the verdict move.
3. Prices
Raw near-mint price comes from public market data. Graded prices come from recent sales when we have them. When we don't, we estimate graded prices with multipliers on raw, and those numbers are always labelled estimate.
4. Costs
- The grading fee for the tier that matches the card's likely graded value (see the fee table).
- Shipping and insurance per card, both ways.
- Selling fees on the graded sale (default 13%) and, for a fair comparison, on the raw sale too.
- Optionally, price drift over the turnaround time. Markets move while your card sits in a queue.
5. The verdict
Expected value of grading = Σ P(grade) × net graded sale − costs. Compare to the net raw sale. The difference is the headline number, along with the probability that grading loses money and the lowest grade that breaks even.
Expected gain is negative, or the upside depends on a long-shot 10.
Roughly break-even, or the answer swings on an assumption you should check.
Positive expected gain with a reasonable chance of not losing money.
What we can't do
- See surface. Scratches, print lines and dents under holo foil are what usually separate a 9 from a 10, and a phone photo doesn't reliably catch them. That's why your condition check matters, and why our odds are a range.
- Know your card's history. We see one photo. We don't know if it spent a year in a top-loader or a shoebox.
- Predict the market. Graded prices move, sometimes a lot, between submission and return.
- Read a grader's mind. Grading is done by people. The same card can come back differently on different days.
The honesty rules
- We say “PSA-style grade estimate,” never “PSA grade.” We aren't affiliated with any grader.
- Every verdict shows its assumptions and a probability range, never a promised grade.
- Every estimated number is labelled “estimate.”
- The fee table carries a last-verified date.
- We publish our misses.