Open Source Grading
Community 1000. Not a cert. Beta
Front
One photo. Fill the frame. JPEG or PNG.
Tap Take or Choose
Front SHA-256 —
Continue to back
Back
Required — cannot skip.
Same card, other side
Back SHA-256 —
This reads the photo in front of you, not a lab scan. A phone can measure centering more consistently than eyeballing. Glare and JPEG lie about surface. That does not make a camera better than a human grader overall.
Card
UnknownPhotograph front, then back, to read the card.
Unknown card
Name it yourself — stored as user-entered
Centering
Add a front photo of the card to measure L/R and T/B.
Centering why — photograph front and back, then measure or set manual.
offoff-mixedmixedmixed-eveneven
Corners, edges, surface
crushedroundedmixedsoftsharp
Corners why — set the slider. Optional phone-photo hint looks at corner crops.
chippedwornmixednickedclean
Edges why — set the slider. Optional phone-photo hint looks at edge strips.
damagedscuffedmixedmarkedclean
Surface why — set the slider. Optional phone-photo hint looks at interior pixels.
Published rubric
The tables below are the grade. Not PSA language. There is no category detector.
How percents are counted
Left% = round(L / (L+R) × 100), Right% = 100 − Left%. Top% = round(T / (T+B) × 100), Bottom% = 100 − Top%. Worse L/R = max(Left%, Right%). Worse T/B = max(Top%, Bottom%).
Axis table — L/R and T/B each (0–125)
Find the row for worse-side percent. Do this twice. Centering = L/R points + T/B points (0–250).
| Row | Worse side % | Shown as | Points |
|---|---|---|---|
| A1 | 50 | 50/50 | 125 |
| A2 | 51–52 | 51/49–52/48 | 118 |
| A3 | 53–55 | 53/47–55/45 | 105 |
| A4 | 56–60 | 56/44–60/40 | 88 |
| A5 | 61–65 | 61/39–65/35 | 68 |
| A6 | 66–70 | 66/34–70/30 | 45 |
| A7 | 71–75 | 71/29–75/25 | 25 |
| A8 | 76–85 | 76/24–85/15 | 10 |
| A9 | 86–100 | 86/14 or worse | 0 |
Example: L/R 60/40 → A4 = 88. T/B 50/50 → A1 = 125. 88 + 125 = 213.
Feel table — corners, edges, surface, and manual centering (0–250)
| Band | Range | Centering (manual) | Corners | Edges | Surface |
|---|---|---|---|---|---|
| B1 | 0–50 | off | crushed | chipped | damaged |
| B2 | 51–100 | off-mixed | rounded | worn | scuffed |
| B3 | 101–150 | mixed | mixed | mixed | mixed |
| B4 | 151–200 | mixed-even | soft | nicked | marked |
| B5 | 201–250 | even | sharp | clean | clean |
Total = Centering + Corners + Edges + Surface, shown as those four integers added, /1000. Catalog identity is not part of the total.
Open Source Grading
Community receipt
Sports cards and TCG. Not a certificate. Not TAG. Not PSA.
This reads the photo in front of you, not a lab scan. A phone can measure centering more consistently than eyeballing. Glare and JPEG lie about surface. That does not make a camera better than a human grader overall.
Card — Unknown
—/1000
Total why —
- Centering—
- Corners—
- Edges—
- Surface—
Community receipt. Not a certificate. Not TAG. Not PSA.
Does not publish. Saves the receipt and both photos in this browser. My catalog lives on this device. Clearing browser data can erase it.
Publish is optional. If the photo write fails, grades still list without images. Nothing is invented.
Open Source Grading
QR only. Payload is this card page URL. About 1.25 inch square for a magnetic case.
QR could not draw. Copy the URL above.
Copyable JSON
Copied.
Recent community grades
A rolling meter of anonymous receipts. Newest activity first. Each QR and /c page is one physical card instance, not a player/set/number. Regrades of that instance stay together; two copies of the same catalog row do not share a code. Lookup by name/year/set lists every published copy, each with its own page. An instance average is the mean of that copy’s grades, not a rank. Not a leaderboard, not a ranking, not a pop report. Average uses round half even (ties to even).
Loading the meter…