Product
Everything below is built and working today. What it does not do is at the bottom, said plainly.
Feature 01
Point it at the card and it finds the four border edges, then compares the gaps and gives you a real percentage.
No machine learning and no server round trip. It sums how sharply brightness changes along each axis, which puts a spike exactly where a border begins. When it cannot find four clean edges it says so and returns nothing, because a made up measurement is worse than none.
Measured, then priced
Feature by feature
It does not tell you the card is an eight. It tells you how likely every grade from one to ten is, and computes the money over all of them.
Two graders can look at the same card and differ, so a single number would be false precision. The weakest attribute caps the whole card, forgiven by at most one point, because a card with one scratched surface is a scratched card rather than an average.
Postage both ways, insurance that scales with value, sleeves and toploaders, then the marketplace fees at the other end.
Postage is charged per package rather than per card, so the batch size changes what every card in it costs. Drag the batch slider to one and watch the answer fall apart.
The lowest grade at which submitting beats selling the card raw today, and your chance of landing at or above it.
This is the number the decision actually turns on. A card that needs a ten to break even is a lottery ticket; a card that breaks even at a seven is a different proposition entirely, even when the headline value is the same.
A sold price you entered, a market data provider, or the app’s own estimate. Marked at every grade.
And when the whole curve is estimated, the app refuses to say send unless the edge would survive the estimate being wrong by the size of the fee. It would rather say borderline and explain why than be confidently wrong with your money.
A card can clear every money test and still be refused for being worth more than the tier accepts.
Saying no when the real answer is yes on a different tier is a bad answer, so the app names the cheapest tier that would take the card instead.
Add everything you are considering and it chooses the tier, the batch size and which cards actually go in.
Because postage is shared, the best box is not the best card repeated. It also flags cards that only look worth sending because better cards are carrying the postage for them.
Both piles side by side, each valued the same way, with the answer between them so the phone can be turned around mid trade.
When more of the trade is estimated than the gap between the two sides, it says it cannot call it rather than picking a winner. A kid with a binder is the case it has to be right for, and a confident answer built on guesses is the way that kid gets taken apart.
Real sold listings, a fair value with its range and confidence, a price chart at any grade, and every grade priced side by side.
Kept beside the verdict rather than folded into it. A handful of sales and a model estimate are evidence, and each one is labelled with where it came from and how far to trust it.
This week’s biggest risers in every category, and any graded card checked by its certificate number.
Implausible gains are filtered out before they reach the list. A certificate lookup says whether that number is on record, which card it was issued to, and what that grade has sold for lately, and a Pro member can watch that grade’s price straight from the answer.
If a lower grade ends up worth more than a higher one, something is wrong and it says so.
The estimates can never do this, so an inversion always means a supplied price is off. Usually a sold price typed against the wrong grade, or a comp from a different variant.
Point the camera at a card and it comes back as cards from the catalog, with the parallel worked out: the base card and the silver refractor of the same number are told apart, which is the distinction that decides whether a verdict is right or wrong by a factor of seven.
It never picks for you. The match it chose is marked with how confident it says it is, every other candidate stays on screen, and nothing is priced until you tap one. The card is in your hand where you can look at it, and a model's confidence in its own answer is not the same thing as being right. One player in one year has ten cards that differ in price by thousands, so a tool that chose silently would be wrong while looking completely right.
On the trade screen, point the camera at the label rather than the card. It reads the grading company and the certificate number off the plastic and looks up what is inside, so a slab somebody just handed you lands in the pile as that card at that grade, with no searching and no choosing a grade from a menu.
The certificate is shown as read, because it is checkable on the grading company's own site while you are standing there, and that is worth more than anything this app can tell you. When the provider matched the card from the words on the label rather than by number, it says so and shows how sure it is.
Half grades come along too. An SGC 9.5 goes in as a 9.5, priced from what 9.5s of that card actually sold for. When there is no such sale the row says est as 9 and prices it as the grade below, so a slab is never promised more than it has shown it can fetch.
Pick a card and the app asks what else shares its set, its number and its subset. One player in one year can have ten cards that differ in price by sixty times, and the dangerous case is not picking the wrong row. It is picking the right row and never learning the others exist, so the verdict is computed from the base card while a gold refractor sits in your hand.
A search needs every word to match, so it can come back empty for a card you could describe in one sentence. When that happens, the app offers its best guess instead: the single closest card, how sure it is, and one line on why it picked that card.
That line is there so you can disagree with it. If nothing comes close enough, the app tells you so plainly rather than showing you a card that is probably wrong.
Two questions a search box is bad at: what is in this box worth, and what else is there of this player. Pick a set or a player and you get every card in it, sixty at a time, with what each one sells for: best selling first, or the highest priced, the fastest rising over 30 days, or A to Z.
With Pro, a player also comes with their market: twelve weeks of sales, and whether the last four are running ahead of or behind the four before. It counts every sale that names them, across all of their cards, and says so.
Browse only looks things up when you tap Look, so nothing loads that you did not ask for.
With Pro, the market page answers the question under every grading decision: what does this card actually sell for at the grade I would get? Name a card, pick PSA 10, PSA 9, BGS 10, BGS 9.5, SGC 10 or CGC 10, and see every sale at that grade in the last 90 days, averaged.
The number of sales sits beside each average, and an average from one or two sales says so, because one odd sale can make it. A card that did not sell at that grade says that in words rather than showing $0.
With Pro, any card page has a Watch the price button. Pick a grade the card has a sale at, and the Monday email tells you when the latest sale at that grade moves a tenth either way, or 5% or a quarter if you choose. A card is reported once per move, then watched from the new price.
Or name your own price. Set a target on any watched card, one number: under the latest sale and the email waits for a drop to it, over it and the email waits for a rise. You are emailed once, the same day the daily price check sees a sale get there, not the following Monday.
Prices are rechecked in turn, a few cards a day, the longest unchecked first, so watching a card never slows the site down for anybody else. Your watchlist, with every starting price and every change, has its own page, where you can put it in order by the biggest move and download it as a spreadsheet. Cards you watch carry a star in Browse, and each card's page draws where its price has been since you started watching. Own one already? Note what you paid, and every row says how the latest sale compares, with the whole list added up at the top and an order that puts your best buy first. Each card also takes a short note of your own, like which binder it is in.
Sports cards. Priced by era. Cards from the junk wax years, printed in such numbers that grading adds little, get their own lower figures. The card you pick in the search box is the same card that gets priced at every grade.
The games, in the same search. Pokémon, Magic, Yu-Gi-Oh and the newer games sit next to the sports cards, so one search box covers the whole table rather than half of it.
Keep lists without Pro. With Pro, your box plans, trades and tracked submissions are saved to your account and follow you to every phone and computer you sign in on, the tracker adds every box up into a track record, and the account page downloads all of it as one file. Without it they are kept in the browser you made them in, and clearing this browser's data clears them.
Automatic grading. Nothing here gives a card a grade. It works out what a grade would be worth and how likely each one is, which is a different and more useful thing than a number no buyer would trust.
Automatic condition from photos. Centering is measurable. Corners, edges and surface are not, not reliably, and an app that pretended otherwise would be confidently wrong about the attributes that cap a grade.
The house card
Every number on the back is read from the same tables the calculator runs on: 4 grading companies, 21 service tiers priced, a 1–10 scale and four verdicts. When the app adds a company, the card shows it too.
It is a card we drew ourselves, so it shows no set and no character that belongs to anybody else.