Grade a card only when two things are true at once: a graded market for that exact card visibly exists in recent sold listings, and the money you expect back after grading beats what the card nets you raw today by enough to justify waiting months and risking a lower grade than you hoped for. Run that as a real calculation and most cards fail it.
The mistake is almost always the same one. People compare what the card sells for in a 10 against what it sells for raw, and skip every line in between — the fee, both legs of shipping, the marketplace cut, and the odds that the card doesn’t come back a 10.
What follows is the whole method: where the graded premium comes from, how to measure it for the card in your hand, how to grade your own card honestly before you pay anyone, and how to run the arithmetic. Budget twenty minutes for the first card. Every card after that takes about two.
What you’re actually buying when you pay for grading
Grading doesn’t improve a card. Four practical things create the gap:
- The buyer doesn’t have to trust you. A raw card in a photo could have a crease the light didn’t catch. A number on a label is a third party’s opinion, and it removes the risk the buyer was pricing in.
- Authenticity. On expensive vintage and on autographs, the grader is also confirming the card isn’t a reprint, a forgery, or a trimmed copy. For some cards that’s most of the premium.
- Liquidity. Graded cards sell faster and to a wider pool, including buyers who won’t touch raw. A card that sells in three days at a fair price is worth more in practice than one that sits for four months.
- It stops the clock. Sealed in the holder, the card can’t get worse. Handling, humidity, and one bad shipping envelope stop mattering.
Notice what isn’t on that list: the card itself didn’t change. Grading describes it, credibly. That’s the entire product.
The raw vs graded gap is a curve, not a multiplier
Anyone who tells you “graded cards are worth three times raw” is selling something. The gap runs from negative — where grading cost more than it added — to enormous, and where you land depends on the grade you get and on whether the card has a graded market at all.
What is reliably true is the shape. The jump from a 9 to a 10 is usually far larger than the jump from raw to a 9. Most of the money lives in that top step, which is also the step you can’t control. Vintage behaves differently enough that it needs its own treatment — the rules that apply to older cards are not the ones below.
| Grade | What it means in practice | Typical relationship to raw |
|---|---|---|
| Gem Mint 10 | No flaw visible at 10x magnification; centering tight front and back | The largest premium by a wide margin. This is where the money is. |
| Mint 9 | One minor flaw — a slight centering miss, a hint of corner softness | A modest premium on modern cards, often a large one on vintage. |
| Near Mint-Mint 8 | A couple of minor flaws, or one you notice immediately | On modern cards, frequently at or below what grading cost. On vintage, still meaningful. |
| 7 and below | Visible wear | On modern cards, usually a loss. On scarce vintage, the holder still adds trust and value. |
So submitting a modern card is a bet on getting a 10. If your honest odds of a 10 are one in three, and the 9 sells for barely more than raw, the submission can lose money even though the 10 looks like a big win.
How to measure the premium for your specific card
- Identify the card exactly. Year, brand, set, card number, and parallel. A base card and a colored parallel of the same player have completely different graded markets. The back carries the set name and card number; the parallel is usually named in small print or identified by border color and finish.
- Pull raw sold comps. Search the exact card on eBay, filter to Sold Items, sort by most recent, and look at the last 30 to 90 days. Ignore active asking prices completely — anyone can ask anything.
- Pull graded sold comps grade by grade. Run the same search with “PSA 10” appended, then “PSA 9”, then “PSA 8”, and again for SGC and BGS if copies have sold. Record the median of each, not the highest sale. The same number from different companies doesn’t always fetch the same price, so comp the company you actually plan to use.
- Count the sales. If only two 10s have sold in six months, that’s thin evidence. Thin markets swing hard, and one optimistic sale can set a comp that never repeats.
- Do the subtraction. Graded sale price, minus the marketplace fee, minus the grading fee, minus shipping and insurance both directions, minus the raw price you could get today. What’s left is the real gain, and it’s usually much smaller than the headline gap suggested.
Write those numbers on paper. The comparison talks people into submissions precisely because they hold the 10 comp in one hand and the raw comp in the other and never write the middle lines.
Where to find the numbers this depends on
Two inputs change often enough that no article should hand you a figure. Get them yourself, the week you plan to submit.
- The grading fee. Every grading company publishes a current service-level page listing its tiers. The thing to understand is that tiers are set by the card’s declared value, not by how fast you want it back — a modest card and an expensive card don’t pay the same fee even in the same envelope. Find the tier your card actually qualifies for. That entry tier is the number that decides whether this math works at all.
- The marketplace fee. eBay and every other platform publish a fee schedule with a final value fee for trading cards, and the rate can differ by category and by seller account type. Look up the rate that applies to you, and remember you pay it on the higher graded price — it scales up along with the premium.
Turnaround works the same way. Companies post current estimated turnaround per tier on their own sites and it moves with submission volume, so check the live estimate rather than what a forum thread said a year ago — then assume it runs longer than posted.
Pre-grade the card yourself, honestly
You can’t estimate the odds without assessing the card the way a grader will. Four things, in this order, with a bright light and 10x magnification — a cheap loupe or your phone’s macro mode is enough.
Centering
Compare the left border to the right, and the top to the bottom. It’s expressed as a ratio: 50/50 is perfect, 60/40 means one side’s border is half again as wide as the other. Top grades generally want the front within roughly 55/45 to 60/40, with more tolerance on the back. Check the back anyway — plenty of cards are centered on the front and clearly off on the reverse, and the back counts.
Centering is the most common reason a clean-looking card comes back a 9, and coming straight out of a pack says nothing about it. Cards are damaged during cutting and packing, before you ever see them.
Corners
All four, magnified. Fuzz, softness, a rounded point, or white paper showing through a colored border. A corner that looks perfect to the naked eye often shows wear at 10x, and the grader is using magnification.
Edges
Run your eye down all four for nicks, chipping, and rough factory cuts. Dark borders — black, navy, deep red — show every speck of edge wear, which is why dark-bordered cards grade high less often than they look like they should.
Surface
Tilt the card and look across it under the light rather than straight down at it. Print lines, scratches, dimples, gloss breaks, and fingerprints only appear at an angle. Chrome and foil show surface problems that plain paper stock hides.
Then write down the grade you’d give the card if it belonged to someone else. That number, not your hope, goes into the calculation.
The one calculation
In plain English:
What you expect from grading = (chance of a 10 x what a 10 nets) + (chance of a 9 x what a 9 nets) + (chance of an 8 or lower x what that nets) – grading fee – shipping both ways – supplies
Compare that against what the card nets you raw today, after marketplace fees and shipping. If grading doesn’t win by a comfortable margin, sell it raw. If you’d rather see the same subtraction laid out line by line with somewhere to put each number, the break-even walkthrough does that.
Here’s the arithmetic on made-up numbers — illustration, not comps for any real card, written in generic “units” so nobody mistakes them for prices. Call the raw net 100 units. Say a 10 nets 500, a 9 nets 130, an 8 nets 75, and grading costs 60 all-in. If you honestly believe the card is 25% a 10, 50% a 9, and 25% an 8, the expected value is (0.25 x 500) + (0.50 x 130) + (0.25 x 75) – 60, which works out to 125 + 65 + 19 – 60 = 149, against 100 for selling raw. That clears the bar.
Now change one input. Drop the odds of a 10 from 25% to 10%, leave everything else alone, and the same card falls to roughly 100 — a coin flip that costs you three months. The probability estimate does most of the work here, which is why the self-grading section matters more than the arithmetic does.
The costs people forget
- Shipping to the grader with tracking and insurance, plus the return shipping charge.
- Supplies — semi-rigid card holders, penny sleeves, a box that won’t crush.
- Your time filling out the submission, packing carefully, and tracking it.
- Money tied up. You pay before you know the outcome, and it’s gone whatever comes back.
- The bulk trap. Cheaper per-card bulk tiers tempt people into submitting fifty marginal cards. Fifty small fees is a large fee, and if forty come back worth less than what they cost to grade, the bulk rate saved you nothing.
When to grade and when to sell raw
| Usually worth grading | Almost never worth grading |
|---|---|
| Vintage star cards, where authentication matters as much as condition | Cards worth single digits raw — the fee exceeds the card |
| High-value modern cards that survive the four-point check, with a graded market already visible in sold listings | Cards with a flaw you can already see: a soft corner, a crease, a print line |
| Cards from high-print-run eras that somehow survived immaculate, where the top grade is scarce even though the card isn’t | Cards with no graded sold comps at all — you’d be paying to find out whether a market exists |
| Anything you’re keeping long term and want protected and authenticated regardless of resale | Anything you need to turn into cash soon, and base cards of current players in ordinary condition |
Two rows in that table deserve a footnote. The high-print-run row is why cards from the Jordan and Kobe era keep climbing despite millions of copies existing: raw supply is enormous, top-grade supply isn’t. And the no-comps row has an extreme version — a one-of-one, where there is nothing to compare against at all and scarcity alone doesn’t set a price. The full list of situations where the answer is simply no is in when not to grade a card.
If you sell raw, sell it well
A raw card sells for less because the buyer is pricing uncertainty. You can shrink that discount without paying anyone a fee.
- Photograph the front and back straight on, in focus, filling the frame. No angle, no glare, no hand in the shot.
- Add one angled shot under a light that shows the surface honestly. Buyers trust a seller who shows the surface instead of hiding it.
- Describe the flaws in words. “Slight softness on the bottom-left corner, print line visible at an angle across the jersey” turns a risky purchase into a known one, and it heads off returns.
- Ship it properly — penny sleeve, toploader taped shut, bubble mailer, tracking on anything with real value. Packing a card so it arrives flat takes a minute longer and prevents the most common kind of dispute.
A well-photographed, honestly described raw card often sells close to what a mid-grade holder would bring, without the fee or the wait.
What to do next
Take your single best card and fill in the four inputs: raw median, graded medians at 10, 9, and 8, your honest self-grade after the four-point check, and total cost from the current fee schedule. One card, twenty minutes. Whatever the answer is, you’ve learned the method.
If the numbers come out close, sell raw — a close call plus a months-long wait plus the risk of a lower grade is not a close call. And if they say sell raw outright, that’s a real answer rather than a consolation prize. Selling for a fair price today beats paying a fee to find out your card was a 9.
Common questions
Should I grade or sell raw if I don’t know what the card is worth?
Sell raw, or find out first. Grading a card whose value you haven’t established is paying a fee to learn something a free sold-listings search would have told you in five minutes.
My kid pulled a big card out of a pack. Should we grade it?
Check it the same way you’d check any other card, and know that pack-pulled cards fail on centering all the time. If the graded comps are strong and the card survives the four-point check, it can be worth it. If what you really want is to protect and display it, a one-touch magnetic holder does that for a fraction of the cost and no waiting.
Is it worth grading a whole inherited collection or a full set?
Not as a batch. Sort first, comp the best ten or fifteen cards individually, and grade only the ones where the math works on their own. If the collection came from someone you lost, it’s also worth separating the cards you’re keeping from the ones you’re deciding about financially. Those are two different piles and they need two different decisions.
Why did my card come back a 9 when it looks perfect?
Almost always centering or surface. Measure the border widths on the back as well as the front — a card can be centered on the front and obviously off on the reverse. Surface flaws like print lines and gloss breaks only appear when you tilt the card under a light, which is what the grader did and you probably didn’t.
