Raw vs Graded Cards: When to Grade
Keep a card raw when the raw sold price covers what you paid and the expected graded price (probability weighted across likely grades, minus total grading cost) does not beat raw by a clear margin. Grade when the expected Gem Mint price after cost exceeds raw by roughly 30 percent or more. The decision is a probability-weighted math problem, not a gut call.
What "raw" and "graded" actually mean
A raw card is a card in a penny sleeve, toploader, or semi-rigid holder with no third-party assessment. It trades on trust. Buyers look at photos, read seller reviews, and accept some risk that the card's centering, corners, edges, and surface are not exactly what they seem.
A graded card has been sent to a third-party authenticator (PSA, BGS, SGC, CGC, or a smaller player), examined, assigned a numeric grade on a 1 to 10 scale, and sealed in a tamper-evident plastic slab with a unique certification number. The slab removes authenticity risk and replaces "I think this is Mint" with a documented grade that the market can price. For the AI-extractable quick read on the same split, our graded and raw cards answer page covers the basics in a tighter format.
The raw card and the graded card are the same piece of cardboard. What differs is information. Grading converts subjective condition guesses into a publicly verifiable number, and buyers pay a premium for that certainty. The entire raw vs graded question is whether that premium, after fees, beats holding or selling the card raw.
If you want the full grading-scale mechanics, start at our PSA grading guide, BGS grading guide, SGC grading guide, or CGC grading guide. This page is about the economics of the decision.
When you are ready to apply the economics to a specific card in your hand, take it through our grading ROI decision tree. That page runs the same raw-vs-graded math on a per-card basis with a 10-question screen and three worked examples.
The real cost of grading
Grading fees are the visible part of the cost. They are not the whole cost. A realistic all-in number for a single card submitted through a mid-tier service in 2026 runs roughly 25 to 60 dollars, depending on declared value, grader choice, and whether you use a group submitter.
| Cost item | Typical range (USD) | Notes |
|---|---|---|
| Grading fee (base tier) | $18 to $45 | Value / Economy tiers at PSA, SGC, and CGC. BGS runs slightly higher on average. |
| Shipping to grader | $5 to $12 | Depends on insurance and distance. A single card ships cheaper as part of a bulk. |
| Return shipping | $8 to $25 | Insured return is non-negotiable on anything over a few hundred dollars of declared value. |
| Supplies (sleeves, semi-rigid) | $0.25 to $1 | Small per card, adds up at scale. |
| Your time | Non-zero | Prep, submission form, wait, receive, list. Worth pricing in on small-dollar cards. |
| Turnaround opportunity cost | Varies | Value tier can run 30 to 90 days. The market can move in that window. |
Use 25 to 35 dollars as a working floor for a single modern card through a Value or Economy tier, and push the number up to 50 to 60 dollars for faster service or higher declared value. Bulk submissions (50 plus cards) and group submitters can drop the per-card number to the mid teens in some cases, which changes the math materially on lower-priced cards.
The expected-value formula
The whole decision collapses to one equation. Call it the grade EV. If grade EV minus total cost beats raw price by a meaningful margin, grade the card. If it does not, keep it raw.
Worked example on a modern basketball rookie parallel (illustrative only, not a recommendation):
- Raw sold price: 140 dollars
- PSA 10 sold price: 520 dollars
- PSA 9 sold price: 150 dollars
- PSA 8 sold price: 90 dollars
- All-in grading cost: 35 dollars
- Your honest pre-grade probabilities: 55 percent PSA 10, 35 percent PSA 9, 10 percent PSA 8 or worse
EV = (0.55 × 520) + (0.35 × 150) + (0.10 × 90) − 35 = 286 + 52.5 + 9 − 35 = 312.5 dollars.
Raw is 140. Grade EV is 312.5. The expected gain is 172.5 dollars, more than double raw. Grading is the better call, as long as your probability estimates are honest.
Now run the same formula with worse inputs: centering is a little off and your PSA 10 probability is really 20 percent, not 55.
EV = (0.20 × 520) + (0.60 × 150) + (0.20 × 90) − 35 = 104 + 90 + 18 − 35 = 177 dollars.
Raw is 140. Grade EV is 177. The gain is 37 dollars pre-variance. That is inside the "grade only if you feel lucky" zone, and most serious submitters would keep this card raw or sell it raw to avoid the downside.
Where graded wins
Some cards belong in a slab. The graded market has real structure, and the cards that benefit most share a few features.
- High-value modern rookies with a clear grade premium. A PSA 10 of a star rookie in a premium product routinely trades at 3x to 10x the PSA 9, and the raw price sits well below the PSA 9. The grade multiplier covers the fee with room to spare.
- Vintage cards with authenticity risk. A 1952 Topps Mantle, a Tobacco-era T206, or a 1980-81 Jordan rookie raw is buyable only by someone willing to absorb counterfeit risk. Slabbing removes that risk and opens the bid to the full market.
- Cards where population data drives demand. Pop-report watchers pay for the documented grade and the cert number. They will not pay raw prices for cards they cannot verify.
- Cards with visible Gem Mint potential. Tight centering, crisp corners, no visible surface flaws, and a clean edge profile. If the card is an obvious 10 candidate, the expected value math almost always favors grading.
- Cards in thin-float markets. Pre-war cards, vintage hockey, and some international soccer issues have so few copies moving that a slabbed grade is the only way to establish a market price at all.
If you collect modern basketball, football, baseball, or Pokemon, grading is a first-order decision for most of the cards you pull that have long-term upside. Our sport hubs (basketball, baseball, football, Pokemon) flag which products are producing graded trade flow and which are not.
Where raw wins
Not every card should be graded. A lot of raw cards generate a better net outcome if they stay in a toploader.
- Junk-wax era commons and minor stars. PSA 10 on a 1989 Topps common might trade at 8 to 15 dollars. The all-in grading cost is 25 to 35. You are guaranteed to lose money.
- Cards with a visible flaw. A soft corner, a print line on the foil, or an edge nick caps the grade at 8 or 9 and takes the Gem Mint upside off the table. Raw sells the card honestly and avoids a fee on a known 9 ceiling.
- Cards you plan to sell quickly. If your exit is in two weeks, the 30 to 90 day turnaround window at base service tiers does not fit. Submitting a card you need to sell fast is a recipe for selling into a different market than the one you projected.
- Cards where the 9 to 10 gap is small. When PSA 10 trades at only 1.3x the PSA 9, a failed 10 wipes out the grading fee. The gap has to be big enough to absorb variance.
- Cards of players with uncertain career arcs. A rookie whose value is held up by a hot start can compress fast. Grading costs go to zero in a hurry if the player cools. Consider the career-risk curve before you submit.
- Trade-only cards inside a collection. If you are swapping within a local group who buys raw, grading is friction, not value.
The brute honest truth is that most printed trading cards should stay raw. The pool of cards that genuinely benefit from a slab is narrower than the hobby narrative suggests, and grading companies have a built-in incentive to push a wider submission funnel than the expected-value math supports.
The pre-grade quality check
Before you drop a card into a submission stack, run it through a quick physical inspection under good lighting. A phone loupe, a desk lamp, and a dark mat are enough. Four attributes, same as the grader.
- Centering. Measure the left-right and top-bottom borders. Modern Gem Mint tolerance on the front is roughly 55/45 to 60/40 or better on both axes, with the back closer to 75/25. A visibly off-center card is capped at 9 most of the time.
- Corners. Sharp tips, no rounding, no fuzzing. A single soft corner drops the grade. Check all four corners under the loupe; the back corners matter too.
- Edges. No whitening on the color borders, no chipping, no micro-rolls. Run a fingernail lightly along each edge; a catch means a problem.
- Surface. No print lines, no scratches, no touched foil, no indentations from the back of the next card in the pack. Tilt the card under the light so you catch angle-only flaws.
If the card fails any one attribute at the Gem Mint threshold, adjust your probabilities down. Our what is a PSA 10 guide covers the Gem Mint criteria in more depth, and the PSA grading guide includes the full 1 to 10 scale for context.
Choosing the right grader
The grader matters. Not every slab trades at the same premium, and the right grader is the one whose population, reputation, and turnaround fit the card.
| Card category | Strong grader choices | Why |
|---|---|---|
| Modern sports (NBA, NFL, MLB) | PSA, then BGS | PSA has the deepest graded population and the tightest sold-comp liquidity. BGS still trades a premium on Black Label. |
| Vintage baseball (pre-1980) | PSA, SGC | PSA is the default. SGC has strong vintage respect and sometimes a higher Gem Mint rate. |
| Pre-war (T206, E-series) | PSA, SGC | PSA and SGC dominate pre-war. BGS and CGC are rare in that market. |
| Pokemon / TCG | PSA, CGC | CGC has carved out significant TCG share. PSA still runs premiums on English Pokemon. |
| Modern parallels with subgrades | BGS | Black Label 10 trades at a premium to PSA 10 on select modern cards. |
| Bulk low-value cards | CGC or SGC via bulk | Lower per-card fees and faster turnaround on bulk-tier submissions. |
Before you pick, check the live sold comps for the exact card in each grader's slab. A PSA 10 and a BGS 9.5 on the same card can trade 20 to 40 percent apart in either direction, and the premium structure changes with cycles. Do not assume yesterday's grader preference still holds.
Five-step decision framework
Run every card through the same five steps. Skip no steps.
- Pull the raw sold comp. Three to five recent sold listings, same year, set, card number, parallel. Sold prices only. Drop outliers.
- Pull the graded sold comps. PSA 10, PSA 9, BGS 9.5 if applicable. Note dates. Recent comps beat stale ones.
- Run the pre-grade quality check. Honest probabilities on 10, 9, and below. Write the numbers down. Use them.
- Compute grade EV minus total cost. Apply the formula from the expected-value section above.
- Decide. Grade if EV after cost exceeds raw price by 30 percent or more. Keep raw if the margin is thinner. Sell raw if you need fast cash.
This framework works for one card and scales cleanly to a 50-card submission stack. At scale, sort cards by expected-value ratio and submit only the top tier; the long tail is usually better off sold raw or kept raw.
Common misconceptions
- "Grading always adds value." Grading always adds information. Information adds value only when the market pays for it. On cheap cards, the fee exceeds the uplift.
- "If it's a star player, grade it." Star players have cards at every price point, from 5 dollar junk-wax inserts to 50,000 dollar rookie parallels. Name alone does not override the expected-value math.
- "A raw card is worth nothing without a grade." Raw comps exist for most cards. A clean raw card with a good photo set sells every day. The slab adds liquidity and removes authenticity risk; it does not conjure value from zero.
- "The grade I get will match the grade I assigned." Pre-grade self-scoring runs optimistic. Professional submitters hit their target grade about 60 to 70 percent of the time on pre-screened cards and far lower on un-screened cards.
- "Fees are the only cost." Time, shipping, insurance, and turnaround opportunity cost all matter. A 25 dollar fee is really a 35 to 50 dollar all-in cost.
- "The grader does not matter." The grader sets the premium. A PSA 10 and a CGC 10 on the same card can trade at different prices in the same week, and the gap moves over time.
How HCI supports the decision
On every card page we publish, you will see the raw sold comp and the graded comps in the public tiers (PSA 9, PSA 10, BGS 9.5 where available), dated, with volume indicators. That is the input data for the expected-value formula above. We do not score cards as "grade" or "do not grade" in public pages, because the decision depends on your specific cost structure, your grading hit rate, and your time horizon. Those are collector-specific variables, and we keep the analytical tools that fold them in behind the paywall, out of respect for the people paying for them.
What we do in public: surface honest sold data, date every price we cite, and name our sources. Our independence pledge spells out why we do not grade cards, run a marketplace, or push a pro service. It protects our signal. For how our data compares to the most commonly cited alternative, see our CardLadder alternative overview. And for the market context that should inform every grade-or-hold call in 2026, our K-shape report walks through where premiums are expanding and where they are compressing this year.