Every grading calculator on the internet shares one optimistic assumption: that submitting your card is, on average, worth doing. Our EV block computes that average from recorded registry outcomes, and for 94 of the 97 dossiers carrying one on our book as of August 28, 2026, the number comes out positive somewhere between barely and wildly. This is about the other three, and about the Wizards Black Star promo Eevee, the card that earned the first negative stamp our desk ever wrote.

31PSA 10s in existenceon a registry of 10,685 slabs
0.29%lifetime gem ratebreak-even needs 2.23%
-$3.97expected net per submissionat registry-average outcomes
83.34%of modeled submissions return rawthe modal outcome is nothing
The anatomy

How a card becomes structurally ungradeable

The EV arithmetic has three inputs: what the grades sell for, how often each lands, and the fee. Eevee fails it in the purest possible way. The prize is enormous, a $1,290.50 PSA 10 median on 48 comps, 24.87 times the $51.89 raw ask. But the registry says the prize essentially never arrives: 31 gems in 10,685 attempts, 0.29%, and across our recorded window of August 4 to August 11 the registry added 18 slabs and not one new ten. A 24.87x payout on sub-0.3% odds, minus a $20 fee, minus the 83% of outcomes that come back raw, prints minus $3.97 per submission. This is a handout-era league card that children played with in 1999, the same preservation story The Handout Grails documents, and the gem gate is not strict grading. It is history: the centered, clean copies mostly do not exist.

Note what negative EV does not mean. The card is not a bad asset; our house view treats the 31 existing tens as a functionally closed pool per the closed-registry playbook, and the smart seat, buying the $167.50 nine on the open market at 3.23x the ask, works fine. It means the SUBMISSION is a bad trade at average outcomes. The desk's rule from this card: when the registry-average gem rate sits below the break-even rate, you are not grading a card, you are buying a lottery ticket where the house keeps the stub. Grade the market, not the card.

Verified, card by card

The ranked list: the worst mail in the hobby

Eevee was the first negative stamp on the desk's hundred-card book, but sweeping all 97 dossiers carrying an EV block on that August 28 book turns up two more, and the deepest is a surprise: Kieran, a modern Twilight Masquerade trainer that gems a healthy 48.64% and STILL loses $7.58 per submission. Its failure mode is the opposite of Eevee's: the prize is tiny. The ten sells for $66.38, the nine for $15.50 against a $28.76 ask, so even the good outcomes barely cover the fee and the modal nine is a $13 haircut on the raw card. The 151 Pikachu illustration rare completes the trio at minus $1.58: a $141.50 ten over an $88.29 ask is only a 1.6x prize, and the strictest drift in its set keeps the gem odds under the 50.03% the arithmetic needs. Three cards, three different ways to lose: odds too thin, prize too small, spread too narrow.

The ten worst submissions on the live book, by expected net

CardEV / submissionGem oddsBreak-evenThe problem
Kieran, Twilight Masquerade-$7.5848.64%57.98%prize too small: $66.38 ten, $28.76 ask
Eevee, Wizards promos-$3.970.29%2.23%odds below break-even, 31 tens ever
Pikachu IR, 151-$1.5826.76%50.03%1.6x prize on a hardening gate
Hisuian Sneasler V, Astral Radiance+$1.6849.73%43.30%positive by a rounding error
Ethan's Adventure, Destined Rivals+$27.0342.05%22.53%thin cushion on a fresh registry
Charizard V, SWSH promos+$29.5456.31%27.53%90,489 pop; the ten is a commodity
Eevee, SVP promos+$34.2337.90%17.55%81,385 pop, industrial supply
Charizard VMAX, Champion's Path+$36.8969.56%41.47%gems easily, so tens carry no scarcity
Arceus VSTAR, Crown Zenith GG+$40.0556.85%32.59%$345.88 ten over a $217.13 ask
Entei, Wizards promos+$50.284.60%2.13%positive, but 58% of outcomes return raw

Read the bottom half carefully, because near-zero positive is its own warning. Entei prints +$50.28, technically fine, but the cushion rides entirely on a 4.6% gem event while 58% of modeled submissions come back raw; our eight-to-one study covers what happens to the impatient there. And the two promo giants, Charizard V and the SVP Eevee, are positive-EV mail with a different catch: their registries are so deep that every fresh ten joins a commodity pile, the exact dynamic The Population Growth Tax prices. Expected value is the entry ticket to a grading decision, not the decision. Our grading ROI guide carries the rest of the checklist.

EV figures are the recorded EV blocks from the 97 live report dossiers carrying one on our book as of August 28, 2026: three-outcome model (gem, PSA 9, raw) at registry-average rates with a $20 fee, raw outcomes valued at the ask. The ask-versus-sold caveat from our companion study applies to every raw figure and generally makes marginal trades worse, not better. Registry counts and gem rates are PSA registry snapshots; Eevee's zero-gem window covers August 4 to 11, 2026. Not financial advice.