The pop report is stale in both directions
Recent PSA submissions grade above the printed rate on 2026 sets and far below it on everything older. Across 597 Pokemon cards, the gap tracks the card's age - and the single lifetime number hides both directions. Post this →
Last month we published a finding: across 2,107 Pokémon cards, the median lifetime gem rate was 29.1% while the median marginal gem rate — what the last 30 days of submissions actually returned — was 15.4%. Recent submissions were gemming at roughly half the rate printed on the pop report.
That was true. It was also incomplete, in a way that matters.
Run the same calculation weighted by submission volume instead of taking the median across cards, and the sign flips. Across the 597 Pokémon cards with at least 50 submissions in the window, the lifetime rate is 40.2% and the marginal rate is 49.8%. Recent submissions are gemming better.
Both numbers are correct. They answer different questions, and the gap between them is the actual finding.
The direction depends on how old the card is
| Era | Cards | Recent subs | Lifetime | Last 30 days | Change |
|---|---|---|---|---|---|
| Mega Evolution (2026) | 123 | 126,687 | 57.4% | 65.5% | +8.1 pt |
| Scarlet & Violet, late (2025–26) | 70 | 17,855 | 41.6% | 40.7% | −0.9 pt |
| Scarlet & Violet, early (2023–25) | 132 | 29,869 | 31.1% | 21.3% | −9.8 pt |
| Sword & Shield (2020–23) | 107 | 20,534 | 46.1% | 34.8% | −11.3 pt |
| Pre-2020 | 137 | 16,512 | 32.6% | 20.7% | −11.9 pt |
The gradient is monotonic. The newest sets grade above their printed rate; every older era grades below it, and the gap widens the further back you go. On pre-2020 cards, 131 of 137 are grading worse right now than their pop report implies.
The weighted average flips positive for one reason: Mega Evolution is 123 of 597 cards but 126,687 of 222,914 recent submissions — 57% of all volume. The newest, best-grading material dominates the pool. Take the median across cards and you see the typical card. Weight by submissions and you see the typical submission. Those are different animals, and quoting either one without saying which is how a real effect gets argued into nonsense.
Two mechanisms, pulling opposite ways
On new cards, supply is pack-fresh. Mega Evolution product is months old. The copies entering PSA came out of sealed product recently, were handled once, and went straight into a submission. There is no accumulated damage and no decade of shifting standards to average against. A 2026 set's lifetime rate is itself mostly recent submissions, so the two numbers converge — and where volume is still accelerating, the marginal rate runs ahead.
On old cards, the good copies are already gone. This is the selection effect, and it is the stronger of the two. Every clean Base Set copy that was worth grading has been graded. What is left entering PSA in 2026 is the residue — the copies that sat in binders, that someone passed on twice, that are being submitted now because prices rose rather than because the card is clean. The lifetime rate on a 1999 card is an average that includes the 2004 submissions of cards pulled from fresh boxes. Nothing being sent today resembles those.
A third factor sits underneath both: grading standards are not fixed across two decades. We can measure the outcome, not decompose it. We are not claiming to know how much of the old-card decline is condition selection versus tightened grading, and anyone who tells you they can separate those from public pop data is guessing.
The individual movers
Cards with the largest gap between lifetime and last-30-days, minimum 50 recent submissions:
| Card | Set | Lifetime | Last 30 days | Recent subs |
|---|---|---|---|---|
| Flying Pikachu | Evolutions | 33.4% | 2.2% | 92 |
| Chi-Yu ex | Paldea Evolved | 53.4% | 20.7% | 58 |
| Mr. Mime | 151 | 33.9% | 4.8% | 83 |
| Snorlax | Paldean Fates | 41.3% | 12.9% | 70 |
| Morty's Conviction | Temporal Forces | 39.7% | 12.0% | 75 |
| Charmander | Paldean Fates | 46.2% | 19.7% | 66 |
| Charizard ex | Paldean Fates | 40.2% | 16.4% | 532 |
| Gardevoir ex | Paldean Fates | 42.8% | 18.8% | 319 |
| Charizard ex (promo) | SVP | 46.4% | 22.3% | 202 |
Charizard ex from Paldean Fates is the one to sit with. It has 40,736 PSA submissions lifetime and 532 in the last 30 days — this is not a small-sample artifact. Its pop report says four in ten come back a 10. The last month says fewer than two.
Flying Pikachu is the most extreme number on the list and the one to trust least: 2.2% of 92 submissions is two cards. The rate is real, the precision is not.
How to check this
Every number here comes from PSA's own public population reports. The only thing we do that a pop report doesn't is store yesterday's copy. A gem rate is pop10 / total_pop; a marginal gem rate is (pop10_now − pop10_then) / (total_now − total_then) over a 30-day window. That is the whole method. The reason almost nobody publishes it is not that the math is hard — it is that you cannot backfill a delta you never recorded.
Per-card pages, with the raw counts behind both numbers, are at cashemapp.com/grading-odds. Coverage is English sets only. Cards need at least 20 recent submissions to get a page at all, which is why plenty of vintage you might want is missing — the cards where this effect is largest are often the ones with too little recent volume to measure.
What we are not saying
We are not saying the pop report is wrong. It answers "how has this card graded, ever," accurately. That is a reasonable question. It is just not the question you are asking when you are deciding whether to put a card in an envelope this week.
We are also not saying recent rates predict your card. A gem rate is a base rate for a population of submissions, not a forecast for the specific copy in your hand — condition still does the work. What the marginal rate buys you is a better prior than a number averaged over a decade of other people's cards.
Data: PSA population reports, 597 Pokémon cards with ≥50 submissions in a trailing 30-day window, as of 3 September 2026. Method and per-card figures: cashemapp.com/grading-odds.