Team Rocket Returns Card Prices: All 111

Team Rocket Returns is the second most valuable EX-era set and the most lopsided one but for [Dragon Frontiers](/research/set-screen/dragon%20frontiers). 111 cards, 109 of them priced on September 23, 2026, summing to $10,931.31, and a single card carries 41.2% of it: Torchic ☆ at $4,500, which is tied with Lugia 149 from Aquapolis for the dearest English card on our tape. The two cards missing from that count are the set's other two Gold Stars, Mudkip ☆ and Treecko ☆, neither of which carried a reading on the census day. Underneath the Gold Stars, the set is thirteen Rocket's ex cards and a lot of 50-cent commons.

The nine bands hold all 109 priced cards. The $250 to $999.99 band is unusually crowded for a set this size, and the reason is a design decision rather than a market one: Team Rocket Returns printed eleven Rocket's ex cards at numbers 96 to 106, and nine of them sit in that band, alongside the Trainer card Here Comes Team Rocket!. The set's median card asks $7.67, which is the highest median of the sixteen EX-era sets and three times the era median of $2.48, so beneath the single chase card this is not a bulk-heavy checklist by the era's standards. Its cheapest cards are a Hoppip and a Remoraid at 50 cents each.

Ranks 1 to 20 are contiguous. Eleven of the top thirteen are the Rocket's ex subset; the other two are Torchic ☆ at the top and the Trainer card Here Comes Team Rocket! at rank 10. Two more English cards are named Rocket's something ex and neither is in this set: Rocket's Raikou ex was printed in [Deoxys](/research/set-screen/deoxys) and asks $284.99, and Rocket's Persian ex in [Unseen Forces](/research/set-screen/unseen%20forces) at $194.99, which makes thirteen in English and no more. Twenty-three further English cards are spelled Team Rocket's something ex, every one of them printed in 2025 or later, and they are a different family that a substring match will hand you anyway. That is the kind of thing a name search gets wrong and a name-and-number lookup does not.

63 of the 111 Team Rocket Returns cards carry a PSA 10 sold median and 66 carry a PSA 9 one, which makes it by a wide margin the best-documented set of the sixteen. The next deepest is [Sandstorm](/research/set-screen/sandstorm) at 30, and nine of the sixteen sets carry fewer than 25. That is not because Team Rocket Returns cards are dearer, because [Unseen Forces](/research/set-screen/unseen%20forces) is worth more and carries 27. It is because the Dark Pokémon subset of this set is cheap, plentiful and heavily graded, so the sold record reaches down into cards asking $3.

The table is the four deepest and the four thinnest of the sixteen by PSA 10 coverage, printed as two contiguous slices of one ordered list. The rightmost column runs the opposite way to the third, and the reason is selection rather than economics: a set where only eight cards have ever recorded a PSA 10 sale records them on its eight most expensive cards, where the multiple is largest. Team Rocket Returns has the lowest median multiple of the sixteen precisely because its record reaches down to the cheap end. That is the same selection effect our study 14,466 Cards With a Price and No Sale documents across the catalogue, and it is why a set-level PSA 10 multiple is a claim about coverage as much as about grading.

43 Team Rocket Returns cards carry 100 or more recorded PSA submissions, 26,502 between them, returning 2,362 tens. The median card in that group gems 8.8% of the time, close to the EX-era median of 9.4%. The extremes inside the set are wide: Torchic ☆ gems 1.30% of the time, 19 tens from 1,467 submissions, the lowest rate of any card in the era with a hundred submissions on file, while a Dark Magneton asking $3.56 gems 39.5% of the time and is the highest. The most expensive card in the era and the easiest card in the era to gem are in the same set, and they are at opposite ends of it. The whole-era treatment is EX Era PSA 10 Gem Rates; the cross-era one is Hard to Gem.

The headline is one card's share of one set on one day, and it has an obvious weakness: Torchic ☆ carries no recorded PSA 10 sold median, which means fewer than three verified PSA 10 sales are on file, so the $4,500 is an asking price with no sold ten behind it. If a census of sold prices put the card at half that, the set would be worth about $8,700 and the top share would fall to roughly 26%. The evidence pointing the other way is the PSA 9 record, ten verified sales at a median of $14,800, which is three times the raw ask for a grade below the one being asked for. A second break is the two missing Gold Stars. Treecko ☆ last read $2,400 on September 20; add that back and the set sums to $13,331.31 and Torchic's share falls to 33.8%. Mudkip ☆ has no reading anywhere in the 25-day tape, so nothing can be added back for it and the true denominator is higher again. A concentration figure computed over a checklist that is not fully listed overstates the top card, and by how much is bounded below at 7.4 percentage points. Third, the graded-coverage claim would break if a set we have not looked at carries more than 63 PSA 10 medians; the count was run over all sixteen and the full ordered list is in the scratch dump.

Two of the 111 cards carry no reading on the census day, and both are chase cards, so every share in this piece is computed over a denominator of $10,931.31 that is missing two of the set's three Gold Stars. That is stated in the falsifiability section with the alternative number beside it rather than buried. Raw prices are asks on September 23, 2026 and graded medians are sold prices over verified sales; sample sizes are printed beside every graded figure and several sit on three to six sales. Team Rocket Returns carries a recorded release date of November 1, 2004, and our daily tape begins on August 30, 2026, so nothing is windowed to its launch; its cards have no weekly series before August 30, 2026 either, so no return figure exists at any horizon. Team Rocket Returns has no `/set/` page because no recorded value series exists for it; the set screen resolves and was fetched to confirm it renders this set. Gem rates are registry counts, a sample of what people paid to grade rather than of what was printed. Nothing here is investment advice.

Population: all 111 `cards` rows with `lang = 'en'` and `set_name = 'Team Rocket Returns'`, deduplicated on `tcg_id`, which collapsed zero rows. Prices are `price_history.price` on day 20719, which is 2026-09-23 and the fullest of the tape's 25 days at 52,378 catalogue rows. 109 of the 111 carry a row on it; the two that do not are Mudkip ☆ 107 and Treecko ☆ 109, both named in the article, excluded from every price figure rather than zeroed, and their effect on the headline share is quantified under falsifiability from Treecko ☆'s last recorded reading of $2,400 on day 20716, which is 2026-09-20; Mudkip ☆ carries no reading at all on any of the tape's 25 days, so no add-back is possible for it and the stated 33.8% is an upper bound on the corrected share rather than the corrected share. `price_history` records each product's CANONICAL printing under the shared price rule in `packages/core`, deliberately not the dump's flagship figure, so a multi-printing product is quoted at the printing the market trades rather than at its dearest row. Summed ask is $10,931.31 over the 109. Share of set is the card's ask over that total. Medians take the upper middle value on even samples. Graded figures come from `graded_snapshots`, latest row per card and grade across both the catalogue-id and TCGplayer-id keyspaces with the fresher row taken, are SOLD medians over verified sales, and are printed with their sale counts; the cron writes a median only on three or more sales, so "not recorded" means fewer than three on file. A graded row is dropped when the fresher of the two keyspaces still predates the tape's first day, 20695. Across the 2,227 products of this ground that removes seven PSA 9 and PSA 10 medians on five cards, all seven in the catalogue-id keyspace: the PSA 10 and PSA 9 of Rocket's Scizor ex 101 at day 20655, which is 2026-07-21, the same pair for Rocket's Mewtwo ex 99 at day 20661, which is 2026-07-27, the PSA 10 of Metagross Star 113 at day 20655, and the PSA 9 medians of Gyarados H10 in Skyridge and Tyranitar H28 in Aquapolis at day 20675. Three of the seven are PSA 10 medians. None of the seven feeds a printed figure anywhere in this module; the five belonging to the three cards named in the callout above are quoted there only to show what was removed. The median PSA 10 multiple is PSA 10 median over raw ask, taken over the cards in each set carrying both a PSA 10 median and an ask of $5 or more. The $5 floor removes 26 of Team Rocket Returns' 63 rows, which is the largest removal of the sixteen sets and is the set's cheap graded tail showing up; the next largest is Sandstorm at 8 of 30 and eight of the sixteen lose nothing at all. The eight multiples in the table run over 37 cards of 63 in Team Rocket Returns, 22 of 30 in Sandstorm, 27 of 27 in Unseen Forces, 23 of 27 in Dragon, 12 of 14 in Dragon Frontiers, 11 of 11 in Ruby & Sapphire, 11 of 11 in Deoxys and 8 of 8 in Emerald. The graded-coverage table prints ranks 1 to 4 and 13 to 16 of one ordered list over all sixteen sets, ordered by PSA 10 coverage with ties broken by the larger card count. Gem rates come from `population_snapshots`, latest day per key across both keyspaces, and pass `gemRateIsTrustworthy()` from `@cardapp/core/server` with the recorded PSA 10 sold count supplied; the cohort is the 43 cards with 100 or more recorded submissions. The claim that Torchic ☆ has the lowest gem rate in the era was run across all 519 EX-era cards clearing that bar, not across this set. Every link was fetched from production on 2026-09-23 and its title read back. Scripts: apps/web/scripts/vex-build.mts, vex-sets.mts and vex-figures.mts; figures in apps/web/.scratch-vintage-ex/figures.json. Nothing here is investment advice.