Which Pokémon Sets Are Mostly Bulk? 129 Ranked

The belief is that Pokémon sets keep getting more worthless, year after year, and that today's are the worst. Half of that is right. We ranked 129 English sets by the share of their priced cards asking under a dollar on September 22, 2026, and the most bulk-heavy set is a 2025 one: 157 of the 190 cards we price in Journey Together ask under a dollar, which is 82.6%. But the ranking does not sort by year. Second place is Obsidian Flames from 2023 at 82.2%, fifth is Emerging Powers from 2011 at 78.6%, and twelfth is Flashfire from 2014 at 74.3%. Grouped by era, the share climbed from 22.7% to 69.5% over the seventeen years to 2016 and then stopped: the five era medians from 2014 onward all sit between 63.2% and 72.5%. What kept moving is the price of the bulk, not the amount of it. The median card in a 2014 to 2016 set asks 40 cents; in a 2025 or 2026 set it asks 19 cents.

Every figure here is one day of the daily card tape: `price_history` day 20718, which is September 22, 2026. A row is an English catalogue card carrying a TCGplayer product id, collapsed to one row per product id, because 240 English catalogue rows share an id with another row and counting both would inflate every share. A set enters when it carries a recorded release date and holds 60 or more such products. That gives 129 sets and 18,698 products, from 21,416 priced English products in all. Prices are TCGplayer market prices for the flagship printing, which are ASKS, not sales. A card asking 16 cents is a card nobody has bid on, and the gap between an ask and a sale is the subject of Ask Versus Sold.

The 60-product floor admits three annexes that are printed inside another set rather than sold as their own release: Hidden Fates Shiny Vault, Shining Fates Shiny Vault and Crown Zenith Galarian Gallery. All three are at the bottom of the ranking with no card under a dollar, which is what an all-holo annex looks like. They are kept in the table and marked rather than deleted, because a selection that removes exactly the rows that would look odd is how a ranking lies. The Set Inside the Set is the study of those annexes on their own terms.

Read the last two columns against the third. The share of a set that is sub-dollar has been flat for a decade. The price of that sub-dollar half has not been: the median card fell from 40 cents to 19 cents over the same period, and the share of a set asking under 25 cents went from 25.0% to 60.8%. The four zeros in the lowest-set column come from Skyridge and Aquapolis in the first row and from the three annexes named above, one each in the 2017-2019, 2020-2022 and 2023-2025 rows. The last row has no zero in it: the lowest of the seven Mega Evolution era sets is 57.8%. That column is not a useful summary of an era on its own.

Eight of the fourteen were released before 2022 and four before 2017. The last column is the one worth staring at. Those 157 sub-dollar cards in Journey Together are 82.6% of the checklist and 6.5% of what the set is worth. In Fusion Strike the 228 of them are 1.9%. The exception is Emerging Powers, where the sub-dollar cards are 26.5% of the set's value, and the reason is not that its commons are dear. It is that the whole set is worth $127 at ask, and $33.73 of that sits in cards asking under a dollar. Bulk becomes a large share of a set's value only when the set has no top end left.

Skyridge and Aquapolis are the only two full expansions in the universe with no card under a dollar at all: 363 priced cards between them and a floor above it on every one. The four unlinked names in this table, Team Magma vs Team Aqua, Delta Species, Sandstorm and Team Rocket Returns, are EX-era sets we hold prices for and no set page, so they are named rather than linked. Three of the five sets at ranks 120 to 124, Gym Heroes, Neo Destiny and Gym Challenge, were requoted on September 7, 2026, and their position needs the caveat in the next section before anyone quotes it.

First objection: the vintage numbers moved recently and you are quoting the new ones. True, and it matters more than anything else here. On September 7, 2026 the price feed changed which printing of each Wizards-era vintage card it quotes, and it did so for nine sets on one day. Seven of the nine had every single constituent reprice between day 20702 and day 20703; the other two, Neo Revelation and Neo Destiny, had 97.0% and 96.5% of theirs. Comparing day 20702 with day 20718 over the same fixed product ids, Jungle went from 0 of 64 cards under a dollar to 16, Fossil from 1 of 62 to 19, Team Rocket from 0 of 83 to 16, Gym Heroes from 1 of 132 to 16, Gym Challenge from 0 of 132 to 5, Neo Genesis from 2 of 111 to 26, Neo Discovery from 1 of 75 to 17, Neo Revelation from 1 of 66 to 15 and Neo Destiny from 1 of 113 to 13. The event is documented in How Concentrated Is the Pokémon Card Market, and it is why Bulk's Hidden Floor, measured on August 28, put the WotC-era sub-dollar share at 4.3% where this piece, measured on September 22, puts the pooled share over the 1999 to 2003 row at 14.3%, 244 of its 1,707 cards. Neither number is a mistake. They are two different printings of the same cards. All nine requoted sets sit in the 1999 to 2003 row and no set outside it repriced that way, so that row's median is a statement about which printing we quote as much as about the market.

Second objection: a dollar is an arbitrary line and you picked it to get an answer. The 25-cent column in the era table is the same measurement at a quarter of the threshold and it tells a different and sharper story, so the finding is not a threshold artefact: at $1 the eras since 2014 are flat, at 25 cents they are not. Both columns are printed rather than only the one that suits the headline. The under-$5 share, which we computed and did not print as a column because it saturates, runs from 84.5% to 95.9% across the top fourteen.

Third objection: your catalogue is missing cards, and modern sets look bulky because you price more of theirs. The ranking is a share, not a count, so a set with 284 priced cards and one with 98 are on the same scale. The relevant check is whether coverage is thinner on old sets, and it is not: we price 101 of Base Set's 102 printed cards, 132 of Gym Heroes' 132 and 178 in Skyridge against a 144-card printed checklist plus its holo and crystal numbering. What the universe does leave out is 2,718 priced English products, the difference between 21,416 and 18,698: 1,971 of them sit in the 53 catalogue sets carrying no recorded release date, the six largest of which are all Black Star Promo sets, and the other 747 sit in 25 dated sets holding fewer than 60 priced products, mostly trainer kits and starter decks. Those 2,718 are not a hidden pile of cheap moderns. Only 29.5% of them ask under a dollar against 54.7% of the universe, so admitting them would pull the pooled bulk share down, not up. If anything the modern risk runs the other way, because a set still in print has cards that have not been listed yet. Does Set Size Change What a Pokémon Set Is Worth measures the size effect directly and finds checklist size correlating -0.402 with the median card price, which is the same effect seen from another angle.

The flat-since-2014 finding dies if the Mega Evolution era's median climbs above 80% as more of its sets are released. It sits at 70.2% on seven sets today, and one era above 80% against four in the 63% to 73% band would be a resumed climb rather than a plateau. It also dies if the same measurement run on a different price source puts the XY era well below 69.5%: this is one feed's flagship-printing ask, and a source that quotes reverse holos or a different default printing would move the whole curve. The sharpest test is the cheapest to run, and we ran it: repeat the era table on day 20702, before the September 7 requote, and on day 20718 after it. The 1999 to 2003 row moves from 1.3% to 22.7%. Every other row moves by 2.2 points or less and five of the eight do not move at all. If a future requote moves a row other than the first one by more than a couple of points, the instrument is less stable than this study assumes.

As of `price_history` day 20718, which is 2026-09-22. Population: `cards` rows with lang = 'en' and a non-empty tcg_id, collapsed to one row per tcg_id, preferring a catalogue-native id over a synthetic `enx-` one and then the lexically first id; 240 rows collapse away, leaving 21,452 distinct product ids of which 21,416 carry a price on day 20718. A set enters the universe with a recorded release date and 60 or more of those products: 129 sets, 18,698 products. Release dates come from `report:spread:setsmeta` joined on `normalizeCardIdentity.setKey` from src/server/db/card-identity.ts, then on the same key after stripping one leading era code from the PROVIDER name only, then through a hand-audited alias table of 27 entries; stripping is never applied to catalogue names, so 'Expedition Base Set' cannot become 'Base Set', and its alias to the provider's 'Expedition' is written out in the script. 53 catalogue sets remain undated. All 53 are promo sets, trainer kits, McDonald's collections, POP series or one-off products; six of them do hold 60 or more priced products, all six being Black Star Promo sets, so they are excluded by the release-date rule and not by size. Prices are TCGplayer market prices for the flagship printing, which are asks. Bulk is defined as an ask strictly under $1.00. Medians take the upper-middle value on even samples. The ranking is over all 129 sets and is printed as two contiguous slices, ranks 1 to 14 and ranks 116 to 129; the full ordered list is in the scratch dump. Ties in the share are broken by the higher median card price ranking as the less bulky set, which is what separates ranks 125 to 129, all at 0.0%. Era is a half-open range of release dates, cut at 2003-06-01, 2007-05-01, 2011-01-01, 2014-01-01, 2017-01-01, 2020-01-01, 2023-01-01 and 2025-09-01; the nine rows partition all 129 sets with no gap and no overlap, and the script asserts that they sum to 129. An earlier draft cut on fractional years with gaps between the bounds and lost three sets, Skyridge, Diamond & Pearl and Mega Evolution, which is why the sum is asserted rather than assumed. The era columns are medians across sets, not pooled shares. Both pooled shares quoted in the sceptic section are taken over the era rows themselves, not over calendar years: the 1999 to 2003 row pools 244 of 1,707 cards, 14.3%, and the 2023 to 2026 rows pool 3,288 of 4,878, 67.4%. The September 7 requote figures compare day 20702 with day 20718 over the same fixed list of product ids per set, so no card enters or leaves across the comparison. The requoted sets were identified from the step itself rather than assumed: on day 20702 to day 20703 every constituent of Jungle, Fossil, Team Rocket, Gym Heroes, Gym Challenge, Neo Genesis and Neo Discovery repriced, along with 97.0% of Neo Revelation and 96.5% of Neo Destiny, and no other set in the universe combined a move on that scale with a change in its sub-dollar count. On the three neighbouring day pairs we checked, no set at all repriced in full. Scripts: apps/web/scripts/rs-build.mts, rs-figures.mts and rs-extra.mts; every printed number with its query is in apps/web/.scratch-reddit-sets/figures.json. Nothing here is investment advice.