What Completing a Set Actually Costs, and Why It Is Two Purchases

Ask what a set costs and you get a number. Ask what the number is made of and the answer stops being a number and becomes a decision. On September 2, 2026 our daily price dump carried 51,843 priced rows across 647 set buckets, summing to $909,227.42 if you bought exactly one of everything. Across the 295 buckets carrying at least 60 priced rows, the median set costs $1,224 to complete, and the median set's ten most expensive cards are 68.8% of that. The other ninety-odd percent of the checklist, all of it, has a median cost of $345.89. But the sharpest thing in the data is not that split. It is that the checklists where every single card costs real money are the ones that were never sold in packs.

Rank the 647 buckets by the median card, the honest measure of what an average purchase off a checklist costs, and the entire top of the board is promos: Nintendo Promos at a $59.55 median card across 75, WOTC Promo at $51.65 across 69, Diamond and Pearl Promos at $37.80 across 60, Black and White Promos at $32.08 across 128, XY Promos at $22.45 across 241. For comparison, the median card in the median modern booster set is nineteen cents.

The mechanism is supply, and it is the cleanest natural experiment in the catalog. A booster set's cheap cards are cheap because millions of them fell out of packs into a market that cannot absorb them. A promo bin has no packs. Every card in it entered the world through a tournament, a box topper, a magazine, or a retail tie-in, in quantities set by an event rather than by a print schedule, and so the bin has no bulk floor to sit on. That is why a promo checklist costs $22 to $60 a card all the way down while a booster checklist costs under a dollar a card all the way down, and it is why promo bins spread their value evenly (32.0% top-ten share for XY Promos, 23.8% for League and Championship Cards) instead of concentrating it in a chase. If you want a completion project whose every purchase holds value, the data points at the bins nobody markets. If you want a cheap completion project, it points at exactly the sets that ship in boxes.

Lost Origin and Phantasmal Flames are trophy sets. In Lost Origin one card, the Giratina V alternate art at $822.95, is 59.9% of a $1,374 bill, and the entire remaining 229 priced rows cost $187.00 together, an average of 85 cents each. That 59.9% needs one qualification stated where it can be seen rather than in a footnote: the dump files Lost Origin's trainer gallery as a SEPARATE bucket of 30 cards worth $594.08, and a collector completing the set as sold would buy both. On the combined 260-card checklist the Giratina is 41.8% of a $1,968 bill. Either number is defensible; what is not defensible is quoting the higher one without saying which checklist it describes. The completion decision in a set like this is still a single-card decision wearing a checklist costume.

Skyridge is the opposite and it is expensive for the opposite reason. Its top card is only 6.4% of the bill and its top ten are only 38.4%, but the remaining 168 cards cost $11,638.81, an average of $69.28 each. Nothing in Skyridge is a trophy purchase and everything in Skyridge is expensive, which is what a closed print run from 2003 does to a checklist over twenty-three years. The same shape appears in Aquapolis (top-10 share 44.7%, $44.24 per remaining card) and Expedition (45.0%, $34.96). If you are budgeting a vintage e-Card set, the top-ten line is nearly useless to you and the per-remaining-card line is the whole forecast.

Temporal Forces and Paradox Rift show that the modern era prints both shapes. Their top cards, a $95.55 Gastly and a $118.92 Groudon, are 7.6% and 9.7% of their bills, the flattest modern checklists in the 295, and their remaining cards run $3.10 and $2.70 apiece. These are genuinely completable sets: a patient buyer working a want list finishes them, and no single purchase dominates. The trade-off is that nothing in them ever becomes a store of value either, which is the same fact stated from the other side.

For the whole dump: the 25th percentile row is $0.30, the median is $1.50, the 75th is $7.23, the 90th is $30.99 and the 99th is $299.99. The single most expensive priced row is $4,500.00. The top 10 rows hold 3.6% of total value, the top 100 hold 15.2%, the top 1,000 hold 47.7% and the top 5,000 hold 79.7%. Meanwhile 22,650 rows, 43.7% of everything priced, cost less than a dollar, and all of them together are 0.84% of the money. This is the fact that makes a completion bill readable: nearly half of any checklist is free in every sense that matters, and the bill is decided by a handful of cards at the other end.

One caution about what a bucket contains. 7,744 of the 51,843 priced rows, carrying $143,241 of the total, do not resolve to a card in our catalog: they are code cards, sealed accessories, deck products and mis-filed rows that the upstream dump prices anyway. Evolving Skies' 253 rows include 16 of them, chiefly Code Cards. Two buckets are made of them entirely: League and Championship Cards (598 rows, $29,911) and World Championship Decks (1,917 rows, $7,379) are the two largest checklists in the dump and neither is a set anyone can collect card by card. Bucket totals are what the dump prices, not what a checklist contains.

Source: moonstone's own daily price-history table, the free CSV dump we accrete rather than re-buy, read at day bucket 20698 (September 2, 2026). That day carried 51,843 rows with a price across 647 set buckets, summing to $909,227.42. The dump's per-row price is the FLAGSHIP printing's market price, not the canonical printing price the card pages show, so a card quoted here can read higher than the same card's page; the two instruments are named rather than mixed, and every figure in this piece comes from the dump. Set buckets are the dump's own lowercased set names, so English and Japanese sets appear as separate buckets, trainer galleries appear as separate buckets from their parent sets, and promo bins appear as sets; concentration and bill figures are computed over the rows actually priced in each bucket, not over a published checklist. 7,744 rows do not resolve to a card in our catalog and are counted in bucket totals as the dump supplies them. Distributional cuts (top-10 share, rest-of-checklist bill, cost per remaining card) are restricted to the 295 buckets with at least 60 priced rows; medians quoted for that group are medians across buckets, not across cards. The table holds four day buckets, August 30 to September 2, 2026, but only three distinct snapshots: the September 1 bucket is byte-identical to August 31 on all 51,827 rows it shares, so the feed did not update that day. Three snapshots is far too short to describe a trend and no movement figures are reported from it. Nothing here is financial advice.