What a reprint actually does to card prices

475 reprints against matched cards left alone: Commander precons nearly halved the median older printing within six months, while Secret Lairs left it above where it started.

Aymeric Lamontagne

Nov 10, 2026

A reprint is supposed to crater the copy you already own. Measured across 475 reprints released between September 2024 and April 2026, the median older printing stood 8% below its pre-reveal price six months after release, and 40% of those printings were above where they started.

That median is close to useless, because it blends different events. When the new copy came in a Commander precon, the median older printing lost almost half its price in six months: index 52, where 100 is its price about a hundred days before release. When the new copy came in a Secret Lair, the median older printing ended up higher, at index 112. The finding is not that reprints hurt, or that they do not. It is that the vehicle, and what the new copy sells for, explain far more than the word “reprint” does.

A comparison group changes the size of those numbers, though not the order. Matched cards that were not reprinted rose 20% over the same six months, which turns the 8% fall into a 25% gap. How much of that gap belongs to the reprint depends on which cards you compare against, so this piece shows the range rather than the most dramatic figure. It covers the rule, the overall result, the split by vehicle, the mechanism, the timing, how many printings recover, and how far to trust any of it.

Older printing, 6 months after release

91.6 — All vehicles; 40% stood at 100 or above

Commander precon reprints

52.4 — n = 107; 6% stood at 100 or above

Secret Lair reprints

112.0 — n = 154; 65% stood at 100 or above

Matched cards, not reprinted

120.0 — Five per reprint, on the same dates

All figures are medians over the 475 reprints in the sample, read from Manamancer’s price archive through November 2, 2026. Index 100 is the older printing’s median price from 120 to 90 days before release.

The rule, fixed before the results

An event study is only as honest as its sample, so the rule was written down before any outcome was computed:

1. The event. A new English paper printing of a card, in a set released between September 5, 2024 and May 6, 2026, dated by the release date in our card index. At least one copy in that set must exist in nonfoil, which removes foil-only oddities. Promos, The List, memorabilia, tokens, Alchemy, joke sets and basic lands do not count.
2. The older printing. An English nonfoil paper printing released at least a year before the event, excluding promos, memorabilia and The List. Where a card has several, we follow the cheapest at baseline, because that is the copy a player would otherwise buy.
3. The baseline. The median TCGplayer market price from 120 to 90 days before release, with at least 20 days of prices. It must be at least $5.00.
4. Clean events only. No other qualifying reprint of the card within 180 days on either side, and the card is not banned in Commander today.
5. The control. For each reprint, the five nearest cards with no new paper printing, promos included, within a year on either side. They are not on the Reserved List, they share a rarity tier, and their baseline is within 25%. “Nearest” weighs price, EDHREC rank, printing age and the price change over the months before baseline. Each control is read on the same calendar dates as its reprint.
6. The measure. A printing’s index is its median price within three days of a date, divided by its baseline. A relative index divides that by the median of its five controls. Everything below is a median with its spread, and 95% intervals come from resampling whole sets, or whole release dates within a vehicle.
7. The split. A vehicle gets a row only with at least 30 events.

The window holds 7,324 card-and-set pairs that meet the first rule, brand-new cards included. Of those, 3,985 had an older printing with enough baseline days, and 553 cleared $5. Three fell to the ban rule (Griselbrand, Dockside Extortionist and Primeval Titan), 73 to overlapping reprints and two for lack of five controls. That leaves 475 events: 433 cards across 25 set codes and 59 separate set-and-date releases, from September 27, 2024 to April 27, 2026. The median baseline was $10.41, with quartiles at $7.10 and $17.49.

Three things changed after the rule was written, and they should be on the record. Before any outcome was computed, an “other” bucket was split into bonus sheets, Eternal sets and Foundations Jumpstart. After the first results, intervals within a vehicle moved to resampling by release date, since every Secret Lair drop shares one or two set codes. And comparisons were added, because the first control looked too strong: the lower rows of the table under “How far to trust this”, plus a benchmark and a raw-price version of the recovery test, flagged where it appears.

The archive holds a price for every day from February 8, 2024 through July 2026. It thins after that, to 13 days in September and 12 in October 2026, with no gap longer than seven days. A date with no price within three days widens its window to seven.

The overall result

Each column is its own median, and the relative index is the median of each event’s own ratio, so the columns do not divide exactly.

The spread is wide enough to matter. At six months, the middle half of older printings ran from 63.0 to 114.8. Some 35% had lost more than a quarter of their baseline price and 15% had lost more than half, while 40% stood at or above it. The middle half of the relative index ran from 53.4 to 95.7, and the 95% interval on its median is about 57 to 84. Four in five sat below their controls.

The controls deserve suspicion of their own. A 20% rise in six months is not “the market”. These are popular cards, matched on EDHREC rank, that nobody reprinted for two years. Every card with a floor of $5 or more, leaving out only Commander bans and the Reserved List, rose about 2% at six months and 15% at a year on the same dates. The real comparison sits somewhere between those lines, which is why the headline here is the range and not one figure.

The vehicle decides

Six months after release, by vehicle:

Commander precons are the cleanest result in the study. The fall appears in all seven products in the sample, from a median of 40.9 for Aetherdrift Commander (n = 11) to 74.2 for Edge of Eternities Commander (n = 8). No comparison group rescues it.

Masters-style sets clear the bar on count and not on breadth. Of the 32 events, 25 are Innistrad Remastered (median 43.7) and seven are Universes Within (77.8). This is close to one product, so it carries no interval and should be read as one.

Eternal sets have the same weakness: 30 of the 34 events are Avatar: The Last Airbender Eternal. Bonus sheets are broader, spread across Special Guests, the Mystical Archive and four similar sheets, and they held roughly level. The Secrets of Strixhaven Mystical Archive was the weak one, at 74.3 (n = 17).

Secret Lairs ran across 30 separate release dates, and none of the 154 had lost half its price at six months. Whether they lost anything at all depends on the comparison, as shown below.

Main sets do not get a row. Only 23 events from five sets survived the rule, which misses the bar we set. All 23 sat below their controls at six months, so the direction is not in doubt, but the size would be an anecdote with a decimal point. Foundations Jumpstart, at 13, misses it too. New-printing rarity was meant to be a split as well, and only six events were reprinted at common or uncommon, so there is nothing to report.

Why: the price of the new copy

A player who wants the card to play it is choosing between copies. The older printing can only hold a price well above the new one if someone will pay for the difference. So the number to watch is what the new copy sells for, and it lines up with the vehicles almost exactly. Here is the cheapest new copy at six months, as a share of the older printing’s baseline:

A $10 card whose new copy settles near $2.60 has a substitute that resets its price. A $10 card whose new copy settles near $11 has no substitute for anyone trying to save money, so the older printing has nothing to fall toward. The older printing rarely falls all the way, either. Six months on, the median Commander-precon older printing still sold at 1.57 times the new copy, with a middle half from 1.08 to 4.40 times. In Secret Lairs the ratio was 0.97: the old card was the cheaper of the two.

The new copy’s price only 30 days after release already sorts the outcome, which is the part that lets you read the next case:

The rank correlation between the two is 0.61 across the sample, and it holds up within a vehicle: 0.35 inside Commander precons, 0.32 inside Secret Lairs and 0.44 inside bonus sheets. Two cautions. The new copy’s price is itself a result of the reprint, not an independent cause, and it stands in for supply, which we cannot measure: the archive holds no print runs.

The same logic explains what happens to premium printings. Following the most expensive qualifying older printing instead of the cheapest gives a different and larger sample (n = 764, because a card now qualifies if any older printing clears $5). There the median index at six months is 96.5, and 95.4 for Commander precons (n = 229) against 52.4 for the cheapest copy. The reprint mostly hits the copy that was competing to be the cheap one.

To illustrate without hand-picking, these are the three events in each group closest to that group’s six-month median, among those with a baseline of at least $10 and a clean full year:

Each later price is a median over the seven days around that date, and the Krark-Clan year-on read stops at the archive’s last day, November 2, 2026.

When it happens: mostly after release

The reveal is the obvious place to expect the fall. In this sample, most of the movement comes later. At release, the median relative index was 97.2 (95% interval about 94 to 100). Even for Commander precons it was only 93.8, about 6 points of an eventual fall of more than 50. Their older printings were still at 106 in raw terms on release day, then 97 four weeks later, 85 at eight weeks, 68 at twelve and 53 at six months. The halfway point of that decline came two to three months after the product was out. Our Hobbit piece found old printings barely moving during previews and warned that the softening had not had time to appear. This study suggests that warning was the right one to give.

Reveal dates are the weak spot, because the index only has a release date for every card. Scryfall’s data carries a preview date on some printings, and 124 of the 475 events (26%) have one. Their median lead was 21.5 days before release, with quartiles at 17 and 24, and 92% came less than 90 days before release, after the baseline had closed. Masters-style sets are the exception at 65%, because Innistrad Remastered’s dated previews began in late October 2024, as its baseline window closed.

Around those dated previews (n = 113, leaving out previews too close to baseline), the relative index read 99.7 two weeks before, 98.8 on the day, 96.1 two weeks after and 93.7 thirty days after. That is a slide, not a cliff, and for most of these cards the last read already falls after release. What no date in our data can show is an effect that arrives earlier, through a product announcement or a leak. If a card fell before day −120, its baseline is already lower and this study misses that part.

How many come back, and how fast

For this we need a clean full year, which leaves 257 events and no Eternal sets. In raw prices, against the baseline:

The benchmark row matters, because prices are noisy and a card can dip 10% with no reprint in sight. It takes each reprint’s nearest control and applies the same test. Of the non-reprinted cards that dipped, 64% were back at or above baseline a year on. Of the reprinted ones, 20% were.

That table is a raw-price version, added because it is easier to read. The test fixed in advance was relative: 210 of the 257 (82%) fell at least 10% behind their controls within six months, and 20 of those (10%) were back within 5% of them at one year. For the benchmark, the same figures were 147 of 241 (61%) and 60 (41%). Either way, the gap between reprinted and not reprinted is the finding, not the absolute count.

The speed is unimpressive where it happens at all. The median low came around day 154. Printings that did get back took a median of 196 days from release, measured on the two-week grid, against 154 for the benchmark. No Commander-precon older printing in this group got back inside six months, and the seven that did so within the year took a median of 224 days.

A partial bounce is more common than a recovery. For the same 57 Commander events, the median index went from 54.4 at six months to 61.0 at one year, while their controls went from 115.5 to 154.0, so the relative index ended the year at 38.5. That 154 is the strongest control reading in the study; without rank matching it is 130.3, and the relative index 48.6. Secret Lairs rose in raw terms over the same stretch, from 107.2 to 114.7, and their relative index was 84.0 (95% interval about 74 to 92). In this window, a reprinted card’s lost ground mostly stayed lost, and the market around it did not wait.

How far to trust this

The control is the softest part of the design, so here is the six-month relative index under each comparison, most tightly matched first:

Read across and the conclusions separate cleanly. Commander precons land between 47 and 54 on every line, so that result does not depend on the control. Masters-style sets stay well below 100. Secret Lairs run from 89 to 101 here and read 84 on the later-reprint check below, so the dent is somewhere from nothing to about 15%, and we cannot say which. Across all vehicles, the gap is 16% to 25% here and 28% on that check.

Four other checks:

• A fake reprint date. Moving each event 200 days earlier, with nothing being reprinted yet (n = 371), the relative index 90 days later was 96.5, with a 95% interval of about 91 to 101. Events from the same period on their real dates read 85.1. So the fake shows no clear effect, but its median is not quite 100, and 3 or 4 points of the real gap could be softness that was already there.
• Controls posing as reprints. Matching non-reprinted cards to other non-reprinted cards gave 100.9 at 90 days and 99.0 at 180. The procedure does not manufacture a gap by itself.
• Cards reprinted later. Using only controls whose own reprint comes more than 180 days after the event (n = 437) gave 71.9 at six months, with Commander precons at 45.3 and Secret Lairs at 84.3, slightly lower than the main result. Being chosen for a reprint does not appear to be what the control is missing.
• Different samples. At a $2 floor (n = 770), the six-month index was 89.8 with a relative index of 74.7. At $20 (n = 88), 98.6 and 80.7. Adding the overlapping reprints back (n = 548), 89.1 and 74.2. Extending the window to June 2024, which adds Modern Horizons 3 Commander and Bloomburrow Commander among others (n = 567, with no pre-trend match for the early events), 86.2 and 73.3, with Commander precons at 51.5 and 46.7 (n = 141).

The groups were balanced on what could be measured: a median pre-baseline trend of +0.5% against +0.7%, and a baseline of $10.41 against $10.31. They were less close on popularity (EDHREC rank 1,388 against 1,968) and printing age (2,044 days against 1,809).

Now the limits, of which there are six.

First, dates. Release dates are real, and reveal dates are partial, as above. For a Secret Lair in particular, the index date may not be the day copies reached the singles market, and we hold no shipping data.

Second, selection. Reprints are not assigned at random. A publisher’s choice of what to reprint can track things we did not match on, and the matching covers only what the data shows.

Third, look-ahead in the control. EDHREC rank is today’s rank, not the rank at baseline, and the no-reprint filter looks a year forward. Both can favour cards that went on to do well, which is why the table above drops them one at a time.

Fourth, survivorship. Attrition inside the sample is tiny: two of 477 events fell out, and every remaining one has a six-month read. But the archive only follows printings still in today’s index, and the one-year rows are a different mix of 257 events, with no Eternal sets and proportionally more Innistrad Remastered.

Fifth, bans and other shocks. The index knows a card’s legality today, not when it changed, so a ban or unban inside a window cannot be dated. Stricter versions barely move the result: also dropping cards banned in Modern, Pioneer, Legacy or Standard today gives 76.4 (n = 455), and dropping today’s Game Changers gives 75.1 (n = 449). A new combo, a popular new commander or a set’s mechanics can still move any single card.

Sixth, the price itself. TCGplayer describes its market price as “compiled from recent sales completed on TCGplayer.com”. So it lags a fast move, it is thin for quiet cards, and it reflects one US marketplace. This is nonfoil only, and 19 months of releases in one market is not a law of nature.

What would change this reading is a larger main-set sample and a reliable reveal date for every reprint; we have neither yet. Until then, the summary fits in a sentence. When a reprint is announced, the useful question is not whether the card got reprinted, but in what, and at what price the new copy settles.

If you want to see what a product is about to reprint, the precon catalogue has the lists. And may your oldest copies stay safely in their sleeves, where no chart can reach them.

Browse the Commander precons — Each product’s decklist, so the reprints are there to read.