The Weakest Sun & Moon Era Cards Rose 2.5% While the Rest of Their Sets Rose 7.7%. In Other Sets They Kept Up

Sets on the site now carry a list called Weakest in the set. It ranks the cards whose price did worst over the last 90 days, counting only cards that started that stretch at $5 or more, and you can open it for any set we cover, like [Lost Thunder](/research/set-screen/lost%20thunder?lens=laggards). People will read it as a bargain bin, so we tested that. We ran the same rule on six past dates, a quarter apart, from April 28, 2025 to July 27, 2026, and followed the three weakest cards in each set for the next 13 weeks. Our weekly prices only reach back to January 2025 for 5,041 of the 27,576 English cards we track, so this looks at the sets those cards come from. We left out promo, deck and trainer kit sets. A card counts on a date if it was $5 or more 13 weeks earlier and still has a price, in a set where at least 10 cards count. That gives 5,708 qualifying cards across the six dates, 1,244 different cards in 32 sets, 14 of them from Sun & Moon, and 516 picks. Every pick is measured against the other cards in its own set over the same 13 weeks, so a market that was rising anyway can't pass for a comeback.

One word before the numbers. A point here is a percentage point: a card up 3% in a set whose other cards rose 7% finished 4 points behind. The gaps we quote are worked out set by set and date by date, and then we take the middle one. That's why they don't equal the plain difference between the overall figures.

Put all 32 sets together and the weakest three rose a median 3.0% over the next 13 weeks while the rest of their sets rose 6.3%. Only 43.1% of them beat the typical card in their own set, where a card from the rest of the set does it 47.5% of the time. We then cut the picks five ways, by price, era, set age, how far behind they were and date, and the era cut is the one that stood out. Sun & Moon is 251 of the 508 picks we could measure, and most of the gap sits in it. Sun & Moon's gap is 5.6 points worse than the other sets', with a likely range of about 0.3 to 9.8 points. We found it after looking at the pooled result, so treat it as a pattern to watch.

In Sun & Moon sets the weakest finished under the rest at five of six dates. Outside Sun & Moon the two lines track each other, with the weakest ahead in the three quarters the sets fell and behind in the three they rose. The two groups had different years. The rest of the Sun & Moon sets went up in all six quarters, by 1.7% to 15.4%. The rest of the other sets fell in three of them, by 2.5%, 3.7% and 9.9%. Whether a set had been rising before the date, which you'd know at the time, doesn't explain the difference: in Sun & Moon sets that had been rising the weakest finished 4.3 points behind, and in other rising sets 1.4 points ahead. What the rest of the set did afterwards, which you wouldn't know, explains part of it. In the 139 times the rest of a set went up, the weakest finished 3.2 points behind and beat it 38.1% of the time. In the 33 times it went down, the weakest finished 2.0 points ahead and beat it 63.6% of the time. So the weakest cards tended to move less than their set in both directions, and Sun & Moon's other cards went up in 80 of the 84 times we checked. Even where the rest of the set rose, though, Sun & Moon's weakest did worse, 4.9 points behind against 2.0 for the other sets.

Half the Sun & Moon picks fell more than 5 points further behind their set. Outside Sun & Moon the weakest caught up about as often as they slipped, 37.0% against 35.4%. Umbreon GX from Sun & Moon is what the Sun & Moon pattern looks like. It was the weakest card in its set the week of July 28, 2025 after falling from $89.69 to $48.02, and it lost another 14.2%. It was on the list again the week of October 27, 2025 and lost 11.1% more. Umbreon V from Evolving Skies shows how wide the outcomes run elsewhere. It was second weakest of 59 cards the week of July 28, 2025, down from $290.57 to $221.71, and 13 weeks later it was $573.77, up 158.8% while the rest of the set rose 22.1%. It topped the list again the week of January 26, 2026 at $297.36 and rose 11.8% while the rest rose 16.5%.

The figures from here on are card by card: each pick against the middle of the rest of its own set on its own date. On that basis Sun & Moon's weakest finished 5.2 points behind. Price made it worse. Its weakest cards at $10 and up finished 7.4 points behind their set (likely range 4.2 to 8.7 behind), and under $10 they finished 2.6 behind (0.2 to 5.1 behind). Outside Sun & Moon, the $10-and-up picks finished 2.5 points behind, with a likely range from 5.3 behind to 0.9 ahead, and the under-$10 picks finished 3.8 points ahead (1.4 to 6.7 ahead). That last one shrinks to 0.1 behind if you buy a week after the date instead of on it, so part of it is a weekly price bouncing back.

Being on the list twice in a row was one of the worst signs we found, and it's also a Sun & Moon thing. A Sun & Moon card that was among the three weakest at one date and was still there a quarter later finished 8.6 points behind its set over the next 13 weeks (likely range 6.4 to 11.8 behind). Only 15.4% of those 39 picks beat their set, and 56.4% lost money outright. The 28 repeat picks from other sets finished 3.4 points ahead, with a likely range from 2.8 behind to 10.3 ahead. Repeats come up more than chance would give you: 69 of the 392 picks we could follow were on the list again the next quarter, 17.6%, against 11.8% if each set's list were drawn at random.

All five are trainer cards and four are from Sun & Moon sets. The four Sun & Moon cards had 16 chances to catch up and didn't beat their set once. Nessa, the Sword & Shield card, beat hers twice. A card wasn't on the list the whole time between its first and last appearance, so each row shows where the card went, not what holding it from one list to the next would have returned.

How far behind a card already was didn't change much. Picks that trailed their set by 10 to 20 points finished 2.0 points further behind, and picks that trailed by 20 to 40 finished 2.9 behind. The 40 picks that trailed by under 10 points and the 30 that trailed by more than 40 both came out near even, with wide ranges.

In sets less than a year old, the weakest beat the rest of their set by 4.8 points over the next 13 weeks, on just 36 picks, 21 of them from the four Mega Evolution sets in the study. The week of January 26, 2026, the typical card in Mega Evolution was down 29.6% over 13 weeks and Mega Lucario ex 179 was down 44.6%, from $265.75 to $147.35. It then rose 30.3% while the rest of the set rose 8.6%.

Eleven of the twelve could be measured, and four beat their set: Vaporeon Star, Latias & Latios GX, Giratina V the second time and Charizard GX. Vaporeon Star's win is one recorded jump. Its price sat at $650 in June and July 2025, read $1,999.99 from the week of July 28, 2025, and sat around $450 to $457 from February 2026, so we wouldn't read that +239.7% as a real move. Latias & Latios GX is jumpy too: its weekly price moved 20% or more in 5 of 31 weeks in 2026, once by 42%. Jolteon Star, Kyogre-EX and Groudon-EX held exactly the same price for all 13 weeks. Pricey cards often sit still like that. In The Frozen Catalog, 51.5% of cards above $500 held one price from August 30 to September 19, 2026, so check recent sold listings before you read much into an expensive card's place on the list.

Pooled, the gap is small and points the same way however it's run. Taken era by era, it's clear in Sun & Moon and absent elsewhere. Sun & Moon also holds with a one-week delay, at 3.8 points behind (likely range 1.2 to 6.7 behind), and the other sets land at 0.3 behind (2.1 behind to 2.9 ahead). The delay matters because one low weekly price can put a card on the list and then bounce back by itself, and buying a week later takes that bounce out.

Our 52,041-card study of whether rising cards keep rising found that cards down more than 10% over 90 days beat the average card by 7.36 points over the next 90. It replaced an earlier, smaller study that had found rising cards kept rising. In our pool, cards down more than 10% did 5.1 points worse than the average card on the same date, on the mean. Measured against their own set, the same 590 cases came out even, a median gap of 0.0 points. Their shortfall came from the sets they were in, and 197 of the 590 were in sets whose other cards fell too. That study used 52,041 cards with no $5 floor and this one uses 1,244 cards at $5 and up in 32 sets, so the two shouldn't be read as one result.

Never Buy the Small Dip measured how far a card sits below its own high, and found that cards 30% or more under their high for five weeks or longer did better than the rest. This list compares a card with its set over 13 weeks. The median pick here was down 9.6% over the 13 weeks that put it on the list while the rest of its set rose 10.0%. The Pitch Black piece followed one new set's fallers over thirty days. This one covers 32 sets over 18 months.

From April 2025 to October 2026, a Sun & Moon card on Weakest in the set usually kept losing ground to its set: about 5 points a quarter card by card, and about 7 at $10 and up. Sun & Moon cards on the list a second quarter running did worse, 8.6 points behind, and more than half of them lost money. Those were also quarters when the rest of Sun & Moon was going up, and the weakest cards in any set tended to trail a rising set, by 4.9 points in Sun & Moon and 2.0 elsewhere. So if you want one of these cards, the past says there was no rush to buy it before it recovered. In the other 18 sets the list didn't tell you much on average: the weakest kept pace with their set and 51.4% beat it, but that average mixes quarters when the set rose and the weakest trailed with quarters when it fell and the weakest held up better. In sets less than a year old, like Mega Evolution, the weakest pulled ahead by 4.8 points, on just 36 picks.

Source: the per-card weekly price series this site stores for each TCGplayer product (the `ppthist:` records), which holds the market price our provider recorded for the product, weekly before the last 60 days and daily within them, merged with our own readings. 5,041 of the 27,576 English series reach back to January 27, 2025. The pull was made once on November 15, 2026 at 12:07 UTC, by primary key, and every figure here was computed offline from that copy. Catalog: 28,054 English catalog rows carry a TCGplayer product id. 325 rows claim an id already claimed by another English row and were collapsed to the lower card id, leaving 27,729 products, of which 27,576 have a series. Sets: the catalog's own set name. 71 catalog sets were left out: every promo set, every trainer kit, every McDonald's Collection, the three Trick or Trade bundles, the three Battle Academy sets, World Championship Decks, Deck Exclusives, League & Championship Cards, Blister Exclusives, Jumbo Cards, First Partner Pack, First Partner Collection 2026, the two Classic Collections and the two Shiny Vault subsets. That leaves 21,344 products in 155 sets, and putting the 71 back is one of the runs in the table. Weekly price: the last reading in each Monday to Sunday week from January 27, 2025, or, if that week has none, the last reading in the two weeks before it, so a price dated to a week can be any day of that week. Dates: the weeks of April 28, 2025, July 28, 2025, October 27, 2025, January 26, 2026, April 27, 2026 and July 27, 2026, 13 weeks apart, so no two forward windows overlap. Entry on a date: a weekly price 13 weeks earlier of $5.00 or more (the list's own floor, MOVE_FLOOR in set-screen.ts) and a weekly price that week, which gave 5,887 card-dates. 85 were dropped because the 13 weeks before the date started before the set's release (release dates from our provider's set list), and 3 were dropped at the July 27, 2026 date because their series steps by more than 25% across September 7, 2026, when our pricing rule changed. No series was cut short. Sets with fewer than 10 entered cards on a date were dropped, 31 set-dates and 91 card-dates, leaving 5,708 card-dates on 1,244 cards in 172 set-dates across 32 sets. The list: each set's cards on each date ranked by their move over the 13 weeks before it, weakest first, ties broken by card id, and the first three taken, 516 picks on 385 cards. Forward move: the weekly price 13 weeks after the date over the price that week. The one-week-delay run uses weeks 1 and 14 after. Missing forward prices: 8 picks and 45 other card-dates, 52 products in all, 46 card-dates whose series ended before the forward week and 7 with a gap at it. Every product was looked up in our daily price table, and all 52 have daily prices from August 30, 2026 through November 15, 2026 (77 days for 49 of them, 76 for two, 52 for Flareon Star). They are left out, with both bounds in the callout. Points are percentage points. Gap: for each set and date, the median forward move of its three weakest minus the median of its other cards, and the published figure is the median of those gaps, so it is not the difference of the pooled medians. The card-by-card figures (price, era and price, repeats, set age, how far behind) are each pick's forward move minus the median of the rest of its own set on its own date, and the share that beat their set counts picks above that median. Comparing a pick only with cards from its own set in its own price range (under $10, $10 to $25, $25 to $100, $100 and up) needs three or more such cards, which 449 of the 508 measured picks have. Medians are upper-middle on even counts, one convention for every cell. Likely ranges are 95% intervals from 5,000 bootstrap draws that resample whole sets, every date of a set together, with a fixed seed. The Sun & Moon against other sets difference resamples sets within each side. Price groups use the price the week of the date. Era is read from the card id (sm Sun & Moon, swsh Sword & Shield, sv Scarlet & Violet, me Mega Evolution, xy and g1 XY, bw Black & White, ex EX). Black Bolt cards whose catalog id carries no era code are counted as Scarlet & Violet. No set mixes eras, so the era runs are whole sets. The era cut came out of five cuts made after the pooled result (price, era, set age, how far behind, date). Whether a set had risen beforehand uses the median move of all its entered cards over the 13 weeks before the date; whether the rest rose afterwards uses the median forward move of its non-picks, which is not known on the date. Set age is the date minus the release date. Second quarter on the list: a pick that was also among its set's three weakest at the previous date. First time counts only dates from July 28, 2025, which have a previous date. The chance rate for a repeat is the average, over the 392 picks still qualifying a quarter later, of 3 divided by the number of qualifying cards in the set at that next date. The doubled-or-halved run drops any card-date whose weekly price more than doubles or falls below half from one available week to the next anywhere from 13 weeks before to 14 weeks after the date, 37 card-dates. The $400 table lists every pick priced $400 or more the week of its date, with no row left out. Cost floors: 14.94% is 1 divided by 0.87 minus 1, fees and postage on a sale. 24.14% is 1.08 divided by 0.87 minus 1, adding an 8% gap for buying at the ask. They are the same two floors Never Buy the Small Dip uses. The dataset and the scripts that produced every number here are kept for an independent recompute.