The Hobbit, Three Weeks Later: Which Cards Actually Stuck
Standard's most-played Hobbit card is a two-mana common that sits in 120 sideboards and two main decks. Three weeks of DeckCortex decklist data across seven formats, August 11 to September 4: Warg Tactics in 122 of 1,265 Standard lists, Giant's Boulder locked into Pauper Tron at 9.9% since day one, and Bilbo, Thief in the Night in 16.2% of Legacy and 21.1% of Vintage lists, with a Legacy deck now named after him. The six cards that stuck, counted down to #1, plus the preview darlings that faded and the calls we got wrong ourselves.
Standard's most-played Hobbit card is a two-mana common that almost nobody casts in game one. Warg Tactics sits in 122 of the 1,265 Standard lists our corpus registered since the set went live on August 11, and in exactly two of them it's in the main deck. The other 120 copies live in the sideboard of Mono-Green Landfall, a deck that was already the biggest thing in the format a week before anyone noticed the card. Three weeks ago, in our 72-hour read, Warg Tactics didn't make the list. Last week it was in 67 of 355 registered lists, more than any other new card in any format that week.
That's the shape of this set's Constructed story: not the cards the preview season crowned, but the ones that turned up in decklists once the format's actual decks got around to asking what they needed. Below is the ranking of what stuck, counted down to the one card that has genuinely changed two formats, followed by the cards that faded and the calls we got wrong ourselves.
How we counted
Every number here is a decklist-appearance count in DeckCortex's own scraped tournament corpus: MTGO Challenges and Leagues plus paper events, canonical lists only, window August 11 to September 4, 2026. That is 24 days of online legality and 21 days of paper. The set's 186 new cards were checked; reprints were excluded so Settle the Wreckage couldn't sneak in as a "Hobbit card". Denominators are the lists registered in that window per format: Modern 2,362, Pauper 1,542, Legacy 1,317, Standard 1,265, Pioneer 723, Vintage 493, Duel Commander 459. Where a card grew or shrank, we split the window into week one (August 11 to 17) and the last week (August 28 to September 4). These are appearance shares, not field shares and not win rates. We don't track match results, and nobody should read "16% of lists" as "16% of the room".
cEDH is deliberately left out. That corpus is full of brew lists whose dates say nothing about when a card became legal, which is how a three-day-old card ended up "in" more cEDH lists than the corpus had lists. Casual Commander adoption is EDHREC's number, quoted as such further down.
6. The Queen of Dale, the pre-release favourite that found a smaller home
The Queen of Dale was our own best Constructed guess before release, and the honest scorecard is: right card, wrong formats. In Standard she's in 38 lists, but 20 of those are sideboard copies, spread over Boros Aggro, 4/5c Aggro and Boros Mobilize, brought in against decks that cast noncreature spells all day. A 2/1 that recruits every time an opponent casts their first noncreature spell in a turn is exactly what you want against a control deck's removal and counters, and exactly what you don't want maindeck against a creature mirror.
Where she's actually the headline is Duel Commander. She is the most-played new card in that corpus, 34 of 459 lists, ahead of Kíli the Resourceful at 20, and she's in 9 of the 10 Pioneer Weenie White lists we have. That is a real home, and a small one. The mythic price tag, $25 in Scryfall's paper aggregate on September 4, is Commander demand, not Standard demand.
5. Belladonna Took, real in three formats and growing in none
Belladonna Took is the card outside sources most wanted to call a Constructed surprise, and our data half agrees. She's in 33 Legacy lists (Boros Energy and Boros Aggro, one energy deck under two scraper labels), 21 Standard lists and 15 Modern lists. The mechanism is honest value: the first token entering each turn gains you a life, the second draws a card, the third puts a +1/+1 counter on your whole team, and the ability is capped there, so a token-flood deck gets one card and one anthem per turn, not a combo.
The problem is direction. Legacy went from 6 lists in week one to 3 in the last week. Modern went from 8 to 2. Only Standard rose, from 4 to 11, and almost all of that is one deck: Boros Mobilize, where she's in 12 of 14 lists. Real, useful, and not going anywhere bigger.
4. Key to the Side-Door, Modern's artifact deck found a draw engine
Modern's most-played Hobbit card isn't a creature. Key to the Side-Door is in 32 lists, and 18 of the 34 Cutter Affinity lists in the window run it. A one-mana artifact that taps to make a creature unblockable is a fine Affinity card on its own; the second ability, discard a legendary card with the same name as a legendary permanent you control to draw two, is what an artifact deck full of legendary four-ofs actually wants. The same deck runs Bilbo, Thief in the Night in 17 of those 34 lists, which is the first sign of a pattern we'll get to at number one.
Two smaller Modern pickups deserve a line. Kíli the Resourceful is in every one of the 11 Hammer Time lists we have, a 100% inclusion rate in a tiny sample, because once you're storied the first equip each turn costs , and a deck built on equipping cheaply draws a card off every Equipment that enters as well. And Giant's Boulder went from zero Modern lists in week one to 26 in the last week; Eldrazi Tron runs it in 14 of its 22 lists. That one is about to come up again.
3. Warg Tactics, the sideboard card that took over Standard
Here's the deck. Mono-Green Landfall existed before the set: it was the single biggest label in Standard's first Hobbit week, 46 of 480 lists, with none of the new cards in it. The core hasn't moved since, Llanowar Elves into Sazh's Chocobo and Icetill Explorer into Mightform Harmonizer, with fetchable lands doing the work. What changed is the fifteen cards next to it.
Warg Tactics does two things a mono-green deck couldn't do before at instant speed for two mana. Mode one destroys a creature with flying. Mode two puts a +1/+1 counter on your creature and gives it trample and hexproof until end of turn, which in this deck means you respond to removal on the Chocobo that Harmonizer just doubled, and it connects anyway. Across the five spellings the scraper uses for the deck (Mono Green Landfall, Mono-green Landfall, Mono G Landfall, Monogreen Landfall and plain Landfall, 248 lists in the one-month matrix), 107 lists run it. Its sister deck Mono Green Aggro adds 13 more.
It brought a friend. Gigantic Big Bear is a seven-mana 10/7 that can't be countered and has hexproof and haste, and it's in 78 Standard lists, all of them sideboard, 65 of them in the Landfall labels. Against 4/5C Control and the Doomsday decks, a green deck with Llanowar Elves and a landfall ramp plan boards in a threat that the counter-and-sweep plan simply can't interact with, and swings for ten the turn it lands. Week one: zero lists. Last week: 51 of 355. Elven Passage completes the package on the maindeck side, 40 lists and rising from zero: for one life, sacrifice it to fetch a basic tapped, behold an Elf, and the basic untaps, which is two landfall triggers and an untapped land in a deck that plays Llanowar Elves in every list.
The list that finished second at the MTGO Standard Challenge on August 30, 53 players, shows the whole package: Bear and Warg Tactics in the fifteen, next to Mossborn Hydra and Torpor Orb.
For scale, the Landfall labels together were 18.9% of the 465 Standard lists registered from August 25 onward, the biggest deck in the format, ahead of Izzet Spellementals at 11.0%. The Hobbit didn't build that deck. It built its sideboard.
2. Giant's Boulder, a common that was a staple on day one and never wobbled
Some cards need three weeks to find a home. Giant's Boulder needed three days. In our 72-hour read it was in 11 of 114 Pauper lists; three weeks later it's in 152 of 1,542, and the weekly share has sat between 8.9% and 9.4% the entire time. Inside the Tron labels it's not a choice anymore: 148 of the 175 Tron lists in the matrix run it, and the four smaller Tron variants (Pinger, Altar, Flicker, Temur Monster) are at 100%. Outside Tron it barely exists, eight lists.
The card reads like it was designed by someone who has played the deck. Scry 2 on entry digs for the third Urza's land. , for a mana of any color solves the deck's colored-splash problem with an artifact it can find with Ancient Stirrings. And once the lands are assembled, seven mana and a sacrifice destroys any permanent, which turns the mana rock into the deck's answer to whatever it couldn't answer before. It doesn't replace Expedition Map; the list below runs four of each.
That's the Urzatron list that made the top 8 of the MTGO Pauper Challenge on August 29 with 47 players. What outside coverage missed is how far the card has travelled. In Legacy it's in 96 lists, 7.3% of the format, and the labels tell you why: Trini Tron Karn 16 of 17, TES 10 of 12, Beseech the Storm 14 of 20, Artifacts Prison 8 of 11. Any deck that generates colorless mana in bulk and needs a colored splash or a late-game answer has picked it up. And in Modern, as above, it went from nothing to 26 lists in the last week on the back of Eldrazi Tron. A $0.30 common is the second-best Constructed card in the set, and it's not close to third.
1. Bilbo, Thief in the Night, the card two formats renamed a deck after
In our 72-hour dossier Bilbo was the anomaly: a Standard-legal mythic that was in 10 Legacy lists and one Standard list. Three weeks on, the anomaly is the format. Bilbo, Thief in the Night is in 214 of 1,317 Legacy lists, 16.2%, and the weekly share went from 10.6% in week one to 19.0% in the last week. In Vintage it's the only new Hobbit card in the corpus at all, in 104 of 493 lists, 21.1%, with the last week at 27.8%. No other card in the set is in double digits anywhere.
Read the text, because the reasons are all on it. A 2/2 with two abilities. The first: spells you cast from anywhere other than your hand cost less. The second: whenever Bilbo attacks, you may cast an artifact, instant or sorcery from your graveyard, and an instant or sorcery cast this way is exiled instead of going back. The reduction only touches generic mana, so a spell whose cost is all colored gets nothing. That matters for what the decks chose to pair him with.
Legacy built a deck around him and the scraper noticed. Our corpus now carries a label it didn't have before the set: "Dimir Bilbo", 31 lists, Bilbo in every one. It's the same Dimir Tempo shell (87% overlap in the top 40 cards), tuned for him: Flow State in 94% of the Dimir Bilbo lists, Mishra's Bauble in 84%, Daze in 94%, Thoughtseize and Force of Will in all of them. The engine is simple. Bauble is free from hand and free again from the graveyard on an attack, one extra card a turn. Flow State from the graveyard costs instead of , and with an instant and a sorcery already in the yard it puts two of the top three into your hand. Every cantrip you've already cast becomes a second cantrip the turn Bilbo swings, and a 2/2 with Daze and Force behind it swings a lot. Across the whole tempo cluster (Dimir Tempo, Ub Tempo, Dimir Bilbo, Dimir Delver: 217 lists) 125 run him, 57.6%; the Doomsday cluster is at 36 of 83.
The results followed. Dimir Bilbo won the Legacy Challenge 32 on August 28 (harmonywoods, three Bilbo, three Flow State, three Bauble) and again on August 30 (Slasher21, four Bilbo, four Flow State), where four of the top eight were the same deck. It finished second at the 50-player MTGO Challenge on September 1, and took first at a 22-player paper event in São Paulo the same weekend. Here's the August 28 winner:
Vintage found a different reason. There, Bilbo is in 21 of the 47 Dimir Lurrus lists, 15 of the 19 Esper Lurrus lists, and 18 of 40 Dimir Control. Lurrus of the Dream-Den casts one permanent with mana value two or less from the graveyard each turn; Bilbo makes each of those casts cost less, and since Bilbo is himself a two-mana permanent, Lurrus recasts him for whenever he dies. A recursive cost-reducer inside a companion deck built on recursion, with Force of Will to protect it. That's the Vintage interaction, and it's why the card's Vintage share is now higher than its Legacy share.
The list that this card will never make: Standard. Four lists in 1,265, none in the last week. The Izzet-Prowess-in-Standard prediction from preview season is dead, and it was wrong for a legible reason. Standard has no zero-cost artifact and no one-mana cantrip worth recasting from a graveyard every turn; Legacy and Vintage have a deck full of them.
The fades
Smaug the Magnificent was the preview season's red card. Smaug the Magnificent is a four-mana 4/3 flying haste dragon that makes a Treasure every upkeep and hits any target for your Treasure count when he attacks. Week one in Standard: 24 of 480 lists, 5.0%, the most-played new card that week. Last week: 6 of 355, 1.7%. The Boros Dragons and Mono Red Dragons shells that carried him (10 of 11 and 7 of 7 lists) haven't kept up with the format, and the $30 mythic price is Commander money.
Belladonna in Legacy and Modern, above: 6 to 3 and 8 to 2 weekly.
Bilbo as a commander. This is the split that tells you something about the card. On EDHREC (external, their count) Bilbo is built as a commander in 816 decks, ranked #1,791, while as a card in other people's 99 he's in roughly 6,700 decks. Our own pre-release commander ranking put him at #11 and called him "maybe the most underrated card" in the set. We were right about the card and wrong about the format. He's a Legacy and Vintage card that happens to be legendary.
The ones that never showed. Gollum, Riddle Master is in 17 Standard lists and 4 Pioneer lists, Gollum the Abandoned in 10 Standard lists and nothing else, The Lord of the Eagles in 8 Standard lists overall, 5 of them in week one and 1 in the last week. Inside Information is the one black card with a pulse, 21 Standard lists, 12 of them Golgari Control. None of these is a mistake to own in Commander. None of them is a Constructed card as of today.
The Commander column, and one thing we can't measure
The most-played new Hobbit card in Commander, per EDHREC on September 4, is Gleaming Splendor, 7.8% inclusion across about 40,300 decks, and at $44 it's the most expensive non-chase card in the set. Whenever an opponent draws their second card in a turn you make a Treasure, and for two target players each draw a card, so you can trigger it yourself. In our Duel Commander corpus it's in one list out of 459. That's not a contradiction; it's the same card seen from a four-player table and a two-player one, and we can only count the latter. Casual Commander adoption isn't something our tournament corpus can see, so the EDHREC figure stands alone.
The commander story from this set is still Thranduil, the Elvenking, EDHREC's most-built Hobbit commander at 5,488 decks as of August 23. If you're building him, the full Thranduil primer is here.
What we got wrong, and what to watch
Our pre-release review named The Queen of Dale as the best Constructed argument in the set and told Standard players to watch Settle the Wreckage. It didn't mention Bilbo, Warg Tactics or Giant's Boulder once. The 72-hour follow-up caught Bilbo and Boulder and missed Standard entirely, because the Standard story didn't exist yet. Three weeks is the shortest window in which a set's Constructed footprint stops being a preview-season argument and starts being a decklist count, and the lesson from this one is that the count and the argument shared exactly one card.
Three things to watch. Whether Bilbo's weekly Legacy share keeps climbing past 19% or the format starts boarding Surgical Extraction and graveyard hate against a 2/2. Whether Giant's Boulder in Modern Eldrazi Tron is a two-week spike or a third format. And whether the Mono-Green Landfall sideboard survives contact with the Reality Fracture release on October 2, which will reshuffle Standard before the Hobbit's second month is up.
Method: decklist-appearance counts in DeckCortex's scraped tournament corpus (MTGO Challenges and Leagues, paper events via MTGTop8 and Topdeck), canonical lists only, window August 11 to September 4, 2026, per-format n as listed above; weekly splits August 11–17 and August 28–September 4. Archetype attributions come from the scraper's labels, with same-deck spellings merged where the card overlap makes it obvious. Not field share, not win rate, not an external aggregator. Card wording validated against live Oracle text on September 4, 2026. Prices are Scryfall paper aggregates on the same date, not a live TCGplayer feed. EDHREC figures are theirs and labelled as such.