Modern Before RC Baltimore: The Deck to Beat Is Two Decks, Ten Lists Apart

Modern · by Andreas Graw · September 10, 2026

One Ephemerate on a reanimated Atraxa and the Angel stays forever — that rules trick took Esper Goryo's from 2.7% of Modern lists in February to the top of the format, while Boros Energy fell from 19.5% in March to 4.9% in August. But the post-Hobbit month is a photo finish: Izzet Prowess 237 published lists, Esper Goryo's 227, out of 2,554 — and Prowess just won both Hobbit Spotlights with two cards Standard banned in August. Goryo's beats Prowess (55%), Esper Blink beats both, Boros beats Blink. Here's the Modern that 1,333 players are registering for RC Baltimore, on a corpus we had to de-duplicate first.

Modern Before RC Baltimore: The Deck to Beat Is Two Decks, Ten Lists Apart

As of September 10, 2026. Decklists for the Regional Championship at SCG CON Baltimore lock at 6 PM on Friday, 1,333 of the 2,500 seats are already claimed, and the format those players are registering looks nothing like the one we described in July. Back then Boros Energy was the deck to beat at 12% of our lists. In the thirty days since The Hobbit went live on Magic Online, the energy deck is 3.8% of the Modern lists we track, ninth in the format. What replaced it isn't one deck. It's two: over those same thirty days, Izzet Prowess has 237 published lists and Esper Goryo's has 227, out of 2,554. Stretch the window to sixty days and Goryo's leads; look at the trophies and Prowess leads; look at the match data and neither of them is favored against the other's hunters. The honest headline for Baltimore is that Modern has a photo finish where it used to have a king, and the deck that actually beats both finalists is sitting quietly in fourth.

First, the number that changes every Modern table you'll read this week

One methodology note, and it's not the usual boilerplate. While merging archetype labels for this piece we found that every MTGO Challenge sits in our corpus twice: once from mtgo.com, which publishes the top 32 under its own archetype names, and once from MTGTop8, which republishes the same event under different names. "Esper Goryo" and "Instant Reanimator" aren't two decks — they're the same lists arriving through two feeds, and the same pilot shows up under both labels on the same weekend. Because our dedup hashes the archetype label into the deck's identity, the twins never collided, and the Challenge top finishers were counted twice. Over the sixty days before Baltimore that was 1,207 of 6,335 rows.

So for this article we removed the republished twins and merged the feed-split labels (Esper Goryo + Instant Reanimator + Goryo's Vengeance; UR Cutter Prowess + Izzet Prowess + Ur Prowess + UR Aggro; Boros Energy + Boros Aggro; Devoted Combo + Creatures Toolbox; and so on). The numbers below are on that cleaned basis: 5,128 published Modern lists from July 12 through September 10 across 295 events, of which 147 were paper events, and 2,554 lists from August 11 through September 10, the post-Hobbit month. The ranks barely move between the raw and cleaned bases; the decimals do, and the July figures we published were a touch too generous to whoever was winning Challenges that month. Consider that corrected.

How Boros Energy lost the crown

Boros Energy didn't get banned out of the format. It got out-positioned. In our monthly data the deck peaked at 19.5% of Modern lists in March, took the Phlage ban on May 18 and slid to 12.7% in May, was still 9.7% in July, and then fell to 4.9% in August. That's not a wobble; the deck lost three quarters of its share in five months, and the last third of that fall came with no ban at all.

The Pro Tour in Amsterdam did it. Boros Energy and Izzet Affinity were the two most-registered Modern decks there and, in Frank Karsten's words on magic.gg, both "collapsed in popularity" after "winning merely around 45% of their matches." The match data since hasn't been kinder: on mtgdecks.net's 30-day matrix the energy deck sits at 51% overall (847 matches, their data, not ours) but loses to both of the new top decks, 46% against Esper Reanimator and 46% against Izzet Prowess. A fair midrange aggro deck that's an underdog to the two most-played decks in the room is not a deck to beat, whatever its pedigree.

The cards tell the same story from the inside. Guide of Souls, Ocelot Pride, Ragavan, Nimble Pilferer and Galvanic Discharge are still in effectively every Boros list. What changed is the top end: The Legend of Roku // Avatar Roku, the Saga we called the Phlage replacement in June when it was in 87% of the deck's lists, is now in 14% of them. Pilots cut the four-mana value card and went back to being a one-drop deck with Lightning Bolt (up to 52% of lists) and a maindeck Solitude creeping in. It's a leaner deck than it was in July. It is also, by our conversion metric, the worst-converting deck among the format's top ten — 14.9% of its published full-standings finishes reached a top cut in the last month, against 33.7% for Esper Blink. The energy deck still wins small paper events on Sunday afternoons (18 wins in our sixty-day sample, 14 of them at local events), and only four of those came after the Hobbit release.

Esper Goryo's: the deck that keeps Atraxa

The statistical leader over the full sixty days is Esper Goryo's at 8.5% of lists (434 of 5,128), and in the August calendar month alone it was 9.2%, up from 2.7% in February. Karsten's performance-weighted winner's metagame on magic.gg had it at 11% for the two weeks after the Hobbit release, the format's most prominent deck in his words, and then, in the same breath: "that's a relatively modest number for the format's most prominent deck." Our decklist-appearance share reads lower than his winner's share by construction, and it agrees with him on the shape: a leader, not a tyrant.

The reason this deck went from fringe to first is a rules interaction, and it's worth being exact about it. Goryo's Vengeance reads: "Return target legendary creature card from your graveyard to the battlefield. That creature gains haste. Exile it at the beginning of the next end step." The target is Atraxa, Grand Unifier — a 7/7 with flying, vigilance, deathtouch and lifelink whose enter trigger reveals the top ten cards of your library and lets you take one card of each card type. That alone is a two-mana turn-three play that hits for seven and draws something like five cards. The trick is what happens next. Ephemerate exiles Atraxa and returns it to the battlefield. The Atraxa that comes back is a new object; the delayed "exile it" from Goryo's Vengeance was attached to the old one and does nothing. So you keep the Angel, you get the ten-card reveal a second time, and because Ephemerate rebounds, you get to blink it a third time on your next upkeep. Griselbrand, the other legendary target, does the same thing with "pay 7 life: draw seven cards" and lifelink to buy the life back.

Getting the legend into the graveyard is the deck's other half, and it's where the fair cards live. Psychic Frog discards Atraxa to put a +1/+1 counter on itself and draws when it connects. Faithful Mending gains 2, draws two, discards two, and comes back with flashback. Every list also carries the full Esper interaction suite: four Solitude (a flashed-in exile effect that Ephemerate happily doubles), Thoughtseize, Force of Negation, Prismatic Ending. That's why the deck is an all-rounder: if the combo isn't there, it's a control deck with a two-mana reanimation spell as its finisher.

And then there's the card that changed the deck this summer. Quantum Riddler is in effectively every Esper Goryo's list and is the archetype's fastest-rising card in our data (trend slope +13.5, the biggest of any card in the deck). It's a five-mana 4/6 flier that draws a card on entering and, with one or fewer cards in hand, turns every draw into a draw-plus-one — but the line that matters is Warp : cast it for two, take the draw, let it leave at end of turn, cast it again later from exile. Ephemerate on a warped Riddler returns it as a new object with no obligation to leave, so the same one-mana spell that keeps Atraxa keeps the Sphinx and draws again. It's the reason the deck doesn't run out of gas when the graveyard plan gets hated out.

Here's the list winning right now — jussupinator's August 30 Modern Challenge trophy, the pilot's second Challenge win in two weeks after two more top-4s:

We ran it through the DeckCortex deck engine, and the result is the useful kind of boring. Zero complete two-card combos: Goryo's plus Atraxa is a value line, not a loop, and the engine correctly refuses to call it infinite. The core/flex split is nearly total — Thoughtseize, Ephemerate, Force of Negation, Prismatic Ending, the Esper mana base, Atraxa, Faithful Mending, Goryo's Vengeance, Psychic Frog, Griselbrand and Solitude all lock at 0.998 or higher, Quantum Riddler at 0.907. The only flex slots the engine finds are Teferi, Time Raveler, Consign to Memory and a Hedge Maze. In other words: the seventy-five is settled, and what you're going to face in Baltimore is this deck, give or take a Teferi.

Its matchup profile is the reason it's the leader and the reason it's beatable. On mtgdecks.net's post-Hobbit matrix (30 days, more than 22,000 matches, their data) Esper Reanimator wins 55% against Izzet Prowess over 221 matches — the single most important number in this article, because it says the volume deck loses to the other volume deck. But it's a dog to the decks that ignore its combo and punish its mana: 43% against Esper Blink, 39% against Devoted Druid, 34% against Eldrazi Tron, 31% against Dimir Frog. And in Dallas, at the biggest paper event of the month, it was the second-most-played deck on Day 1 and put nobody in the Top 8.

Izzet Prowess: Standard banned its engine, Modern crowned it

Izzet Prowess is the other half of the photo finish, and it's the half with the trophies. In the post-Hobbit month it's the most-played deck in our data at 9.3% (237 lists), it has more event wins than any other deck in the sixty-day window (24 of 258, 16 of them in paper), and it won both Magic Spotlight: The Hobbit main events — Jefferson Lee in Brisbane on the last weekend of August, 570 players in our data, and Roy Varney in Dallas on September 5, 921 players, a 15-3 record from the number-eight seed on tiebreakers, beating Liam Kane's Tameshi Belcher in the final. Star City Games' own coverage put Prowess at 10% of the Dallas Day 1 field, the most of any deck.

The shell hasn't changed since Pioneer banned its best card and Modern didn't: Cori-Steel Cutter makes a 1/1 Monk with prowess every turn you cast your second spell and attaches itself for free, Dragon's Rage Channeler becomes a 3/3 flier once you have four card types in the graveyard, Monastery Swiftspear and Slickshot Show-Off carry the damage, and Lava Dart, Mishra's Bauble, Lightning Bolt and Mutagenic Growth are the cheap spells that turn all of that on. The engine analysis of Varney's winning list finds no combo, a core locked from Cutter and Lava Dart at 1.000 down through Expressive Iteration at 0.941, and exactly three flex slots — and those three are the story.

Two of them are Stormchaser's Talent and Boomerang Basics. On August 10, Wizards banned Stormchaser's Talent in Standard, naming its interaction with Boomerang Basics as the problem. Both cards are legal in Modern, and four weeks after that announcement Varney won the largest Modern event of the year so far with two Talents and four Basics in the maindeck. Talent is a one-mana Class that makes a 1/1 blue-red Otter with prowess on entering and, for , returns an instant or sorcery from your graveyard to your hand. Boomerang Basics is a one-mana sorcery — "Return target nonland permanent to its owner's hand. If you controlled that permanent, draw a card" — which in this deck is a second spell for Flurry that either bounces a blocker, resets an opposing Overlord, or bounces your own Cutter to draw a card and replay it. In our data the Talent is in 28% of the deck's lists and rising faster than any other card in the archetype; the mtgo.com feed even files the Talent build under its own label, where both cards sit at 92%. The engine flags both as rising flex, and we'd expect the Baltimore lists to run them.

Now the cold water. Prowess is the most-played deck on mtgdecks.net by matches — 2,355 in thirty days — and it's 49% overall, a hair under even. It beats the decks that stumble early: 63% against Devoted Druid, 60% against Ruby Storm, 54% against Boros Energy. It loses to everything that gains life or gums up the board: 45% against Esper Reanimator, 42% against Esper Blink, 40% against Izzet Metalcraft, 32% against Eldrazi Tron. This is the same profile we flagged in July — a popular deck a good player can plan to beat — with one difference: it's now winning the big paper events anyway, because in a 900-player room the fastest consistent deck gets to play against a lot of people who aren't ready for it.

The three decks that came from nowhere

If Goryo's and Prowess are the two names on the marquee, the format's actual movement in the last thirty days happened underneath them. Three decks that were rounding errors in July are now top-five archetypes.

Esper Blink went from 1.8% of our lists in July to 6.3% in August and 8.0% in the post-Hobbit month, and it's the deck with the best win rate of any of the big five on mtgdecks.net: 55% overall, 57% against Esper Reanimator, 58% against Izzet Prowess, 68% against Devoted Druid. It beats both finalists. The engine is the same Ephemerate the Goryo's deck plays, pointed at different targets. Phelia, Exuberant Shepherd exiles another nonland permanent when it attacks and returns it at end step — your own Solitude for another exile, your own Witch Enchanter // Witch-Blessed Meadow to destroy another artifact. The card that turned the deck into a real one is Overlord of the Balemurk, in 100% of lists: cast it for its impending cost of and it enters as an enchantment with five time counters, not yet a creature, and its enter trigger mills four cards and lets you return a non-Avatar creature card or a planeswalker card from your graveyard to your hand; blink it with Ephemerate or Flickerwisp and it comes back with no time counters, a 5/5 creature right away, and the mill-and-return trigger fires again. Two mana, a 5/5, and two Overlord triggers, on turn three. Colten Fowler won a 115-player ReCQ at SCG CON Dallas with it the day before Varney's title:

The hole is red. Blink is 16% against Boros Wildfire — the land-destruction deck — and 38% against Boros Energy on that same matrix, which is the strongest argument left for the energy deck in Baltimore.

Devoted Druid is the cleanest "from nowhere" story in the data: 0.7% of lists in July, 6.8% in August, and 10 event wins in sixty days. Devoted Druid taps for and untaps by putting a -1/-1 counter on itself; Vizier of Remedies makes that counter "that many minus one," which is zero, so the Druid untaps as often as you like for unlimited green mana. What made the old combo a real deck in 2026 is the mana doublers around it: Leyline of Abundance — free on the battlefield from your opening hand — and Badgermole Cub both add an additional whenever you tap a creature for mana, and Shang-Chi, Master of Kung Fu and Tyvar, Jubilant Brawler let the Druid activate the turn it enters. With effectively every list running four each of Green Sun's Zenith and Nature's Rhythm to find the pieces, the payoff is whatever you like: Tyvar, the Pummeler pumping the team by its own power, a Walking Ballista for X, a hasty Craterhoof Behemoth. The deck's mtgdecks profile is the mirror image of Prowess — 54% overall, beats Esper Reanimator (61%) and Eldrazi Ramp (73%), loses to Esper Blink (32%) and Izzet Prowess (37%). John Wager's NRG $10k win in Milwaukee is the reference list:

Broodscale is the deck the two headliners have been standing on all summer: 1.0% of lists last December, 6.4% in June, 7.3% in August, second in the sixty-day window at 8.2%, second in event wins with 21, and 53% overall on mtgdecks.net. The loop is Basking Broodscale wearing Blade of the Bloodchief: any +1/+1 counter on the Broodscale makes a 0/1 Eldrazi Spawn, sacrificing the Spawn for is a creature dying, which puts a +1/+1 counter on the equipped Broodscale, which makes another Spawn. Unlimited colorless mana and unlimited Spawn entering the battlefield, and Glaring Fleshraker deals 1 damage to each opponent for every one of them. The card gaining fastest in the archetype is Sowing Mycospawn, now in 90% of lists (trend slope +15.1): a land off the top when you cast it, and an opposing land exiled when you kick it, in a deck that already runs four Kozilek's Command. Broodscale is 47% against Prowess and 50% against Reanimator — an even deck against both, with a real combo, which is why it keeps its 8%.

The matchup triangle for Baltimore

Put the two data sets side by side and the tournament writes itself. Our share data says the room will be Prowess and Goryo's at roughly 9% each, Broodscale and Blink at 8%, Devoted Druid around 5%, then a long flat tail — Eldrazi Ramp, Dimir Midrange, Affinity, Boros Energy and Eldrazi Tron all between 3.6% and 4.6%. The matchup data says:

Baltimore's structure rewards exactly this kind of reading. Nine rounds of Swiss on Saturday with eighteen match points to make Sunday, six more rounds and a Top 8 for a $100,000 prize pool, $20,000 to the winner along with invitations to the Pro Tour and the World Championship. The Pro Tour in question is Pro Tour Nauctis at MagicCon Detroit, February 26–28, 2027 — Nauctis Draft plus Modern — so the Baltimore result doesn't just pay out, it decides who's testing Modern again in February.

One more thing the data can't decide for you: Dallas. Esper Goryo's was the second-most-played deck there and missed the Top 8 entirely. That's one event, a 921-player one, and it's the reason we won't call Goryo's the deck to beat on its own. If it converts in Baltimore, the sixty-day share and the August numbers were telling the truth. If it misses again, Karsten's "relatively modest number" was the more important half of his sentence.

The clock, and the set that didn't matter

Wizards left Modern untouched on August 10 — the announcement banned three cards in Standard, including the Stormchaser's Talent that Varney just won with, and opened its Modern paragraph with "Innovation is brewing in Modern," pointing at the Amsterdam decks that "incorporated new tools into existing shells." The next Banned & Restricted announcement is October 12, 2026: after Baltimore, but before the later round of Modern Regional Championships Karsten says will run through September and October, all feeding the same Pro Tour. Nothing in our data argues for a Modern ban. The top deck is under 10% of lists, the top four are within a point and a half of each other, and the two decks with the best match win rates among the big five, Blink and Druid, are 8% and 5.5% of the field.

And The Hobbit itself? Legal in Modern since August 14, it has barely touched the format. Karsten counted the set's most-played Modern cards across more than 800 decks: Key to the Side-Door with 34 copies, Bilbo, Thief in the Night with 23, Kíli the Resourceful with 20. None of them appears in the top cards of any of the decks above; the one Hobbit card in our Boros data is Belladonna Took at 13% of lists. The set that changed Modern this summer was Edge of Eternities, and the card was Quantum Riddler.

Bottom line

If you're registering for Baltimore on Friday, the safe read is this. Izzet Prowess will be the most-registered deck, because it's cheap, it just won two Spotlights, and its core — Cutter, Channeler, Swiftspear, Bolt — costs less than a single Ragavan. Esper Goryo's is the better-positioned of the two leaders, because it beats Prowess and its seventy-five is already settled. Esper Blink is the pick for the player who wants to beat both and is willing to dodge Boros. Devoted Druid is the pick for the player who wants to beat Goryo's and is willing to dodge Prowess. Boros Energy, the July king, is a fine deck for a room full of Blink and a bad one for a room full of Goryo's, which is what the room will be.

What to watch after Sunday: whether Goryo's puts a list in the Top 8 this time, whether Blink crosses 10% once the Baltimore lists are published — that's the point where Boros Wildfire becomes the sharpest deck in the format — and October 12.

Data: DeckCortex Modern corpus. Shares are decklist-appearance shares of published Modern lists (mtgo.com Challenge and League publications, MTGTop8-hosted paper events including the two Magic Spotlight main events and Pro Tour Marvel Super Heroes): 5,128 lists from 295 events, July 12–September 10, 2026, and 2,554 lists from 143 events, August 11–September 10, after removing 1,207 (resp. 593) MTGTop8 republications of mtgo.com events that carried a different archetype label — a cross-feed duplicate found while preparing this piece; our earlier Modern figures, including the July 12.1% Boros Energy share, double-counted Challenge top finishers. Monthly shares use calendar-month buckets on the same basis. Archetype labels were merged only where same-window card cores matched (the full evidence file is versioned with our research). Not field share, not match win rate; DeckCortex does not track match results. Event wins count first-place rows excluding MTGO League 5-0 publications. The conversion figure is the share of a deck's published full-standings finishes that reached a top cut. Win rates and matchup percentages are mtgdecks.net's, August 11–September 10, 2026, more than 22,201 matches, all platforms; cells under about thirty matches are noise. Winner's-metagame shares, the Pro Tour win-rate remark and the Hobbit card counts are Frank Karsten's on magic.gg (August 27, 2026). Spotlight results per Star City Games and magic.gg coverage; RC Baltimore structure and enrollment per melee.gg and starcitygames.com as of September 10. Banned & Restricted facts per the May 18, June 29 and August 10, 2026 announcements, cross-checked against Scryfall legality. The deck engine's combo count is a signal, not a verdict. Card prices are Scryfall paper aggregates as of September 10, 2026.

← All articles