Post-IKO Weeks 4-5 and Solving the Companion Crisis

I took a Week 4 blogging hiatus but continued to track format changes past our Week 3 article. Good news: Wizards finally acknowledged the companion problem in Ian Duke’s May 18, 2020 Banned and Restricted update. Duke even included a broader statement about multi-format companion impact, which we’ll unpack later today. Bad news: companion dominance got worse. Lurrus, Yorion, and the rest of the Ikoria Pokesquad have completely taken over Modern as the format collapses around a few top-tier decks. Burn and Prowess continued to exceed URx Delver shares at the height of the Treasure Cruise reign. The online Modern community is in peak outrage over companions specifically and format management generally: 17 of the top 20 Modern subreddit posts from the last month are criticizing Wizards, blasting companion, or lamenting Modern’s decline. There’s even a community of players who have lost so much confidence in Wizards they’ve taken format management into their own hands. I’m sensing the same despair and rage I wrote about in my inaugural 2020 article, “Fixing Modern: Redefining Format Mission,” which Wizards still has not addressed. All of this makes me fearful for Modern’s direction and future, but for now, we need to resolve the immediate companion crisis.

Today’s metagame breakdown updates the statistics from Week 1, Week 2, and Week 3. We’ll look at top-tier decks, companion prevalence, and changes from Week 3 to Week 5. I’ll then explore Wizards’ acknowledgement of the companion problem, and offer banlist and mechanical solutions Wizards can employ to fix this mess. Like most readers, I’m exhausted from constantly identifying and discussing major Modern problems. I just want to write about fun decks and play a fun format again. But until we pressure Wizards into decisive action on these formative problems, we’re not going to enjoy Modern the way we once did. This requires clear data to inform our conversation and clear conclusions based on that data.

Post-IKO Modern Weeks 4-5

One advantage of taking a Week 4 break was having an opportunity to absorb more Modern content. For instance, my former Modern Nexus colleague, David Ernenwein, wrote his own thorough metagame update the other week. I enjoyed reading his analysis and methods, although I disagreed with some of his conclusions (Lurrus, even downtrending to the mid-40% range, is still massively problematic) and presentation (“Other” decks should never be listed first on any metagame breakdown above a most-played Tier 1 deck). As today’s update will show, unlike Ernenwein, I identify an acute problem with Lurrus specifically and companions broadly, a metagame no longer trending towards diversity, and a format in deep trouble.

Metagame Breakdown

Our Magic: The Gathering Online dataset is now up to a respectable 38 events and 928 datapoints since April. Preliminaries continue to overlap with Challenges, Super Qualifiers, and other premier MTGO events, so I’m going to keep including them in the dataset to guarantee a larger sample. As always, curated Leagues are not included, and our ▲ and ▼ indicators show percentage changes from Week 3 into the end of Week 5. This week’s Tier 1 cutoff is 4%.

Post-IKO Modern Metagame: 04/18 – 05/24/2020 (n=928)

  1. Prowess: 11.4% (106) ▲3.7%
  2. Burn: 8.6% (80) ▼-2.1%
  3. Jund: 7.9% (73) ▲0.5%
  4. Amulet Titan: 4.6% (43) ▼-0.7%
  5. Ponza: 4.6% (43) ▲0.3%
  6. Devoted Devastation: 4.4% (41) ▼-0.7%
  7. Eldrazi Tron: 4% (37) ▲1.4%
  8. Temur Urza: 3.4% (32) ▼-1.1%
  9. Bant Snow Control: 3.2% (30) ▼-1.3%
  10. Humans: 3.1% (29) ▼-0.9%
  11. Hardened Scales: 2.8% (26) ▼-0.6%
  12. Ad Nauseam: 2.7% (25) -0%
  13. Bogles: 2.4% (22) ▲1.1%
  14. Mono G Tron: 2.3% (21) ▲1%
  15. Scapeshift: 1.7% (16) ▲1.4%
  16. The Rock: 1.7% (16) ▼-0.2%
  17. 5C Niv: 1.6% (15) ▲0.1%
  18. Azorius Control: 1.6% (15) ▲1.1%
  19. Dredge: 1.5% (14) -0%
  20. 4C Uro Snow Control: 1.4% (13) ▼-0.1%
  21. Grixis Delver: 1.4% (13) ▼-0.5%

Soul-Scar MageLet’s talk red decks. Prowess has been rising for the past few weeks with Burn on the decline. Specifically, we’re seeing a sharp increase in RBx Prowess, which pairs discard spells and Cling to Dust with traditional Prowess staples like Dart, Swiftspear, Manamorphose, Bauble, etc. I’ve seen scattered debate around whether or not these RBx Prowess upstarts are distinct from traditional Mono R or Boros builds, and by extension, if all of those are separate from Burn. We could theoretically unpack that macro “Prowess” category into at least 2-3 other sub-Prowess builds, which would paint a more positive top-tier picture where no single deck has a 10%+ share. We could also pretend Burn and Prowess are different when assessing format health.

I’m not buying any of that. Today’s Prowess flavors today are no more distinct than 2016’s Eldrazi. Fun fact: colorless, Azorius, and Izzet variants of Eldrazi still had between 30% and 50% of their slots vary from deck to deck. That’s less overlap than 2015’s Grixis Control and Grixis Twin (sharing about 66% of slots). And yet, despite Eldrazi lists sharing no more than half of their cards, these were clearly the “same deck” from a health and gameplay perspective. Rx Aggro presents a similar scenario today, constituting a profoundly unhealthy macro archetype. As I said in Week 3, there’s no appreciable difference between dying on turn four to Guides, Eidolons, and Boros Charms vs. Swiftspears, Mages, and Lava Darts. Or getting buried on turn six to all of that plus Lurrus and Bauble recursion. This Rx experience is at a distressing 20% of the format, worse than Treasure Cruise URx Delver variants at 17.5% in late 2014. Taken in context, the Rx rampage has elevated from an orange flag to a literal red flag. I struggle to remember any healthy Modern metagame where a macro-category like Rx Aggro had an upward-trending share around 20% (see the metagames of BGx Deathrite, URx Delver, Eldrazi, or most controversially, URx Twin) and that hasn’t changed in 2020.

Karn LiberatedLooking beyond the Rx takeover, we see other widening metagame cracks. Every top-tier Lurrus deck except Jund, Burn, and Prowess is down (poor Grixis Delver has been floundering ever since it appeared), defeating our cautious hopes that Lurrus could eningeer a midrange renaissance. Eldrazi and Gx Tron are lumbering back to the top, and Ponza is climbing with them to exploit big mana weaknesses. Meanwhile, grindy Astrolabe decks are declining (Temur Urza and Bant-X Uro Snow both down by over a perventage point) replaced by Yorion-powered combo kills in 80 card Astrolabe Scapeshift. At least Bogles is capitalizing on Rx’s rise, jumping 1.1% but not exactly contributing to the most enticing Modern gameplay. The only positive sign of format adaption is old-school Azorius Control peeking out of Tier 3 irrelevance into the big leagues. Thank Kaheera, the Orphanguard (in 73% of Azorius decks) for giving the control deck some direction.

The archetype breakdown captures all these worrisome developments from a bird’s eye sky noodle’s view:

  • Aggro: 32.8% (304) ▲.2%
  • Midrange: 23.6% (AA 61 + 158) ▼1.2%
  • Combo: 15.8% (AA 23 + 124) ▲.6%
  • Big mana: 15% (139) ▲1.7%
  • Control: 11.2% (AA 75 + 29) ▼.6%
  • Tempo: 1.6% (15) ▼.8%

Jegantha, the WellspringAggro, combo, and big mana are up, midrange, control, and tempo are down. This is a bad look for a format often accused of promoting goldfish-style Magic. Big mana’s continued rise remains unsurprising. As I predicted in basically every metagame update since Week 1, the companion hype died down and players inevitably returned to mainstays Eldrazi and Gx Tron. Except some of those Tron lists, particularly the 1% of players on Gruul Tron, have even upgraded to better support Jegantha. The midrange decline is even more worrisome than big mana’s ascent. I was initially hopeful Lurrus would create grindier games of Magic and incentivize midrange and control players to adopt the Cat into their lagging UBx control, Death’s Shadow, and similar strategies. My Week 1 hopes have imploded as Lurrus just keeps edging towards an increasingly unhealthy and linear metagame. Even grindmaster Yorion is having a similar effect, with the more interactive Bant-plus and Urza decks both declining in favor of a Scapeshift combo kill.

There aren’t a lot of positive metagame takeaways outside of snowless Azorius Control making a quiet entry into the top 20. Midrange and control keep trending down, aggro and big mana keep trending up or reinforcing already high shares. Our previous picture of “relative diversity” is suffocating under the weight of less interactive play patterns, repetitive games, and a mechanic that will likely go down in history as one of contemporary Magic’s most egregious and baffling design mistakes.

Companion breakdown

If the metagame picture is bad, the companion picture is worse. We’ve reached a point in Modern where you are either playing companions, a Titan strategy, or a Tron strategy. You can try and abuse metagame gaps with screwballs like Ad Nauseam, Dredge, or Azorius Kaheera Control, but you’re probably playing a worse version of the omnipresent companion decks. The Week 5 companion table really underscores this problem, with sizable jumps in virtually all categories.

CompanionsMeta %Top-Tier %T8 %
Lurrus

46.3% (430)

▲.9%

49.6% (364)

▲.7%

48.8% (78)

▼.3%

Yorion

11.1% (103)

▲1.2%

11.2% (82)

▲1.2%

13.8% (22)

▲3.1%

Jegantha

4.3% (40)

▲1.6%

3.4% (25)

▲.8%

4.4% (7)

▲.6%

Zirda1.1% (10)0.1% (1)0% (0)
Obosh

4.1% (38)

▲.4%

5% (37)

▲.8%

3.1% (5)

▲1.2%

Gyruda0.4% (4)0% (0)0% (0)
Lutri0.1% (1)0% (0)0% (0)
Umori0.1% (1)0% (0)0% (0)
Kaheera0.8% (7)0% (0)0.6% (1)
ALL COMPS

68.3% (634)

▲4.6%

70.3% (509)

▲4.1%

70.6% (113)

▲4.5%

Kaheera, the Orphangurad2020’s companion coup continues to surpass even the most pessimistic expectations. 70% (!?) of top-tier decks and Top 8 decks are running companions. 50% of top-tier decks are Lurrus decks alone. We had a 05/16 Challenge where all but four decks ran companions and half of those companions were Lurrus. Remember from Week 3’s article that Once Upon a Time, Oko, Dig Through Time and Treasure Cruise combined, and Eye of Ugin never exceeded about a 44% format share. Lurrus at 50% and partner in crime Mishra’s Bauble at around 45% are committing historic diversity violations, to say nothing of Obosh, Yorion, Kaheera, and Jegs being virtually mandatory for the non-Lurrus half of the format. Even worse, all of these measures continue to trend up from week to week, with the Weeks 4-5 jump representing the most alarming increase so far. Week 1 had “only” 57% of all decks using a companion. Now that’s up to 68.3% and even higher for tiered decks. Lurrus has been hovering around 50% saturation since Week 1 and sees no sign of dipping anywhere below 45%, which would still beat the combined TC plus DTT share.

As with the metagame breakdown, I don’t have many positive conclusions to share. I had this conversation with some Modern colleagues on a discussion board, and it’s on my mind today. It’s tempting for data-driven authors to be nuanced and even-handed for the sake of appearances alone. Readers count on us to take that middle ground. I’m all for level-headed cautions and disclaimers, but sometimes we just need to call a spade a spade. Or call a Nightmare a nightmare. Equivocation for the sake of equivocation isn’t helpful, and it would be disingenuous to analyze this collapsed metagame and declare it anything other than a mess. Lurrus and companions are crushing even the rosiest interpretations of the format. For every positive observation (e.g. only 41% of Humans decks experiment with Jegantha despite fears the Elk would be a mandatory Gurmag Angler), there are many more negative ones (e.g. four of the seven Tier 1 decks are Lurrus decks. The remaining two are big mana decks or big mana slayer Ponza).

All of this forces us and Wizards to stop assessing the damage and start making repairs. Between the high-level metagame factors and gossip from the ground (Sanchez’s six Prowess mirrors into a Challenge Top 8 with three Prowess and one Burn), we need to switch gears from diagnosing the problem to identifying solutions.

Taking Aim at Lurrus and Companions

It’s been a relatively dismal two weeks of Magic content, with what feels like endless Wizards missteps and player outcry from even the game’s most loyal supporters. Thankfully, Wizards shined a rare ray of hope through this storm in its May 18 B&R update. The most obvious takeaways were Lurrus’s and Zirda’s exit from Legacy and Vintage. Also, removing obnoxious Winona and Magistrate from Arena. Add all four of these to the blossoming list of multi-format bans we’re going to see throughout 2020, but refocus your attention to the B&R’s last two paragraphs. Today, I’m going to do a close reading of this official acknowledgement as a jumpoff point to discuss potential solutions to the companion crisis.

Unpacking the official companion stance

Hats off to Ian Duke for having the courage to acknowledge companions as a growing, multi-format threat. We shouldn’t have to praise Wizards employees for transparent communication around obvious issues, but that yawning communication gap is a topic for another day. For now, let’s just focus on Duke’s and R&D’s stance on companions. I’m pasting his entire statement in segments for clearer breakdown (emphases added) with commentary:

While this set of changes has focused on Legacy, Vintage, and Brawl, we’re continuing to watch the evolution of the metagame in each other format, including Standard, Pioneer, and Modern. If changes become needed in other formats, we’ll provide those separately in a future announcement. As of now, we’re seeing a diverse and dynamic metagame that changes from week to week in each Standard, Pioneer, and Modern.

Duke starts with a basic acknowledgement of R&D monitoring other formats (i.e. Standard, Pioneer, and Modern). Longtime Moderners know this concession is rare, with Wizards often including no context beyond “No changes.” The existence of these pargraphs alone shows Wizards is having serious conversations about the problem. That said, as of Duke’s publication date, Wizards judged those formats to feature “diverse and dynamic metagame[s]” which did not require bans. I understand this stance and shared it for a few weeks. Data was less clear and less plentiful. We were seeing some exciting Lurrus decks (Grixis Delver, The Rock, 4C Control) competing alongside more tuned options (Burn, Prowess, Jund). We were enjoying widespread experimentation with other companions and cautious optimism. But as we’ve seen from Week 3 to Week 5 in today’s article, the format is rapidly veering off that course in the exact ways Wizards feared.

Before determining whether any changes are necessary, and what the right changes would be, we need to see the metagame come closer to an equilibrium state. Currently, these formats are shifting too quickly for data to indicate what, if any, card or archetype poses a problem.

Bridge From BelowThe first bolded bit underscores my interpretation above. Wizards wants formats to settle into an equilibrium, or approach such a state, prior to action. Having analyzed the format for five weeks now, I understand their initial hesitation and also encourage them to admit the format has now reached this state. Bridge from Below and Hogaak saw decisive action on comparable timelines (particularly Bridge in the barely one month between Modern Horizons’ legalization and Bridge’s ban), and Modern is in worse shape today. As for the second bolded bit, I’m calling shenanigans. Wizards couldn’t identify a single “card” as the main problem? Lurrus was a visible orange flag within an hour of its preview, a budding problem as early as Week 0, and a historic violator by the end of Week 1. We’re over a month past IKO‘s release and it’s maintaining a +/- 50% share. Companions have also stabilized at the 70% level as the format slips further towards a big mana and aggro plurality. We might not be at the true equilibrium state, but it’s clear where Lurrus and Modern are now and will keep heading into June.

We are aware of some players’ concerns about the frequency at which they encounter decks using companions across several formats. While we’re not currently seeing problematic win rates in Standard, Pioneer, or Modern from decks using companions, we are looking at overall metagame share and potential for repetitive gameplay.

Aetherworks MarvelOn the one hand, I agree win rates must be considered in banlist decisions. R&D has been better at promoting data-driven articles and decisions from 2017 through present and it’s good Wizards appreciates this gold standard of data. On the other hand, Wizards correctly identifies win rates alone do not determine bans. This is the exact consideration that led to Aetherworks Marvel’s and Lattice’s banning in their respective formats. Or even Once Upon a Time more recently (prevalence and win rates were responsible). Metagame share and repetitive gameplay are important R&D considerations. Lurrus plus companions as a whole are undoubtedly violating both criteria through the end of Week 5, whether by numbers (Rx should never sustain 20% of the format) or anecdotal experiences. This should push Wizards to actions less palatable in previous weeks.

If we see signs of long-term health issues resulting from high metagame share of companion decks, we’re willing to take steps up to or including changing how the companion mechanic works. For now, metagames need more time to evolve before we can determine whether changes are necessary.

This was by far my biggest takeaway from Duke’s article. Duke not only signaled Wizards’ awareness of the exploding companion issue, he also tipped the company’s hand at an unprecedented solution: changing the mechanic itself. Magic players across the content-sphere had been speculating on this possibility for weeks before Duke’s article, and his official position lends credibility to this approach. The fact Wizards mentioned it at all should encourage us to speculate on these potential changes.

For any Wizards staffers reading my article today, please reconsider your May 18 stance. The Modern metagame is flattening towards an equilibrium where Lurrus, companions, and ramp rule the roost. I understand the desire to preserve an exciting mechanic, maintain product viability, and allow formats time to organically adapt to issues, but we are past that point. Modern requires action that balances damage to the format against reduced sales and destabilized player confidence; hopefully some suggested solutions below will fulfill those demands. As for readers more generally, please spread this message and demand action from Wizards. Wizards does listen, even if their representatives can be puzzlingly silent, and we need them to know the Modern status quo is untenable.

Banlist solutions

I’m keen on mechanical solutions to companion, but also acknowledge the banlist as a viable management tool. Until Wizards addresses some fundamental issues in their set design and playtesting processes, bans are going to remain an inescapable part of the 2020 and beyond Magic experience. As yet another disclaimer, I acknowledge other Modern suspects might require close R&D scrutiny: see Astrolabe, Veil, and others on the Modern community hitlist. Today, however, we need to table those discussions until Wizards has tamed Lurrus and company. Here are some of the likeliest banlist responses, starting with the trap solution everyone is worried about:

NOT a Solution: Ban Mishra’s Bauble

Mishra's BaubleThis is a scary possibility I don’t want to even acknowledge. Unfortunately, Wizards has a alarming track record of banning an older card when a new release creates a busted synergy. See Opal (with Urza) and Lattice (Karn) this year, Bridge (Hogaak), Looting (Hogaak, Phoenix), and KCI (Trawler) last year, and many more before that. In these cases, Wizards hit the old card instead of the new enabler: pre-2018 examples include Electromancer breaking Storm but Song taking the hit, and Oath Eldrazi or SOI block graveyard cards busting their respective decks prior to Eye/GGT getting hammered. Wizards does ban newer cards (Oko, Hogaak, Once Upon a Time, or TC/DTT before), but even there we sometimes see an older card take the initial fall as in Bridge from Below. Wizards also has a record of hitting free or zero-mana effects and indicting cards that reduce variance (cantrips across all formats including Probe, OUaT, and P&P in Modern). All of this suggests Wizards could conceivably terminate Bauble instead of Lurrus, especially to deflect responsibility from their busted packseller to the Coldsnap old-timer.

This bad ban would not solve any fundamental companion problems. We’d still see unprecedented Lurrus saturation as players swapped Bauble with other free or even 1 CMC cantripping effects (Conjurer’s Bauble and Unbridled Growth are possible candidates). Lurrus is good enough on its own with all the other low-cost permanents; indeed, decks like Devoted Devastation don’t even run Bauble! That brings us to a Bridge 2.0 scenario where Lurrus decks evolve beyond a glancing ban and then get banned anyway. We must discourage Wizards from this disastrous approach.

(Real) Solution #1: Ban Lurrus

Lurrus of the Dream DenNo surprises here. If Wizards does not change the mechanic, they at least need to ban Lurrus. The Cat has committed at least three bannable offenses in Modern, any one of which might not be damning, but all three of which are unacceptable. It has unprecedented Modern prevalence in excess of virtually every other card on the format’s banlist. That might be okay if we were entertaining a Legacy Brainstorm scenario, but Lurrus does not promote the same kind of high-skill, distinctive gameplay the venerable cantrip encourages. Instead, Lurrus pushes a metagame where Rx Aggro is 20% and all other non-Jund Lurrus strategies are on the downswing. Add repetitive gameplay due to Lurrus being “drawn” in every starting hand and its mechanical break of being a virtual 0-for-2 in most exchanges, and you have a singular design blunder to cap off a 2019 and 2020 season of offenses. This is exactly what bans are intended to address and if Wizards can’t fix the mechanic, Lurrus must go. This is too bad because I actually think maindeck Lurrus is an interesting choice that may promote some slower, interactive gameplay, but if Wizards won’t change the mechanic then maindeck Lurrus is a necessary casualty.

Solution #2: Ban Lurrus and Yorion

Yorion, Sky NomadLurrus feels like an inevitable ban in these models, but I actually think Wizards also needs to hit Yorion. Failure to do so will create a Modern where Yorion supplants Lurrus as the companion of choice, pushing into similar 40%+ territory over the summer. We already saw Bant Snow and Temur Urza decks leading the pre-IKO format, and with Lurrus gone, we would likely return to that Week 0 metagame with Yorion at the helm. This may result in a more interactive and grindy Modern revolving around Yorion value instead of Lurrus blitzes, but it would still suffer from similar prevalence issues. Wizards has a history of banning both the obvious problem and the natural successor (e.g. TC and DTT, Hogaak and Looting, Oko and Opal, etc.), and I’d be surprised if they didn’t take similar action here. This could also take the form of some kind of heavyhanded Astrolabe ban (and/or Bauble if Wizards really goes bananas), but that seems a lot less likely than just the companions.

Solution #3: Ban Lurrus and/or Yorion alongside a mechanical change

Kill it with fire! In this truly nuclear scenario, Wizards both enacts a sweeping companion change (see next section) and then bans either Lurrus and/or Yorion just to be safe. Even if you added Lurrus to your hand before the game started as a 7th card (drawing one less card to begin with or scrying one to the bottom), that inherent consistency boost of always having your Lurrus would still result in dominance issues. RBx Prowess and Jund are already outfitted for this change with ample discard spells in the maindeck to bin opposing companions. If Wizards changes the mechanic and bans its two worst offenders, however, R&D could allow the more offbeat companions to flourish (get ’em, Lutri) without risking future bans or rules changes in 1-2 months.

I still think pure ban approaches are less likely than mechanical ones. Banning your iconic IKO cards before we’ve been playing the cards in paper (thanks, COVID-19) is a risky economic proposition for a company that needs to generate sales. Mechanical changes might allow the cards to work as intended.

Mechanical solutions

Disclaimer: I have no background in game design and acknowledge there are more qualified experts to discuss companion changes. I’ve also struggled to keep up with the sheer volume of companion revision suggestions on Twitter and Reddit, and am happy to give credit where credit is due if anyone can cite some sources for these ideas. I do, however, know Modern and can discuss companion mechanical changes through our format’s lens.

Wall of WoodBefore talking companion, it’s important to understand how Wizards can change the mechanic at all without functionally rewriting card text. Indeed, companions have italicized text under the mechanic name which reminds you how the mechanic works. Keyword: “reminds.” This so-called reminder text is not actual game text per the comprehensive rules, specifically 207.2 (“The text box may also contain italicized text that has no game function“) and 207.2a (“Reminder text is italicized text within parentheses that summarizes a rule that applies to that card“). Wizards can change the comprehensive rule entry for companion without resulting in functional changes to the rules printed on the cards themselves. Only the reminder text would be obsolete, but if old-school players can get over the defender keyword materializing on all pre-2004 Wall creatures, we can survive this new change too.

Solution #4: Exile a card, cast the companion

In one of the more basic mechanical changes, players would keep companions in their sideboard and just tack on an additional cost of exiling a single card. Exiling that extra card eliminates one of the most troubling “8th card problems” of companion. This change still insulates companions from discard-based interaction but at least makes it harder to cast them from the sideboard. I suspect this change would not significantly impact the prevalence of Lurrus, Yorion, or any of the other top Modern companions, but it would at least increase their opportunity cost.

Solution #5: Draw fewer cards, add companion to hand

The next two solutions actually put the companion in your hand at the start of the game where discard effects can target it. It also addresses the 8th card problem by trading a card in your hand for the companion in your virtual command zone. Solution 6 does so by revealing the least information and giving player the least flexibility. Here, players who announced a companion would draw N-1 cards, where N is 7 minus the number of times mulliganed, and then make mulligan decisions. After making those decisions, they would add the companion. As an example, on the first hand of the game, a Jund player would announce they are playing Lurrus, draw 6 cards, make mulligan decisions, and then add Lurrus to their hand. This would bring it up to a starting 7 cards. If they mulled that 6-carder to 5 and kept, they would still end with a grip of 6 but one would be Lurrus. This allows players to build around a guaranteed companion, but makes your openers worse.

Solution #6: Scry away excess cards, add companion to hand

If Wizards wanted to give players more information when choosing opening hands, they players could draw as normal and then scry one for a companion to replace. Mechanically, players would announce a companion and draw N cards where N is 7 minus the number of times mulligan. They would then make a mulligan decision on that hand and, if satisfied, add the companion before scrying one additional card to the bottom. This gives players more information about what hand they should keep but results in a weird double-scry scenario. For example, consider a Burn player who drew 7 cards and then mulled to 6. They draw 7, scry an excess Eidolon to the bottom, and then add Lurrus to their hand. Now they scry yet another card to the bottom so their final hand size is 6. In theory, Wizards could keep companions pushed by allowing players to keep that companion in their hand without scrying the additional card, but then these players would be starting with literal 8 or 7 cards in hand. That seems like too much even for designers who created companions to begin with, and I doubt it would happen.

Solution #7: Prohibit the mechanic in certain formats

Technically, this is a form of a ban, but it’s still groundbreaking in that you wouldn’t ban any of the cards from maindecks. They’d just be banned as companions, and/or the entire companion mechanic could be barred from a given format. I doubt Wizards would pursue this option because of complexity issues, and because they seem to admit companion is a unique, multi-format problem that demands cross-format solutions. It’s still an option if Wizards doesn’t mind increasing the deckbuilding complexity of older formats while maybe working on mechanical adjustments in Standard.

I’d love to hear other changes by players more familiar with this debate, so throw some my way in the comments or on the various discussion boards we frequent. Overall, this is the direction I think Wizards wants to head, and I would be surprised if a subsequent announcement on companions doesn’t include some kind of mechanical revision.

The State of Constant Crisis

FIres of InventionBeyond Lurrus, I feel like the Modern community and even the Magic community as a whole is just sprinting from fire to fire, trying to call 911 at Wizards while our Washington dispatchers Tweet at us about some cool new product being released in a few months instead of deploying the R&D firefighters. I’ve been zeroed on Lurrus in past weeks, but companions aren’t even close to the the biggest danger to Modern and other nonrotating formats. Those threats are a fundamentally broken design process that is guaranteed to create overpowered, multi-format mistakes, and a disturbing lack of consistent and transparent communication from Wizards’ leadership. I’m hoping Wizards takes actions on the immediate issues of Lurrus and companions so they can shift gears to more publicly and vocally addressing the cracks in the foundation. This will help us recover from our endless crisis states and get back to enjoying the game, especially competitive Constructed players who can’t escape the ceaseless parade of design disasters that have hit Standard (Yorion), Vintage (Lurrus), and everything in between.

As I’ve Tweeted at Wizards and discussed on social media, Wizards employees love this game and are some of its most tireless advocates. At the same time, new and perilous trends which have many veteran players worried. None of this makes folks like Mark Rosewater or Aaron Forsythe bad people or incompetent. It doesn’t mean Play Design needs to be fired, it doesn’t mean the game is going to die, and it doesn’t mean we’re all not going to load up MTGO or Arena less than 12 hours after quitting. It does, however, signal a potential shift in core elements many of us enjoy about Magic, and it might lead to a less rewarding, complex, and engaging game. We have an obligation to speak out about these issues and encourage Wizards to change course. Today it’s Lurrus in Modern but tomorrow it’s bigger issues like card balance and communication.

Thanks for reading and considering these major issues. Let me know in the comments, on Reddit, on MTG Nexus, or via Twitter if I missed anything or got something wrong. Feedback, criticism, ideas, and thoughts are always welcome. Hopefully we’ll see a true Wizards announcement about an announcement foretelling the doom of Lurrus before my next article hits and until then, stay healthy and stay sane in this increasingly polarized and tumultuous Magic world.

 

Post-IKO Week 3 Update and Lurrus Comparisons

Weeks 1 through 3 of our post-Ikoria Lurrus world are in the books. The 05/04 through 05/11/2020 period saw both expected developments (more sweet companion innovations, more online complaints about the mechanic) and some unexpected shifts (an unusually public debate with Mark Rosewater around companion, a refreshingly nuanced take by Eric Froehlich). We’ve also seen Lurrus. A lot of Lurrus. Lurrus is den-master of every format, except in Standard where its “sky noodle” buddy soars supreme. Lurrus’s eternal format impact is so dire we saw an ominous Wizards Tweet Monday evening which will likely spell the companion’s demise in at least Legacy and Vintage. Modern, however, is conspicuously absent from that update. This means Moderners will continue to debate Lurrus’s place in an uprooted metagame. As in previous weeks, today’s goal is contributing clear data to that conversation. This ensures opinions are informed and grounded in real metagame developments, not just gut instincts and anecdotes.

Week 3’s breakdown updates the same measures we’ve seen so far: Magic: The Gathering Online’s top-tier decks, archetype summary, and companion shares. But just as our conversation about Lurrus is evolving, so too must our analysis. To acknowledge the growing complexity of this major Modern issue, I’m also going to compare present Lurrus shares to those of other format icons: Bloodbraid Elf, Jace, the Mind Sculptor, Stoneforge Mystic, and more. This will situate Lurrus in the broader Modern context and help us decide if the Cat’s presence is business-as-usual or a true feline nightmare.

Modern in Post-IKO Week 3

Most of us are familiar with the current state of Modern, so it’s less important to do full recaps of previous weeks. If you want a breakdown of Week 1 or Week 2, check out those respective articles. Today, we’ll just focus on the Week 3 picture with the same gains () and losses () indicators as last week.

Metagame breakdown

As of 05/11/2020, we’re up to 24 events in our post-IKO dataset. I’m still including Preliminaries alongside Challenges, Super Qualifiers, Premiers, and other major MTGO events because these samples continue to overlap. Of the top 21 decks in the major event pool only, 16 overlap with the Preliminary top-tier group. As long as Wizards keeps publishing all 3-2+ Prelim finishes and doesn’t curate the data like they do with Leagues, we’ll keep including them. Here’s our current MTGO top-tier, representing the top 75% of the format. If you need more granularity, we’d pin this week’s Tier 1 cutoff at around 4.5%:

Post-IKO Metagame: 04/18 – 05/09/2020 (N=625)

  1. Burn: 10.7% (64) ▼2.4%
  2. Prowess: 7.7% (43) ▲1.8%
  3. Jund: 7.4% (45) ▲.6%
  4. Amulet Titan: 5.3% (31) ▲1%
  5. Devoted Devastation: 5.1% (31▲.1%
  6. Bant Snow Control: 4.5% (28) .2%
  7. Temur Urza: 4.5% (28.2%
  8. Ponza: 4.3% (24) ▲.9%
  9. Humans: 4% (25) ▼1%
  10. Hardened Scales: 3.4% (21) ▼.2%
  11. Ad Nauseam: 2.7% (15) ▼.2%
  12. Eldrazi Tron: 2.6% (13) ▲.3%
  13. The Rock: 1.9% (12) ▼.4%
  14. Grixis Delver: 1.9% (12) ▼.4%
  15. Infect: 1.6% (10) ▲.2%
  16. Neobrand: 1.6% (9) ▼.2%
  17. Dredge: 1.5% (9) ▼.5%
  18. 5C Niv: 1.5% (9) ▼.3%
  19. 4C Uro Snow Control: 1.5% (9) ▲.1%
  20. Simic Titan: 1.4% (9) ▲1%

Monastery SwiftspearBurn is the big Week 3 loser, but Prowess is the big winner. This means you probably won’t notice any difference in the overall presence of Rx aggro, which remains a major format force (19% down to 18.4%). This is also one of the least healthy aspects of this breakdown. I know Burn and Prowess have technical distinctions, but from a MTGO gameplay experience perspective, there’s no appreciable difference between dying on turn four to Guide and Lava Spike vs. Swiftspear and Lava Dart. We’ll need to keep monitoring these shares to see if Lurrus is disproportionately benefiting these decks more than the rest. Outside of this worrisome trend, Week 3 sees many of the same top-tier decks as Week 2 in virtually identical positions. This suggests the metagame continues to crystallize around a competitive core, which thankfully represents a relatively diverse range of strategies beyond the Rx hegemony.

Glistener ElfThere are also new top-tier entrants. First is the 4-5C Astrolabe madness of Uro-powered control tying up the bottom slots with Dredge, Neobrand, and 5C Niv. This feels like a natural consequence of Yorion’s parallel rise; if you’re going up to 80 cards in an Astrolabe deck, why not play the best cards of every color? Incidentally, 100% of these decks use the Sky Nomad, not to be confused with the 100% of 4C Snow Control decks (N=6) using Lurrus in previous weeks. The second newcomer is Infect (mostly Simic, some Golgari) sneaking into 15th from Tier 3 obscurity. Infect is likely capitalizing on a simultaneous increase in big mana. And speaking of big mana, Simic Titan is the last newcomer. Uro and friends saw a huge boost from last week that aligned with my prediction of increased Titan prevalence. Between Simic Titan and its other variants (Amulet Titan on top, Golgari Titan, Titanshift, and others below), the macro Gx Titan archetype makes up a considerable 8.6% of the metagame: more than the Tron options by far! Don’t sleep on these Titan decks and have a plan to beat them.

Here’s our overall archetype breakdown. I note major Week 2 to Week 3 changes, as well as Astrolabe (AA) decks vs. non-AA options in parentheses:

  1. Aggro: 32.6% (196) ▼2.8%
  2. Midrange: 24.8% (AA 48 + 101) -0%
  3. Combo: 15.2% (AA 6 + 83) ▼.1%
  4. Big mana: 13.3% (73) ▲2.9%
  5. Control: 11.8% (AA 54 + 18) ▲.5%
  6. Tempo: 2.2% (14) ▼.5%

I wouldn’t read too much into the apparent aggro decline, as we’re still seeing a horde of different aggressive builds across top tables: Rx Burn/Prowess occupies 18% of the format and are the likeliest macro-strategy you will face. A more notable shift is the rise in ramp. Week 3 saw a troop of Gx Titan strategies lumber into the spotlight, especially breakout Simic Titan. This aligns with predictions from previous articles. Players likely boarded the companion hype train in Week 1, abandoning boring Modern regulars like Tron and Titan. With the initial buzz dying down, we should see more players returning to ramp staples as they realize Prime Time doesn’t need companions to excel.

Mishra's BaubleThere are some complicated questions about whether or not any of this represents a truly “diverse” metagame. Some claim the format is just fifty shades of Lurrus/Bauble/Seal of Fire decks running lean aggro/combo gameplans. Others believe Lurrus is supporting a range of distinct decks across the spectrum with different playstyles. I’m still leaning towards the latter. As Jeff Hoogland Tweeted last week, critics are performing Olympic-level mental gymnastics to justify their hatred of these new cards. Be very wary of anyone who asserts Devoted Devastation, Burn, Hardened Scales, and Jund are all “the same deck.” There are many legitimate criticisms of Lurrus, its play patterns, its variance reduction, and its metagame impact without grouping obviously different decks based on 10-12 card overlaps.

Going forward, I am developing a more data-driven method to tackle this question of “diversity,” but need more time to crunch numbers. I’m optimistic I’ll complete that analysis in 1-2 weeks. Until then, I’m leaving this as an open question which is much more complicated than extremists on both ends are framing it. Can metagames be diverse/healthy and see card prevalence over 50%? Yes; see many Legacy periods and their Brainstorm hallmark. Does Lurrus have a similar impact on Modern? Maybe, but probably not. If anything, Modern has a much more troubling history of dominant staples leading to unhealthy environments (see Oko and Once Upon a Time). All of this forces us to be careful and nuanced when assessing this diversity question.

Companion breakdown

I’m combining all companion statistics into a single table for easy comparisons. The “Meta %” column shows the share of all decks using a certain companion. Next over is “Top-Tier %” which solely evaluates companion play in the metagame’s top 75% of decks. Finally, we look at “T8 %”; companions as a share of major MTGO T8s. Our ▲ and ▼ indicators denote changes from last week, but I’m only showing those for categories most people seem to care about: Lurrus, Yorion, and companions collectively.

Companions Meta % Top-Tier % T8 %
Lurrus 45.4% (284)

▲1.4%

48.9% (229)

▼.8%

49.1% (55)

▼.9%

Yorion 9.9% (62)

▲.9%

10% (47)

▲2.1%

10.7% (12)

▲1.9%

Jegantha 2.7% (17) 2.6% (12) 3.8% (4)
Zirda 1.1% (7) 0% (0) 0% (0)
Obosh 3.7% (23) 4.2% (22) 1.9% (2)
Gyruda 0.3% (2) 0% (0) 0% (0)
Lutri 0.2% (1) 0% (0) 0% (0)
Umori 0.2% (1) 0% (0) 0% (0)
Kaheera 0.2% (1) 0% (0) 1% (1)
ALL COMPS 63.7% (398)

▲2.8%

66.2% (310)

▲2.4%

66.1% (70)

▲2.3%

Kaheera, the OrphanguradObviously, our big story of the week is a certain cat. Specifically, THE RISE OF KAHEERA! Not only does Kaheera, the Orphanguard jump into Modern for the first time since Week 1, but it instantly claimed a Top 8 slot in a creature- and Astrolabe-free Azorius Control deck. Hats off to aspiringspike for not only ditching Snapcasters and boring Uro but even including a Terminus and Counterbalance (?!) playset. You’re my hero.

Dragging us back to reality, Lurrus continues to dominate despite negligible dips in each category. As a helpful Redditor reminded me last week, the last time we saw a new Modern card with this kind of format-wide penetration was the Treasure Cruise/Dig Through Time era (more on that comparison shortly), and even then we didn’t see those busted delve spells in literally half of all competitive decks. Even if different Lurrus decks are still doing different things, such as creature combo Devoted Devastation vs. midrange Jund, there are alarming parallels between Rx Burn/Prowess’s 18.4% share and the 17.5% share of 2014 URx Delver. The bottom line is this: when a single card is in 50% of your top-tier decks, you need to ask tough questions about the kind of format Moderners want.

Yorion, Sky NomadMoving beyond what feels like the most polarizing card in recent Magic history, Yorion is another Week 3 winner. The Sky Nomad gains points in every column, even increasing its shares between categories. This overperformance from 9.9% of the overall metagame to 10.7% of the T8 metagame, although small in magnitude, does suggest Yorion is performing better than its baseline share. Continue to expect more Yorion as players, particular Standard refugees fleeing the Bird, experiment with its laughable deckbuilding “restrictions.” Of course, the other big winner is the companion category as a whole. Just shy of two-thirds of Modern’s best decks are playing the IKO poster-animals (56% of that is Lurrus and Yorion alone). I’m trying to shy away from the idea of companions as a macro category because the decks do different things and the brunt of that share is just Lurrus, but it’s still hard to not acknowledge their sheer takeover.

Overall, both the metagame and companion breakdowns continue to portray a Modern in varying states of crisis. Lurrus is still undermining even our most positive outlooks. For instance, although the top-tier metagame looks pretty diverse, that 18.1% Rx Lurrus aggro is a major blemish. And even though Yorion holds down a reasonable 10% share, the ~50% Lurrus share is staggering. We’re only on Week 3 of this metagame but these signs keep pointing towards pending Modern intervention by Wizards.

Assessing Diversity in Lurrus Modern

Last week, I participated in a very civil and engaging Twitter exchange about Lurrus. At one point in this exchange, @GermanSpaceAce posed the following question:

This wasn’t the first time I’d seen this kind of question: “how does Lurrus compare to the initial share of Insert-Modern-staple-here?” It was, however, the first time I realized this would be a great topic for a future article. I responded to their Tweet with some snapshot statistics, but realized I could draw this out even further to cover more staples and comparisons. This will provide important context to help us evaluate Lurrus’s sudden rise.

Historical cases: Oko, Bloodbraid, and more

Bloodbraid ElfThe table in this section presents a variety of past Modern players and their respective shares of MTGO decks playing them in different periods: the first month of its legality (or for cards like Eye of Ugin, the first month after it broke out) and its peak share over its last three months (for banned cards) or three months after its legalization (for cards that never got banned). Events include any noncurated or randomly selected dumps, which means some older Leagues are included for earlier eras, like Eldrazi Winter. I won’t include any events after Wizards started throttling data to prevent solved metagames (tangent: this absurdly misguided policy has never prevented any solvable metagames from being rapidly solved).

As a final note, the table counts both maindeck and sideboard slots, but does not double count a deck with copies in both the main 60 and the board. For instance, a 2014 Delver deck with 3 Treasure Cruises in the main and 1 additional in the SB would only count as one representative, not two. But if it had 4 TC in the main and a DTT in the SB, it would contribute +1 to both the TC and DTT tallies.

Here are our companion test standards. We’re looking to see if these standards are above, below, or matched with our comparisons:

  • Lurrus: 45.4% of the metagame after 3 weeks
  • Yorion: 9.9% of the metagame after 3 weeks

And here are the comparison cases. As an example of reading the table, SFM saw play in 14.1% of decks in her first month, which settled at 9.4% by the end of month three. For a banned card, you’d read it as follows: Oko, Thief of Crowns saw play in 27.2% of decks in his first month, which was up to 37.6% of decks in the three months up until its January 2020 ban.

MTGO EVENTS
First month Peak 3 months
Stoneforge Mystic 14.1% (52/368) 9.4% (65/688)
Bloodbraid Elf 10.7% (24/224) 9.4% (60/640)
Jace, the Mind Sculptor 17.9% (40/224) 10.8% (69/640)
Oko, Thief of Crowns 27.2% (74/272) 37.6% (356/946)
Once Upon a Time 20.2% (55/272) 37.8% (386/1020)
Eye of Ugin 40.7% (182/447) 37.6% (339/902)
Treasure Cruise 28.3% (197/695) 26.6% (583/2188)
Dig Through Time 12.4% (86/695) 17.3% (379/2188)

I had three big takeaways while compiling stats for this table. First, just look at how much data we had back in 2014! By the end of TC and DTT’s reign, we had an impressive 2,188 datapoints in our MTGO dataset. Compare with the three month retrospective for Oko and OUaT, with less than half of that data. The good old days.. My second takeaway was the upper end of these three-month peaks: all around the 38% range for the worst offenders. TC and DTT had minimal overlap, so we could even merge those two categories to get a macro Ux delve share of around 40% (+/- 2%). This is a potential warning threshold to monitor going forward.

Treasure CruiseAnd of course, there’s the third takeaways: Lurrus. Always Lurrus. I couldn’t believe the Den Master’s 45.9% monopoly is even greater than the combined shares of TC and DTT (43.9%) at their pre-ban heights. We can’t even explain this away as flavor-of-the-month hype either, as Lurrus handily outpaces the initial shares of much-hyped unbans like JTMS, BBE, and SFM. It’s surpassing the initial Eye of Ugin craze around Pro Tour Eldrazi. This truly suggests Lurrus is at unparalleled levels which pushes Modern into scary and uncharted depths. As for Yorion, our nomadic noodle appears to be at a perfectly acceptable prevalence relative to historical cases. For context, it’s seeing less play now than JTMS did three months after his release. This points to Lurrus being the problem, not other companions or companions generally.

Current staples: Astrolabe, Bolt, and others

While building the above table, I remembered the MTG Goldfish “Format Staples” pages. As of evening hours on 05/13/2020, I’m seeing Bolt at 42% of the format and Veil at 39%, both ahead of Lurrus at just 38%. There are obviously some data aggregation errors at play here (MTG Goldfish regrettably groups curated Leagues with the non-curated results), but this got me thinking. How do these more acceptable staple cards (or equally controversial, in the case of Astrolabe) compare with Lurrus?

To assess this, I checked the prevalence of established and contemporary Modern staples as an additional test. I focused on the pre-IKO and post-OUaT era for the analysis (03/10 – 04/17/2020). This was a period that led to no additional bans and represents our most recent “stable” metagame slice.

MTGO SHARE 03/11 – 04/17/2020
Arcum’s Astrolabe 24.6% (380/1547)
Lightning Bolt 33.7% (522/1547)
Path to Exile 26.1% (404/1547)
Veil of Summer 38.5% (595/1547)

I enjoyed seeing the results of these stats for a number of reasons. For one, we actually have a lot more data than I thought we would in 2020. N is up to 1,547 in a 1.5 month period! I for one am grateful for the scraps our benevolent Wizards datalords have thrown our way and will beg for more. More seriously, this suggests the addition of Prelims to our data picture has greatly enhanced our understanding of the format. I’m also pleased we have some actual numbers behind the constant Astrolabe vs. Veil vs. Bolt debate which has raged for month.

Arcum's AstrolabeThen there’s Lurrus. At this point, I’m not even surprised to see Lurrus enjoying more play than even the most venerable and versatile Modern staples from the pre-IKO metagame. I’m just disappointed (said in your most disappointed dad voice). It is quite alarming that the Lurrus draw/value package is seeing more play than not only the combined DTT/TC shares of 2014 Modern, but also the individual shares of Modern’s best, generic removal spells. If Astrolabe was an issue at 25% of the format with allegations it homogenized the format, I can’t imagine Lurrus at around twice that share should deserve more popularity. And this is coming from someone who thinks Lurrus has a partially positive effect on the format by pushing Modern to grindier games. Even acknowledging that potential effect, it’s implausible to me that a card at 50% of the format is healthy given these comparison cases. Finally, it’s worth noting that non-Lurrus companion shares as a whole (18.2% of the meta, 17.3% of top-tier decks) are significantly less than even just Astrolabe decks as a whole at 24.6%. Again, this points to Lurrus being a problem, not the mechanic.

In Store for Week 4

I don’t expect Lurrus’s share to get much higher than its current 50% throne. It will fluctuate +/- 5%, but it still can’t co-opt other companion decks. Obosh Ponza and the various Ux and toolbox Yorion strategies will still take up a number of top-tier slots, not to mention tried and true Humans or burly big mana. This puts a metagame cap on Lurrus’s likely share, which still isn’t too heartening given the sheer magnitude of that share.

Lurrus of the Dream DenAll of this continues to point towards Lurrus, not companions collectively, being a problem. Unless we accept the Lurrus/Bauble package as the new Modern normal and an equivalent of Legacy’s Brainstorm, there’s just no way we can possibly find this kind of plurality healthy. I struggle to see a world where a format with 44% of decks rocking TC/DTT was so problematic to warrant 2+ bans and a format with 50% Lurrus is somehow “fine.” Doubly so when Lurrus is primarily driving a steady 18% Burn/Prowess reign. Barring some significant (and unlikely) metagame shifts, this will likely end with Modern Lurrus facing the same fate as Legacy/Vintage Lurrus on Monday, 05/18. At that point, the question is no longer “Should Wizards act on Lurrus?” but rather “What kind of action should they take?” I’d prefer a less nuclear option, e.g. banning it as a companion, but I’m open to just about anything to address what appears to be a card-specific problem.

I wish these articles weren’t so doomy and gloomy because I’m just as tired of ban talk as everyone else. But Wizards keeps printing themselves into a banning corner. Until we and Wizards address some fundamental, underlying Magic issues, we are going to keep seeing Hogaaks, Okos, Lurrus’s, and other broken monsters Play Design keeps missing. For now, we’re back in firefighting mode and need to extinguish the Lurrus blaze before we can go back to these bigger problems. As always, let me know in the comments here, on Reddit, MTG Nexus, or Twitter if you have any feedback or changes. See you all after the Week 4 dust settles and we’re invariably still living under Lurrus’s paws.

Modern Metagame Update: Post-IKO Week 2

The COVID-19 quarantine has been bad for Magic, worse for our social lives, but great for metagame updates. Last week’s breakdown had such a positive reception I’m bringing it back for a Week 2 encore. Between schedule shifts and shelter-in-place requirements, I’ve had more time than usual to compile decks and crunch the numbers. The Modern community needs clear data now more than ever as companions continue to scurry, crawl, fly, and slither all over every Constructed format from Standard to Vintage. This takeover has ignited a socially distanced debate over individual companions and the mechanic itself. Without accurate metagame statistics, these conversations will remain disconnected from format realities. We owe it to Modern’s future (and our own Internet sanity) to stay informed.

In today’s article, I’ll do a quick recap of pre-Ikoria and Week 1 post-Ikoria Modern to frame our Week 2 update. I’ll then assess top-tier decks, the overall archetype breakdown, and a companion footprint way too big for one little Nightmare Cat. Week 2 finds us with some lingering health concerns for the format but also some promising developments. To some extent, there’s only so much change to track in a one-week span. But with most of us stuck at home, it’s important to match the pace of data with the blistering pace of online conversation. Whether you’re trying to select your next Magic: The Gathering Online deck to take down the Super Qualifer, evaluating overall format health, or just want to see some eye-popping figures about companions, read on.

Ikoria Week 0 and Week 1 Recap

I can’t believe companions have only been in Magic for two weeks. The rate and rancor of social media posts makes me feel like we’ve been living in a Lurrus world for months, not days, and it’s easy to forget the Modern landscape before IKO hit digital shelves. It’s important to understand this shifting picture to appreciate format evolutions from late April through early May. This section summarizes takeaways from last week’s breakdown of pre-IKO Modern and post-IKO Week 1. For full details about the old metagame picture, revisit last week’s article.

Pre-IKO stability

Uro, Titan of Nature's WrathRemember the good old days when Twitch chat would just complain about Arcum’s Astrolabe, Veil of Summer, and longtime format heel Tron? It wasn’t that long ago but it’s easy to forget the pre-companion picture. This was a format where Astrolabe decks were king but all archetypes saw top-tier representation. Overtuned 2019 and 2020 haymakers defined most games, and Modern saw a mix of standbys like Dredge, Humans, and Tron intermingling with newcomers like Temur Urza and Bant Snow. Here was our top-tier picture according to MTGO Challenges, Premiers, and Super Qualifiers:

Pre-IKO Metagame: 03/21 – 04/11/2020 (n=352)

  1. Bant Snow Control: 11.6% (41)
  2. Ponza: 8.2% (29)
  3. Dredge: 7.4% (26)
  4. Burn: 6.5% (23)
  5. Eldrazi Tron: 5.1% (18)
  6. Temur Urza: 5.1% (18)
  7. Humans: 4.8% (17)
  8. Prowess: 4.8% (17)
  9. Jund: 4.5% (16)
  10. Amulet Titan: 4% (14)
  11. Mono G Tron: 3.7% (13)
  12. Infect: 3.7% (13)
  13. 5C Niv: 2.6% (9)
  14. Death and Taxes: 2.3% (8)
  15. Bant Snowblade: 2% (7)

My biggest takeaways from this metagame were the power of entrenched snow decks, overall archetype and strategic balance, and green’s hegemony over all strategies (especially interactive decks). To emphasize the strategic diversity, here’s an archetype breakdown with Astrolabe decks noted in parentheses:

  • Midrange: 31.5% (AA 45 + 66)
  • Aggro: 29.8% (105)
  • Control: 16.2% (AA 47 + 10)
  • Big mana: 15.9% (56)
  • Combo: 6.5% (AA 6 + 17)
  • Tempo: 0% (0)

(Note these numbers have changed a little since last week’s breakdown. I’ve split the Astrolabe decks into archetype bins, which will make a cleaner comparison to Week 2 numbers later. I’ve also reclassified some decks into more accurate categories)

Although this metagame looked balance on a spreadsheet, I heard many complaints it didn’t feel balanced in games. Big mana, UGx snow, and various aggro forms dominated events. If you played something else it would often feel like a willfully worse version of one of these format heavyweights. Except Ponza. Long live Ponza, one of the cleanest and clearest examples of natural metagame evolutions we’ve seen in year. Of course, this relatively neat Modern picture changed almost overnight once IKO brought us companions and their denmaster, Lurrus.

Post-IKO upheaval

Lurrus of the Dream DenIt’s an overstatement to claim companions reshaped the format into something unrecognizably Modern. The post-IKO Week 1 picture was still very much Modern as we have seen in the past. It didn’t even have any “new” decks, unlike the post-Modern Horizons churn from Urza, Astrolabe, Hogaak, and more. But companions, especially Lurrus, both revitalized a bunch of bygone Modern decks and supercharged many existing Modern favorites. This reshuffled our tidy metagame in a short period of time. Here was the new Modern just one week after IKO arrived, adding MTGO Preliminaries into the analysis to boost our sample size. Relative % point changes from Week 0 to Week 1 are color-coded by gains and losses:

Post-IKO Metagame: 04/18 – 04/25/2020 (n=253)

  1. Burn: 15.4% (39▲8.9%
  2. Jund: 8.3% (21) ▲3.8%
  3. Humans: 5.5% (14) ▲.7%
  4. Bant Snow Control: 4.7% (12) ▼6.9%
  5. Prowess: 4.7% (12) ▲.1%
  6. Devoted Devastation: 4.3% (11) ▲4.1%
  7. Amulet Titan: 4% (10) -0%
  8. Dredge: 3.6% (9) ▼3.8%
  9. Grixis Delver: 3.6% (9) ▲3.6%
  10. Hardened Scales: 3.2% (8) ▲3.2%
  11. Temur Urza: 2.8% (7) ▼2.3%
  12. The Rock: 2.8% (7) ▲2.5%
  13. Ponza: 2.4% (6) ▼5.9%
  14. Eldrazi Tron: 2.4% (6) ▼2.7%
  15. Ad Nauseam: 2.4% (6) ▲1.5%
  16. 5C Niv: 2% (5) ▼.6%
  17. 4C Snow Control: 2% (5) ▲1.4%

No one skimmed this top-tier and confused it with a different format. We were still playing Modern, but this was a Modern in transition. Longtime metagame trackers will be familiar with Burn ascending in a Week 1 environment, especially with Lurrus pushing the archetype. Alongside Burn, we saw familiar contemporaries like Bant Snow Control, Prowess, and Urza. Devoted Devastation recovered after losing Once Upon a Time a month ago, and Humans, Titan, and Dredge maintained ground. But we also saw decks emerge from the ashes of older bans (Hardened Scales sans Opal, Grixis Delver without Probe) and from the netherworld of departed Modern options (The Rock? Ad Nauseam?). In total, five decks were big gainers, five decks were big losers, and a number of decks were just treading water.

All of this happened against a backdrop of continued archetype diversity, which I breakdown below by macro-strategies. AA vs. non-AA options are noted in parentheses, and archetype share changes relative to Week 0 are color-coded for winners and losers:

  1. Aggro: 38.7% (98) ▲8.9%
  2. Midrange: 23.7% (AA 16 + 44) ▼7.8%
  3. Combo: 13.4% (AA 3 + 31) ▲6.9%
  4. Control: 10.3% (AA 19 + 7) ▼5.9%
  5. Big mana: 9.5% (24) ▼6.4%
  6. Tempo: 4.3% (11) ▲4.3%

Hardened ScalesThis remained a relatively balanced metagame where everything was still viable but proactive decks were gaining ground. Aggro, combo, and tempo all improved position with midrange, control, and big mana losing points. This mirrored what we knew about companions in Week 1, especially Lurrus. Although Lurrus improved a lot of interactive decks like Jund and Grixis Delver, it had a disporportionate impact on strategies like Prowess, Devoted Devastation, Hardened Scales, and especially Burn. These four decks collectively gained 16.1% points in the metagame. Even Jund, a staple Lurrus deck, only increased by 3.8% relative to the pre-IKO picture. When you add in all the losses in Urza and Astrolabe decks generally, the overall decline in these midrange/control options made sense. Doubly so because Lurrus was a much more natural fit in aggro and creature-based combo, not just grindy BGx Midrange.

Speaking of Lurrus, here were some top-level companion statistics from Week 1.

  • % of total decks using any companion: 57.7%
  • % of total decks just using Lurrus: 47.4%
  • % of total decks using other companions: 10.3%
  • % of top-tier decks using any companion: 63.1%
  • % of top-tier decks just using Lurrus: 54%
  • % of top-tier decks using other companions: 9.1%
  • % of all companion decks just using Lurrus: 82.2%

It was tempting to characterize Modern as a “companion format,” but these numbers really show Week 1 was more of a “Lurrus format” specifically. Lurrus was rampant, with just under 50% of all decks across MTGO using Lurrus and just over 50% of top-tier decks pairing with the Cat companion. 13 of the 17 top-tier decks used companions in some form. Eight of those decks were distinctively Lurrus decks, and if we isolated the Tier 1 players (those with >4% of the format), we saw four of the seven focusing around Lurrus. None of this is to villify our new smoking, four-eyed overlord, as I think there are some real deck diversity benefits Lurrus is promoting. At the same time, it is important to highlight this disproportionate and rapid format impact.

I concluded last week’s article with five critical dataset limitations. We should never draw conclusions on any metagame pictures, let alone Week 1 following a new set, without remembering these reminders. Carry all of these over to today:

  • Week 1 metagames almost always change.
  • Players love experimenting with new cards.
  • Hyped cards can promote echo chambers.
  • Top-tier regulars are boring in new metagames.
  • N was comparatively small.

Delver of SecretsGiven those disclaimers, I concluded Week 1 Modern remained strategically diverse from an archetype and deck perspective, at least on paper. I gave a caution about high companion prevalence but admitted Lurrus was driving both existing top-tier options and reinvigorating old stalwarts (gogo Grixis Delver!). Finally, I emphasized Lurrus’ extremely high prevalence as both an observation about the metagame and a warning about potential Wizards action on such a prevalent staple. It’s impossible to not consider bans when we’re discussing a card that sees play in 50%+ of all top-tier decks, even if I am utterly exhausted by all the ban talk and ban announcements over the last 1.5 years.

We ended Week 1 with cautious optimism about the resurgence of old strategies and overall format balance. We also ended with general cautions about Lurrus’s metagame footprint and worries about companions collapsing into a few best options. It would be helpful if Week 2 of Ikoria Modern could provide clear answers to these pressing questions, but as we’re about to see, Week 2 just complicates an already complicated picture.

Post-Ikoria Modern: Week 2

Going into Week 2, I was monitoring two trends. First, I wanted to verify if a companion metagame remained a strategically diverse metagame. Second, I wanted to see if Lurrus stayed the driving Modern force or if other companions emerged as competitors. We may not have answers to our overall metagame questions, but Week 2 definitely offered strong indicators of where the format is trending. Unsurprisngly, companions are leading that march forward.

Metagame breakdown

Our event and deck sample increased to 443 in Week 2, representing a solid 200 additional datapoints relative to Week 1. We’re still including Preliminaries both to boost N and because Prelims continue to have relevant overlap with the Challenge/Premier/SQ metagame; 12 of the approximately 17 top-tier decks overlap between major events and the smaller Prelims. I will continue to include these events until I start to see big differences in their pools. With that in mind, here’s our metagame heading through the end of Week 2 noting all gains and losses:

Post-IKO Metagame: 04/18 – 05/03/2020 (n=443)

  1. Burn: 13.1% (58) ▼2.3%
  2. Jund: 6.8% (30) ▼1.5%
  3. Prowess: 5.9% (26) ▲1.2%
  4. Humans: 5% (22) ▼0.5%
  5. Devoted Devastation: 5% (22) ▲0.7%
  6. Bant Snow Control: 4.7% (21) -0%
  7. Temur Urza: 4.7% (21) ▲1.9%
  8. Amulet Titan: 4.3% (19) ▲0.3%
  9. Hardened Scales: 3.6% (16) ▲0.4
  10. Ponza: 3.4% (15) ▲1%
  11. Ad Nauseam: 2.9% (13) ▲0.5%
  12. Eldrazi Tron: 2.3% (10) ▼0.1%
  13. Grixis Delver: 2.3% (10) ▼1.3%
  14. The Rock: 2.3% (10) ▼0.5%
  15. Dredge: 2% (9) ▼1.6%
  16. 5C Niv: 1.8% (8) ▼0.2%
  17. Neobrand: 1.8% (8) -0%

Lava SpikeAlmost every deck on this list switched positions with the notable exceptions of Burn and Jund. Despite drops for both decks, they remain on top. Other contenders maintained their metagame share (Bant Snow, Neobrand), but saw relative position changes. Bant Snow dropped to 6th from 4th as Prowess and Devoted Devastation climbed the rankings. Neobrand was just under the Tier 2 threshold last time, but a four way tie among the 1.4% decks (4C Snow Control, Golgari Titan, Infect, Grixis Shadow) snuck in the Griselbrand combo deck on a technicality. Neobrand has also replaced 4C Snow, the only deck that actually fell out of Tier 1 and Tier 2 over the week. In between these shifts, we also see the usual Modern suspects of Humans, Urza, Titan, Ponza, and others.

Here are some additional observations on Modern’s current top tier:

  • Tier 1 is wide open. We’d define Tier 1 decks as all those with >4% share, and we see an impressive eight decks meeting that criterion. Incidentally, Tier 1 also represents a healthy mix of archetypes (more on archetype balance shortly). This suggests the overall format is healthy, at least from a strategic diversity perspective. We’ll talk about companion impact later.
  • Top-tier Modern has an encouraging mix of format regulars and hungry upstarts. As in the Week 0 picture, this includes; classic Jund, Burn, and Titan decks; contemporary favorites like Prowess, Devoted Devastation, Urza, and Bant snow; and throwbacks like Ponza, Grixis Delver, The Rock, and Ad Nauseam. Again, this underscores widespread diversity across the top-tier landscape.
  • The lower end of Tier 2 is unstable. There’s a weird four-way tie between 4C Snow, Golgari Titan, Infect, and Grixis Shadow just under Neobrand and 5C Niv. I doubt all four of these decks are truly Tier 2, but I suspect one or more will beat out the rest to join the top-tier club. I’d bet on Titan (a deck seeing less play because Moderners want to play companions, not boring Prime Time), and a Shadow variant (Grixis or otherwise; the collective Shadow share is about 2.7% and may narrow to one Lurrus-powered option).
  • Ad Nauseam feels like post-IKO Modern’s Ponza. Like Ponza, the combo veteran is a weird throwback to an older Modern era taking advantage of new cards (Thassa’s Oracle). It also benefits from a grindy companion metagame where players eke out advantages from Lurrus Bauble recursion, Yorion Astrolabe draws, or just the virtual 8th card edge. Neobrand is likely benefitting from similar dynamics; keep your eye out for other decks which can exploit the companion grind.
  • RIP Mono G Tron. I truly don’t remember the last time I did a major metagame breakdown where Mono G Tron was absent. It’s also not like Tron is barely out of Tier 2 reach. It’s way down the list scavenging for metagame points with Bogles, Yorion Chord, and frikkin Goblins. This doesn’t necessarily mean Gx Tron is out of the picture for good, as I suspect the deck is better than its share reflects. But for now? Wow. Cya. Tron.

The overall archetype breakdown reflects this diversity. Note winners and losers relative to Week 1:

  1. Aggro: 35.4% (157) ▼3.3%
  2. Midrange: 24.8% (AA 38 + 72) ▲1.1%
  3. Combo: 15.3% (AA 3 + 65) ▲1.9%
  4. Control: 11.3% (AA 39 + 11) ▲1%
  5. Big mana: 10.4% (46) ▲.9%
  6. Tempo: 2.7% (12) ▼1.6%

Zirda, the DawnwakerAggro’s sizable decline (which is almost entirely from Burn and Dredge) has opened up the format to small shifts across other archetypes. Combo was this week’s big winner, reflected in a gaggle of Zirda, Umori, and Lurrus creature combo strategies emerging across events. Midrange and control are also trending up, driven in large part by an increased willingness to experiment with Yorion as a companion. Big mana enjoyed a parallel rise as players realize Titan decks remain amazing, even if they may feel a bit dull relative to the new companion hotness. My only red flag here is the tempo decline. I fear Lurrus players will shift away from UBx Delver strategies and revert to either Shadow decks (still a plus for diversity) or to more tested aggro/combo Lurrus builds (less healthy). This is an important balance to monitor; are Lurrus players coalescing around a few best options or using the Cat to promote underdogs decks?

As a whole, there aren’t too many warning signs in this metagame breakdown alone. The format looks pretty sweet with a lot of exciting options for players across the strategic spectrum. Unfortunately, as with last week, the Bird Serpent’s-eye view of a spreadsheet misses all the companions playing on the battlefield below. We might be comfortable with this companion dominance if it means a healthy archetype breakdown like we’re seeing above, but we still need to acknowledge their share to honestly assess Modern.

Companion breakdown

Week 1 saw Lurrus break records for fastest and widest-reaching clawprint on top-tier decks. Week 2 is more of the same, with a few other companions gaining a foothold on the edges of Lurrus’s throne. Here are overall companion statistics comparing Week 1 to Week 2 increases and decreases across the broader metagame:

  • % of total decks using any companion: 60.9% ▲3.2%
  • % of total decks using just Lurrus: 44% 3.4%
  • % of total decks using other companions: 16.7% ▲6.4%

Yorion, Sky NomadNo one should be surprised with companions trending up or featuring in 61% of Modern. Lurrus, Yorion, and company are just such powerful additions to so many decks that the only decks which aren’t using companions probably can’t: see Titan, Tron, and Dredge strategies. The sole exception to this is Humans, which can easily run Jegantha but is actually ghosting the poor Elk and sticking with its old-school core. Don’t be fooled by Lurrus’s slight dip either. Lurrus remains a major format force even if the +6.4% increase in other companions is an interesting anecdote. This is particularly true because companions, especially Lurrus, remain extremely dominant in the top-tiers. Positive/negative shifts track changes from Week 1 to Week 2 in both the top decks…

  • % of top-tier decks using any companion: 63.8% ▲1.9%
  • % of top-tier decks using just Lurrus: 49.7% ▼2.9%
  • % of top-tier decks using other companions: 14.2% ▲4.9%

…and in Challenge, SQ, and Premier Top 8s:

  • % of T8 decks using any companion: 63.8% ▼2.3%
  • % of T8 decks using just Lurrus: 50% ▼1.8%
  • % of T8 decks using other companions: 13.8% .5%

If you’re tired of companions, these statistics deliver some good news. Lurrus’s share has declined across not only the entire metagame (-3.4%) but also in both top-tier (-2.9%) and Top 8 (-1.8%) lists. This suggests Lurrus’s current share may overrepresent its true power. We’re also seeing a small downtick in companion T8 prevalence, suggesting players have realized they don’t need the virtual 8th card to succeed.

Obosh, the PreypiercerThere’s also bad news. Despite the small downward changes in companion shares, the discrete shares themselves remain extraordinarily high. Half of all T8 decks are using Lurrus in the first two weeks. More than 60% of all top-tier and T8 decks are using companions, and Lurrus is seeing slight, increasing shares (44% –> 49.7% –> 50%) progressing from all decks to T8 decks. This suggests additional overperformance on the Cat’s part. Contrast with other companions which collectively see decreasing shares from the format-wide picture into T8s (16.7% –> 14.2% –> 13.8%). Despite scattered success of cards like Obosh, it’s possible the other companions just aren’t as good as Lurrus, and/or players haven’t figured out how to use them. To repeat an earlier disclaimer, none of this is to say this is an actionable problem. We still have reassuring archetype diversity. At the same time, we also need to be calling out these unprecedented shares.

On the topic of top-tier companion prevalence, the table below shows companion usage by deck and the percentage of those decks running a preferred critter in crime. As an example of reading this table, 93.3% of Jund players are running Lurrus, but only 36.4% of Humans players are running Jegs. Companion % /s indicate changes in the percentage of a specific deck using companions relative to Week 1 (i.e. 36.4% of Humans players ran Jegantha, down 6.5% from Week 1):

Rank

Deck

Meta %

Comp %

Comp of choice

1 Burn 13.1% 91.4% ▲1.7% Lurrus
2 Jund 6.8% 93.3% ▲2.8% Lurrus
3 Prowess 5.9% 84.6% ▲9.6% Lurrus
4 Humans 5.0% 36.4% ▼6.5% Jegantha
5 Devoted Devastation 5.0% 86.4% ▲4.6% Lurrus
6 Bant Snow Control 4.7% 19% ▲10.7% Yorion
7 Temur Urza 4.7% 90.5% ▲19.1% Yorion
8 Amulet Titan 4.3% 0% (same)
9 Hardened Scales 3.6% 100% (same) Lurrus
10 Ponza 3.4% 66.7% ▲33.4% Obosh
11 Ad Nauseam 2.9% 0% (same)
12 Eldrazi Tron 2.3% 0% (same)
13 Grixis Delver 2.3% 100% (same) Lurrus
14 The Rock 2.3% 100% (same) Lurrus
15 Dredge 2.0% 0% (same)
16 5C Niv 1.8% 50% ▼10% Jegantha
17 Neobrand 1.8% 0% (same)

For the most part, every deck previously using companions is still using companions. We’re just seeing more of these decks sticking with the companion formula for success than in Week 1. Some represent neglibile upticks from already high numbers: 91.4% of Burn players use Lurrus up 1.7% points from Week 1. Others are more significant. In particular, Yoion sees much more Week 2 play than the Sky Nomad enjoyed in Week 1. Yorion is a virtual Temur Urza staple with 90% of these decks using the companion, and even holdout Bant Snow Control is trending up with about 20% of Bant Snow mages adopting Yorion. Similarly, Obosh is rising as a Ponza mainstay, and I expect we’ll see all of these numbers keep climbing into Week 3 and beyond.

Jegantha, the WellspringThen there’s Jegantha. Poor Jegs. Only two decks are using the big, dumb Gurmag Angler imitator, and both have seen marked dips in their willingness to give up the sideboard slot. Humans is a fascinating case study in this regard. This is a Top 5 deck that can use Jegantha and is simply choosing not to. Same with 5C Niv, even though Niv Mizzet looks like the ideal home for the 5C Elemental! Maybe Jegantha will reverse its fortunes, but for now, it looks like a “free” 5/5 just isn’t enough to keep the Elk on rosters.

Finally, for those who can’t get enough companion and want to see how Lurrus’s and Yorion’s less successful cousins are competing, here’s a format-wide companion breakdown as we wrap Week 2:

Companions

Overall %

Top-Tier %

T8 %

Lurrus 44% (195) 49.7% (158) 50% (40)
Yorion 9% (40) 7.9% (25) 8.8% (7)
Jegantha 2.9% (13) 3.1% (10) 3.8% (3)
Zirda 1.6% (7) 0% (0) 0% (0)
Obosh 2.5% (11) 3.1% (10) 1.3% (1)
Gyruda 0.5% (2) 0% (0) 0% (0)
Umori 0.2% (1) 0% (0) 0% (0)
Lutri .2% (1) 0% (0) 0% (0)
ALL COMPS 60.9% (270) 63.8% (203) 63.8% (51)

Umori, the CollectorBlah blah Lurrus Yorion blah. Look at all the cool showings by Zirda, Obosh, and even lovable ooze friend Umori! There are significant gulfs between top companions and those underneath (and even within that, between Tier S Lurrus and everything else), which indicates players are still tinkering with these cards. I had the same reaction as Bob Huang to the amazing Umori experiment that stole 10th in the 05/03 Challenge, and I’m hoping more decks play around with companions. I fear Lurrus will continue to dominate the conversation because it’s often better to just play a Lurrus combo deck than a Zirda variant, or a Lurrus aggro deck than an Umori one, but maybe the brewers will prove me wrong.

Taken as a whole, Lurrus (and other companions to a far lesser extent) remain decisive Modern influencers. As in last week’s update, it’s hard to compare this current Modern epoch to a previous period; I don’t remember a time where a card had such a titanic format imprint but still empowered an overall diverse range of decks. This puts us in a challenging position when assessing metagame health and this is where we are stuck heading into Week 3.

Week 2 Takeaways

My Week 1 disclaimers are still in effect as the format continues to take shape. In particular, even though we enjoy a respectable N=443 sample, we still need a lot more data and time to really see where Modern is heading. Disclaimers aside, here are my biggest conclusions from the Week 3 picture:

  • The format remains strategically diverse. You can play all archetypes, as well as different decks within archetypes. At least from a high-level perspective, the Modern metagame is healthy.
  • Lurrus is still everywhere. This is somewhat true of other companions, but it’s mostly just Lurrus. There are more decks playing Lurrus now than played OUaT when that abominable cantrip got banned. At the same time, Lurrus has a very different metagame impact than the Eldraine design mistake and appears to support many distinct, diverse options, not just a herd of big mana bruisers. This makes it harder to positive or negatively assess Lurrus’s impact.
  • Companion shares are increasing. Players are getting better and bolder at building around these powerful cards, and we will continue to see breakout performances as these decks keep improving.
  • Tier 1 is crystallizing, but Tier 2 remains open. It’s hard to imagine a Tier 1 that looks significantly different from our current Big 7 (Burn/Prowess, Jund, Humans, Devoted Devastation, Bant Snow, Temur Urza, Amulet Titan). But it’s easy to see how Tier 2 and 3 decks will continue to scrap for top-tier presence. Perhaps the biggest indicator of this is Gx Tron’s steep decline, a deck so powerful it is likely to return… even if it’s currently struggling in the muck with Bogles and Goblins.

Weeks 3 and 4 will clarify our metagame picture and give us a better sense of where things are heading. Expect more shifts like Burn dropping, Yorion rising, and Amulet Titan staying exactly where it is as the format progresses.

Bring on Weeks 3 and 4!

I can’t guarantee I’ll have the bandwidth to crank out another metagame update next week, but you can bet we’ll be back for more analysis as May continues. We’re also still waiting for Wizards to release a Modern format vision update, and to generally address many of the sweeping Constructed issues rankling every Discord and Twitter feed online. Personally, I’ll be happy once we get Forsythe’s much-anticipated and badly-needed Modern update. Any day now. Please?

Let me know in the comments here, on Reddit, on Twitter, or on the MTG Nexus Modern subforum if you have any feedback for the article. Ideas, criticism, thoughts, and corrections are always welcome, especially corrections if my Excel automation missed some numbers. See you all soon and until next time, keep your socially distanced friends close and your Modern companions closer.

Modern Metagame Update: Post-IKO Week 1

The last time I wrote a true metagame update was for Modern Nexus. Back then, we were emerging from the darkness of Eldrazi Winter with a Tier 1 that included decks some darn kids these days ain’t even heard of: Affinity, Jeskai Control, Abzan Company, etc. Today, however, we’re just up to our necks in cats. Plus birds, elks, and whatever the heck you call Gyrudas and Oboshes. Ikoria’s new companion mechanic has riled the online Modern community into a frenzy and redefined every Constructed format where it is legal. I don’t remember the last time I saw a ban-related Reddit thread hit almost 1,000 upvotes, let alone in the first week of a new set. Players are clearly concerned about the metagame and companion’s impact on every format from Standard to Vintage. This has led to intensive, ongoing debate about the mechanic itself and whether or not it violates fundamental Magic rules. Unfortunately, there has been one element missing from most of the discussion, especially in Modern: clear metagame data.

I normally post Challenge and Super Qualifier (SQ) breakdowns to the Modern subreddit with a mini write-up. These are generally well-received and always spark interesting conversation. In light of the companion crisis, however, I think it’s much more important to bring back Modern Nexus-style metagame breakdowns. It’s impossible to have meaningful conversations about the format without understanding its landscape. Today, I’ll show the pre-IKO picture before companions hit, present the changed format using non-curated MTGO data from Week 1, and offer takeaways about the format’s health and direction.

Pre-Ikoria Modern

It feels like it’s been over a month since we first met Lurrus, Yorion, and company, but it’s actually been just eight days since we got our post-IKO Challenge results last Sunday. The COVID-19 quarantine time warp has never felt so real. As someone who regularly posted pre-IKO MTGO results, I’ve been in a unique position to notice overall metagame shifts, not just the prevalence of companion. Community members must also understand these shifts to have meaningful discussions about where Modern is heading.

Let’s start with a pre-IKO Modern metagame breakdown. This includes Challenges, SQs, and Premiers on MTGO, totaling 11 events from 03/21/2020 (the first date of published SQ data) through 04/11/2020. There are 352 decklists in this sample. I originally only included SQs under the assumption Challenges were less competitive and couldn’t be merged with SQs, but as I assessed in a separate MTG Nexus forum post, this assumption actually didn’t withstand scrutiny. Challenges actually had fewer Tier 3 spice/jank than SQs while still having the same top-tier representatives. Because of this, I think it’s appropriate to combine the two both to increase N and capture a fuller metagame picture. Here are the Tier 1 and Tier 2 decks of the previous Modern era representing about 75% of the entire format.

Pre-IKO Metagame: 03/21 – 04/11/2020

  1. Bant Snow Control: 11.6% (41)
  2. Ponza: 8.2% (29)
  3. Dredge: 7.4% (26)
  4. Burn: 6.5% (23)
  5. Eldrazi Tron: 5.1% (18)
  6. Temur Urza: 5.1% (18)
  7. Humans: 4.8% (17)
  8. Prowess: 4.8% (17)
  9. Jund: 4.5% (16)
  10. Amulet Titan: 4% (14)
  11. Mono G Tron: 3.7% (13)
  12. Infect: 3.7% (13)
  13. 5C Niv: 2.6% (9)
  14. Death and Taxes: 2.3% (8)
  15. Bant Snowblade: 2% (7)

This is what Modern looked like just over a week ago. Arcum’s Astrolabe (AA), big mana, Jund/Ponz midrange, and aggro defined this format, with virtually all strategies (just over 80%) fitting into one of those categories. Extending the analysis down to Tier 3 and lower strategies (e.g. Infect, Simic Nexus, Ad Nauseam, Mono U Tron, etc.), we can categorize all decks by macro-archetype to capture overall format trends:

  • Aggro Decks: 30.3% (103)
  • AA Decks: 28.2% (96)
  • Midrange Decks: 17.1% (58)
  • Ramp Decks: 16.5% (56)
  • Combo Decks: 5% (17)
  • Non-AA Control Decks: 2.9% (10)

I know many Modern players bemoaned a metagame they experienced as Astrolabe vs. Ramp, but the MTGO reality was more diverse. Pre-IKO Modern had a ton of strategic representation, even if the power-level of deck staples and matchup deciders was incredibly high. This probably led to a feeling of swingy gameplay even if the metagame itself appeared balanced on a spreadsheet. As a whole, here were a few takeaways from this older format:

  • Arcum's AstrolabeBant Snow was probably the secret, not-so-secret best deck, coming in about 3% ahead of the next best options. This notion of a “secret, not-so-secret” best deck is a recurring concept in my writing. It denotes a deck that is not so obviously the best deck where everyone plays it (e.g. Hogaakvine, Simic Oko Urza, Eldrazi, etc.), but is probably the best from a match-win-percentage (MWP) standpoint (e.g. Twin, KCI, Phoenix, etc.).
  • All strategies saw representation, but some had fewer options. I’d characterize this as healthy but somewhat restrictive. Want to play ramp? E Tron, G Tron, and Titan variants await. Want to play aggro? Take your pick of Burn, Prowess, Humans, Dredge, and others. Midrange? You’re mostly stuck on Jund and Ponza. And control? It’s Bant Snow or bust, unless you want a midrange/control hybrid in which case it’s Temur (Snow) Urza or bust.
  • All fairer, more interactive decks played green. The sole exception was Death and Taxes, eking into Tier 2 behind another Astrolabe variant. This isn’t necessarily an issue (Legacy is the blue format, maybe Modern is the green format) but it’s an important observation about Modern’s direction.

Overall, there were both advantages and disadvantages to this metagame. You could play any archetype you wanted, but players sensed a huge power gap between the best decks and the lower-tier options. You could enjoy a range of archetype vs. archetype matchups, but it often felt like swingy haymakers disproportionately influenced game outcome. There was also the Ponza factor. A longtime low-tier player, the Gruul upstart rose through the ranks to become the second most-played deck. Depending on who you asked, that was either metagame adaptation at its finest or an indicator of a warped format. There wasn’t a lot of consensus around this old metagame and I know many players, myself included, were looking for a little bit of a shakeup to see if Ikoria could change the picture. As they say, careful what you wish for.

Post-Ikoria Modern: Week 1

Enter companions. Or, as we’ll see later, enter Lurrus. Although IKO definitely brought us a variety of new Modern playables (check out last week’s Top 25 playable countdown on this site: a fair number of hits!), it’s misleading to compare something like the solid General Kudro to the meteoric impact of companions. And when I say “companions,” I really mean Lurrus of the Dream Den. I don’t remember the last time a single card and its core mechanic had such format-wide penetration. Companions (read: Lurrus) are absolutely everywhere. Our job today is to assess the overall metagame, evaluate the prevalence of companions specifically, and gauge our comfort with this picture.

Metagame Breakdown

Week 1 of Ikoria Modern saw only a few Challenges, SQs, and Premiers, even counting the SQ and Challenge published yesterday. We can increase N by including Preliminary results. To be very clear, this is not going to give us a perfect apples-to-apples comparison with the pre-IKO metagame. Our earlier sample did not include any Preliminaries and only encompassed major events. I’m fine with this for a few reasons. For one, our biggest priority in Week 1 is not necessarily predicting the most competitive metagame in Week 7. It’s establishing an accurate picture of the metagame right now. In that regard, we want to increase N using as many noncurated MTGO events as possible, and Preliminaries show all 5-0 through 3-2 decks. This makes them a natural dataset fit. Leagues, however, are a delibeately filtered list of 5-0s Wizards uses to show off different ideas: we would never include these curated events. Second, and perhaps more importantly, the approximately 75 Prelim decks in this first week mostly align with the SQ/Challenge/Premier picture. Of the top 20 decks in the Prelim-only sample, 15 overlap with the top 20 of the SQ/Challenge/Premier sample. This makes me reasonably confident we can merge the two, acknowledging a bit of rogue deck noise.

Here’s the post-IKO metagame across our 11 events and N=253 decks:

Post-IKO Metagame: 04/18 – 04/25/2020

  1. Burn: 15.4% (39)
  2. Jund: 8.3% (21)
  3. Humans: 5.5% (14)
  4. Bant Snow Control: 4.7% (12)
  5. Prowess: 4.7% (12)
  6. Devoted Devastation: 4.3% (11)
  7. Amulet Titan: 4% (10)
  8. Dredge: 3.6% (9)
  9. Grixis Delver: 3.6% (9)
  10. Hardened Scales: 3.2% (8)
  11. Temur Urza: 2.8% (7)
  12. The Rock: 2.8% (7)
  13. Ponza: 2.4% (6)
  14. Eldrazi Tron: 2.4% (6)
  15. Ad Nauseam: 2.4% (6)
  16. 5C Niv: 2% (5)
  17. 4C Snow Control: 2% (5)

Monastery SwiftspearAlthough this is still a visibly Modern format with tons of recognizable decks, Ikoria has brought a lot of changes. Burn has dethroned Bant Snow. Jund has replaced Ponza as the midrange deck of choice. Devoted Devastation has risen up from complete irrelevance to a top six deck, and some bygone Modern decks are back in the fight: Hardened Scales, Grixis Delver (!), The Frikkin Rock (!!). We also see some decks remaining relatively stable, such as stalwart Humans, trusty Prowess, and the veteran Amulet Titan. As a whole, this is mostly a healthy assortment of decks and archetypes, even if the exact rankings have changed from just a weekend ago. Burn is definitely higher than we’d want to see any single deck, but a) Burn is often a great Week 0/Week 1 choice in any metagame, and b) people want to play the companions and Burn is a natural home for Lurrus. These factors could be artificially increasing Burn’s share so I’m not too concerned about that 16% for now.

The table below summarizes shifts in the pre- to post-IKO metagames. It lists all Tier 1 and Tier 2 decks from both metagames (the four at the end were only in the pre-IKO top 15). It then shows the change (delta, ▲) in rank and percent share. Darker blue indicates a bigger increase, darker orange indicates a bigger drop.

Rank

Deck

Rank ▲

% ▲

1 Burn +3 +8.9%
2 Jund +7 +3.8%
3 Humans +4 +0.7%
4 Bant Snow Control -3 -6.9%
5 Prowess +3 +0.1%
6 Devoted Devastation +27 +4.1%
7 Amulet Titan +3 +0.0%
8 Dredge -5 -3.8%
9 Grixis Delver New! +3.6%
10 Hardened Scales New! +3.2%
11 Temur Urza -5 -2.3%
12 The Rock +15 +2.5%
13 Ponza -12 -5.9%
14 Eldrazi Tron -9 -2.7%
15 Ad Nauseam +8 +1.5%
16 5C Niv -3 -0.6%
17 4C Snow Control +12 +1.4%
Pre-IKO Mono G Tron -10 -1.7%
Pre-IKO Infect -8 -1.7%
Pre-IKO Death and Taxes -7 -0.3%
Pre-IKO Bant Snowblade Gone! -2.0%

Insectile AberrationBig winners are Burn, Devoted Devastation, Jund, Grixis Delver, and Hardened Scales. Delver and Scales are particularly exciting, as the latter hasn’t been viable since Opal’s ban four months ago, and the former hasn’t been viable since Gitaxian Probe ate the banhammer in 2017. Incidentally, every single one of those decks is playing Lurrus; more on that later. The big losers are Temur Urza, Eldrazi Tron, Dredge, Ponza, and Bant Snow Control. Most of these decks do not use companions at all, although both Temur Urza and Bant Snow are experimenting with Yorion, while Ponza tinkers with Obosh.

Despite the clear influence of companions, and the overrepresentation of Burn, it’s still important to emphasize Week 1 presents a mostly healthy metagame picture with multiple viable strategies. Here’s our rough archetype breakdown with changes relative to the pre-IKO breakdown noted in blue/orange:

  • Aggro Decks: 38.7% (98), +8.4%
  • Midrange Decks: 17.4% (44), +.3%
  • AA Decks: 15% (38), -13.2%
  • Non-AA Control Decks: 12.3% (31), +9.3%
  • Ramp Decks: 9.5% (24), -7%
  • Tempo Decks: 4.3% (11), +4.3%
  • Combo Decks: 2.8% (7), -2.2%

Aggro is up by basically the same margin Burn increased. Astrolabe decks are way down with a ton of non-Astrolabe control entering the picture. We’re also seeing the return of true Delver tempo decks (as opposed to Shadow decks, which are also back but folded into the midrange category), which are up 4.3% from a fat 0% in the old metagame. In addition to Astrolabe’s lost ground, big mana decks are also down. With both strategies struggling to adopt Lurrus and pals, players may be shifting away from these format standbys to hot new Ikoria tech.

Overall, the high-level metagame review is relatively positive. If this was any other Week 1 of Modern, there would be nothing remarkable about these Tier 1 and Tier 2 standings. Burn might warrant a footnote about its overrepresentation, but it’s nothing we wouldn’t expect to change as the metagame matures. Unfortunately, this higher level deck breakdown misses the cards that are really driving the changes: IKO’s companions and all those dang Lurrus copies.

Companion Breakdown

Up until this point, I’ve seen a lot of conversation about companions as a general mechanic. I was doing this too but today’s analysis exposed a gap between the prevalence of all companions not named Lurrus, and Lurrus itself. This section will show the huge divide between companion/Lurrus decks and those not running the Nightmare Cat.

Here are summary metagame and companion statistics to describe the Week 1 format. For reference and the sake of simplicity, “top-tier” refers to all Tier 1 and Tier 2 decks.

  • % of total decks using any companion: 57.7%
  • % of total decks just using Lurrus: 47.4%
  • % of total decks using other companions: 10.3%
  • % of top-tier decks using any companion: 63.1%
  • % of top-tier decks just using Lurrus: 54%
  • % of top-tier decks using other companions: 9.1%
  • % of all companion decks just using Lurrus: 82.2%

Lurrus of the Dream DenThat’s a lot of Lurrus. I emphasize Lurrus individually because the overall prevalence of non-Lurrus companions is relatively tame. Lurrus sees twelve times (!!) more play than the next most-played companions in Jegantha and Yorion across Week 1 Modern. More than half of the top-tier decks are running just Lurrus alone with only 9.1% running some other companion. To put that in perspective, there are more decks running Tarmogoyf in Week 1 than are running non-Lurrus companions. This is a critical observation because it shifts our Week 1 scrutiny off Yorion, Jegs, and the companion mechanic generally towards Lurrus.

Of course, this aggregate analysis misses the distinct prevalence of companions/Lurrus within specific decks. Amulet Titan and Dredge can’t run any of the current companions without major changes. On the other hand, about 90% of Burn and Jund decks rely on Lurrus to enjoy their Week 1 success. And then there’s Hardened Scales and Grixis Delver, which wouldn’t be here at all if Lurrus wasn’t joining their 75. Aggregating these decks miss individual Lurrus/companion impact.

To capture these distinctions, the table below presents post-IKO top-tier decks, their prevalence, and the percentage of those decks running both a companion and Lurrus specifically. As an example of how to read the table, 42.9% of Humans decks ran a companion but 0% ran Lurrus; they opted for Jegs instead. For Devoted Devastation, however, 100% of its pilots used a companion, and all of those were Lurrus.

Rank

Deck

%

Companion %

Lurrus %

1 Burn 15.4% 89.7% 89.7%
2 Jund 8.3% 90.5% 90.5%
3 Humans 5.5% 42.9% 0.0% (Jegs)
4 Bant Snow Control 4.7% 8.3% 0.0% (Yorion)
5 Prowess 4.7% 75.0% 75.0%
6 Devoted Devastation 4.3% 81.8% 81.8%
7 Amulet Titan 4.0% 0.0% 0.0% (none)
8 Dredge 3.6% 0.0% 0.0% (none)
9 Grixis Delver 3.6% 100.0% 100.0%
10 Hardened Scales 3.2% 100.0% 100.
11 Temur Urza 2.8% 71.4% 0.0% (Yorion)
12 The Rock 2.8% 100.0% 100.0%
13 Ponza 2.4% 33.3% 0.0% (Obosh)
14 Eldrazi Tron 2.4% 0.0% 0.0% (none)
15 Ad Nauseam 2.4% 0.0% 0.0% (none)
16 5C Niv 2.0% 60.0% 0.0% (Jegs + Yorion)
17 4C Snow Control 2.0% 100.0% 100.0%

Yorion, Sky NomadAgain, that’s a lot of Lurrus. Of decks that tried to run Lurrus at all, we’re seeing 92.1% of them on average sticking with the Cat. Other companions aren’t having this kind of effect except Yorion in Temur Urza (71% of Urza pilots are rocking the Sky Nomad). It’s true there are a few decks which can succeed without any companion. Bant Snow Control is crushing it at 4th place with only 8% of decks trying out Yorion (for now…). Amulet Titan and Dredge don’t even need fancy IKO nonsense to stay in the top 10. But these decks are the exception and not the rule. 13 distinct archetypes in the top 17 are using companions in some capacity. 10 of the top 17 see more than half of their players trying out companions. Of our seven Tier 1 decks, which we’d roughly define as those with >4% shares, four of them are Lurrus decks including the top two, which incidentally have a higher metagame share (23.7%) than the next five decks combined (23.3%).

All of this points to Lurrus redefining the Modern format in the span of a week. I hesitate to call this a totally “new” format because all of the decks are recognizably Modern decks on paper. At the same time, Lurrus is clearly presenting new dynamics Modern has never seen in the past. This pushes us to assess our comfort with Lurrus and the new metagame reality.

Week 1 Disclaimers and Takeaways

I condensed my pre-IKO conclusions into a few bullet-points in the first section. Our post-IKO picture needs a more thorough treatment. Before we get into takeaways, we need to talk about some metagame forces which influence the numbers. We must remember these qualitative caveats before making strong statements about post-IKO Modern:

  • Week 1 metagames almost always change. Players don’t know the best cards, best decks, and where the format is heading. This is true in both genuinely open metagames and formats that are secretly solved. As an extreme example of this, check out Modern just before Pro Tour Oath Eldrazi. Not only did its lone Eldrazi variant have just 5.7% of the metagame, it was probably the worst variant of the lot, a Bx style built around cute Wasteland Strangler synergies over outrageous Eye/Temple/Mimic openers.
  • Players love experimenting with new cards. When new sets enter a stagnant environment, everyone wants to try out the exciting technology. This is why we saw a huge uptick in Stoneforge decks immediately after her unbanning and a quick decline when people realized SFM was merely okay in a much more powerful format.
  • Hyped cards can promote echo chambers. In the absence of high viewership tournaments, accessible written content, and ongoing Twitch coverage, players tend to default to whatever has the most social media buzz. We saw this effect in events like 2015’s PT Fate Reforged, where a huge segment of the pro community talked up the Abzan gameplan and willfully missed out on playing the best decks of the year: Amulet Bloom and URx Twin.
  • Top-tier regulars are boring in new metagames. We just exited a Modern period where most of the online community argued constantly about Astrolabe, Veil, ramp decks, etc. Who wouldn’t want a break from those tired decks to try out the new hotness?
  • N is comparatively small. This obligatory sample size disclaimer is particularly important in Week 1, where we have fewer events and need to include less appropriate events (i.e. Preliminaries) just to boost N.

These reminders will temper our conclusions and prevent us from making sweeping, overstated claims. I’m sure there are other data limitations we need to consider, and if I missed any major ones worth remembering, let me know and I can add them to future articles. With the disclaimers in mind, here are my big Week 1 takeaways:

  • Modern is strategically diverse. You really can play any archetype you want, as long as you are building around a few established pillars. Current pillars include Astrolabe, Tron lands, Titan, and, of course, Lurrus. Those four pillars represent a whopping 75% of the entire metagame, and 83% of the top-tier metagame. At the same time, those four pillars cross all archetypes and give you a lot of options. You can also play a linear strategy that doesn’t necessarily need any of these pillars at all (Humans, Dredge, Ad Nauseam, and Infect are currently in the vogue).
  • Lurrus is improving decks across the tiers. It’s both sharpening existing competitive decks (Burn, Jund) and breathing new life into dearly departed friends (Grixis Delver, Hardened Scales). This may be a net diversity positive if Lurrus keeps spreading the love. It becomes a net negative, however, if the Lurrus decks collapse into just 1-2 best options.
  • The exact metagame breakdown is going to change, but high companion prevalence is not. In fact, I suspect we have not reached peak companion levels yet. Lurrus is a relativey easy buildaround and players will continue to optimize their Dream Den decks. The other companions are harder, especially if fewer people are playing them. This forces brewers to create without the benefit of the hive mind or previous iteration. As Moderners continue to grind on MTGO, expect to see even more of the group, especially Yorion, Zirda, and Jegantha. Yorion in particular is poised for a big uptick as people figure out how to build manabases and cores in 80 card decks, realizing Yorion’s consistency offsets the usual risk of an 80 card strategy.
  • Lurrus. Is. Everywhere. I can’t emphasize this point enough because it’s such a breakout appearance by a new card. For now, it would be disingenuous to offer any Modern metagame analysis that doesn’t carefully track this companion’s prevalence.

Our biggest, lingering questions are about Lurrus. The companion is rampant but how rampant is too rampant? This notion of card prevalence has at the forefront of the last decade of competitive Magic. This is still true today. Prior to IKO there were a number of Reddit posts about the Modern prevalence of Veil of Summer and Lightning Bolt. One was serious, one was snarky, but both set a rough baseline for staple prevalence in Modern in the mid-30% range. Back when Once Upon a Time was legal, the defining cantrip saw play in about 40% of MTGO decks, higher than both Bolt and Veil individually. History offers other examples of problematic card prevalence, although they aren’t all Modern-specific. In the good old Caw Blade days, Aaron Forsythe famously noted that 88% of Day 2 decks at a Grand Prix had Jace, the Mind Sculptor copies. 70% had Stoneforge Mystic. More recently, we saw an Eldrazi Winter with 38% of major event Day 2 decks rocking Eldrazi strategies, peaking at 47% of Detroit and 40% of MTGO. These rough numbers provide important context for Lurrus’s current takeover.

The OzolithIn that regard, I’d classify Lurrus somewhere between an orange flag and a red flag issue. On the one hand, the overall metagame looks good, at least  in an Excel table. You can play a lot of strategies and those strategies cross the archetype divide. Modern sans Lurrus will likely lose a lot of these options (maybe Ozolith could keep Scales around?), especially if Astrolabe decks refine their Yorion options and Bant Snow returns to number one.

On the other hand, there are just so many decks using Lurrus that it’s hard to not draw comparisons to previous banlist announcements. There are more decks using Lurrus than OUaT. There are more decks using Lurrus than Eye of Ugin. It’s not clear if those comparisons are entirely appropriate because Lurrus has a different metagame impact than OUaT/Eye, but it’s still hard to ignore the comparable play rates. Personally, I don’t want to talk about bans anymore and avoided it in my Top 25 last week, but there are just so many people playing Lurrus/companions that it’s hard to avoid the topic. In that regard, Lurrus is either an orange flag issue (if you’re willing to accept the Lurrus/Bauble package if it means strategic diversity/balance), or a red flag issue (if you are reasonably fixated on Lurrus being in 63% of top-tier decks).

Either way, we need more data to see where the metagame heads. It’s going to be an interesting few months.

Modern: Lair of Companions

Wherever you see Modern going by May and June, it’s hard to deny companions will be a big part of its future. Especially Lurrus. It’s also hard for even the most ardent opponent of ban talk to avoid the issue when so many decks are using this card. If we can’t avoid the banning conversation altogether, at least metagame pieces like this will hopefully keep the debates informed.

Mishra's BaubleAs a final note before we end, there are a ton of weighty Modern issues this article either dances around or ignores completely due to our limited scope. Examples include the outrageous uptick in Standard power-level, the failures of Play Design to prevent pushed cards from breaking formats, Wizards’ alarming history of banning older cards when newer cards break a format (be afraid, Mishra’s Bauble, be very afraid), ongoing silence from Wizards representatives about multi-format Constructed concerns, continued data restrictions in thoroughly solved formats, and still no word on an updated Modern vision/mission. We have a lot of work ahead of us to address these issues and others. As long as we focus our arguments around researched cases and evidence, however, we’ll continue to make a strong case for change.

Let me know in the comments, on Reddit, or on MTG Nexus if there’s anything I missed or got wrong. See you all around the forums. While we’re all sheltering in place, where else are we going to debate our favorite format?