Here we are again: another set, another set of problems, and another banning... oh wait, nothing's getting banned in Modern. Awkward. And more to the point, a waste of a perfectly good rant setup. Why am I even bothering... oh right, something else happened this time around. And it is consequential. Though not in the way that I was expecting.

I said in last week's article that I was done talking about companions. It was getting boring for everyone, and I had run out of new insight anyway. That was completely thrown out the window a few hours later when Wizards announced another Banned and Restricted update. There would be no actual changes to Modern, just Standard and Historic. Because Arena is the new, expensive, and special child, and must be protected.

Instead, Wizards issued functional errata as a nerf to companion as a whole. Wizards often changes cards to either fix oversights or update cards as the rules change. However, fixing power level via rules changes is something Wizards avoids these days, largely because of the ridiculous saga of Time Vault. That it had to happen this time speaks to the fundamentally different nature of the problem.

The Decision

Everyone knew that something had to happen about companion. The data had become unequivocal. The problem (in Modern anyway) wasn't that one deck was dominating or that a particular card was obviously broken; it was saturation. Despite the deck and archetype diversity remaining high, companions were omnipresent. The same principle justified banning Once Upon a Time. Companion saturation had reached 76% of MTGO results, with the uptick being spread out amongst many different companions. Therefore, the only way to deal with the problem by normal means would be ineffective, as Wizards acknowledged:

This trend represents a long-term problem for the health and diversity of all formats. Rather than go down the path of making several individual adjustments to the banned list for each format, we feel the better solution is to reduce the advantage gained from using a companion across the board.

While the move to rework an entire mechanic after release is unprecedented, it wasn't completely out of left-field. As early as May 11, Mark Rosewater had signaled unhappiness with companion. And in the May 18 banning, Wizards mentioned:

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.

Rosewater confirmed that extraordinary action may be incoming later that same day. There were plenty of suggestions from players about a mechanical fix, but Wizards went their own way.

The Calculus

The other thing to note is how the decision was reached. Lurrus of the Dream-Den was banned in Legacy and Vintage for being demonstrably too powerful. Wizards justified not taking action on the other formats because:

Currently, these formats are shifting too quickly for data to indicate what, if any, card or archetype poses a problem.

...we're not currently seeing problematic win rates in Standard, Pioneer, or Modern from decks using companions...

The data I collected at that time didn't back up the latter quote, but the former, yes. Modern had greatly shifted and changed over the three weeks I surveyed. It started settling the following week, and became clear last week. Apparently, things were worse than I knew:

As a group, decks using companions have too high of win rates and metagame share in Standard, Pioneer, and Modern, and have already necessitated bans in Legacy and Vintage. This trend represents a long-term problem for the health and diversity of all formats. Rather than go down the path of making several individual adjustments to the banned list for each format, we feel the better solution is to reduce the advantage gained from using a companion across the board.

This section reaffirms that Wizards prioritizes win rates. More importantly, the line appears to have crystalized. In previous announcements, a 55% win rate in non-mirror matches was the cited red-line. Here, it continues to be so, constituting our best indication of typical banning criteria:

Over the course of the last several weeks, Fires of Invention decks have risen to have a dominant win rate and metagame presence in Standard, achieving a 55% win rate and having even or favorable matchups against each of the other top ten archetypes. This indicates that metagame forces alone aren't sufficient to keep the deck in check.

The Change

Which brings us to the actual change. And it's a surprisingly simple one.

Once per game, any time you could cast a sorcery (during your main phase when the stack is empty), you can pay 3 generic mana to put your companion from your sideboard into your hand. This is a special action, not an activated ability.

Companion is now a tutoring mechanic. Rather than directly cast a companion from exile, players must tutor it for three mana. Therefore, it's no longer a complete freeroll, and significantly impacts playability.

A note on special actions: these are abilities like turning face-up a morphed creature or paying for Leonin Arbiter. Special actions don't use the stack. Just pay the cost and do the thing; there's no way to respond and a player can do it whenever they have priority. Once priority changes, then opponents may respond. Therefore, I can't respond to my opponent tutoring for Lurrus and immediately casting it. Similarly, special actions aren't triggered abilities, so Pithing Needle and the like don't affect them.

Paradigm Shift

There are two immediate impacts of this rule change. The first is counterplay. The companion must be tutored into hand as a sorcery, making it vulnerable to discard spells. This is huge, because prior to now, the only way to answer a companion was to counter it or kill it after it had hit board and probably gained value. Killing a companion is often necessary, but feels bad, and not many decks can run counterspells. This change affords counterplay options to more decks and makes companions more like normal cards.

The second is playability. The big draw, and the big problem, to companions was how easily they slotted into decks and provided power boosts: give up a sideboard slot; pay the mana-cost; get a decent creature with a strong ability. Now there's a tempo cost. Three mana is a high price for drawing a single card. Having to do so at sorcery speed means sacrificing the ability to do anything else while telegraphing next turn's play. Doing all of that in one turn requires at least six mana, which few Modern decks want or expect to have.

As a result, the opportunity cost of companions is much higher. However, not so high as to make them all unplayable. Those decks that were maximizing the utility of the companions beforehand will be doing a lot of soul-searching. That tempo hole will be hard to climb out of, and the payoff is definitely lower than that of being a normal, companion-less deck. Whether the guaranteed card is worthwhile remains unclear.

The Outliers

Interestingly, it's the incidental companions that are least affected. Burn ran Lurrus because there was no deckbuilding cost, and there still isn't. It wasn't very good, but it was generally better than not having Lurrus. The only thing Lurrus did for Burn was mitigate flooding out. Lurrus is less efficient now, but it does still give Burn something to do when it has no burn. Therefore, I'd expect Burn decks to soldier on as if nothing's changed. Storm and Humans may keep running Jegantha, the Wellspring for similar reasons. However, Jegantha being so much more expensive than Lurrus makes that outcome less likely, as these decks were casting the 5/5 far less often than Burn would Lurrus.

I can't imagine UW Control sticking with Kaheera, the Orphanguard. Snapcaster Mage has never been that good in straight UW, so it was an easy cut. In exchange, UW got more room for control spells and a three mana win condition that didn't take up deckspace. That basic calculus hasn't changed. However, now they have to pay six mana for a 3/3. For that price they could have Elspeth, Sun's Champion, not to mention the option of 1-2 Snaps in the main and a sideboard slot, something more precious for slow, interactive decks like control. Kaheera is getting harder to justify.

Lurrus is Nerfed

In terms of affected cards, Lurrus is the most affected. It was the most widely played, and arguably the most efficient companion. On turn 3, Lurrus was a 3/2 lifelinker that drew a card with Mishra's Bauble. Assuming Lurrus survived, it started snowballing card advantage every turn thereafter. That cannot happen anymore. For the typical low-land Prowess or GBx decks (Lurrus's primary homes), this means that Lurrus has been delayed by a turn. They can tutor on turn three and play Lurrus on four. Of course, spending turn three tutoring is far from the gameplan of either deck.

Even if they decide to hold off playing Lurrus until later, options are still severely constrained. "Waiting until the time is ripe" is well within GBx's wheelhouse, but still presents a problem. Jund in particular turned Lurrus into a Swiss Army Knife using Seal of Fire and Kroxa, Titan of Death's Hunger in addition to Bauble. That was easy when Lurrus only cost three. Now that it costs six, those late-game option will be more restricted. As a result, the Lurrus-dependent decks have to be reevaluated.

Prowess Must Change

Red-based Prowess was screaming to the top of the metagame on Lurrus' back. Some of that is definitely due to Prowess being a cheap deck online, but in this context, Prowess also got the most out of Lurrus compared to other decks. Recycling Bauble meant not just card advantage but also prowess triggers, which translated into wins. Cracking Bauble, playing Lurrus, then replaying the Bauble was two cards and a threat for most decks. Prowess got all that plus 1-2 extra damage per prowess creature. That's a special level of efficiency, in a deck that's all about maximizing efficiency.

Which is a huge problem now. In order to get any use out of Lurrus, Prowess must spend an entire turn doing nothing but tutoring. Without a single trigger. That's anathema to its gameplan. If Prowess isn't churning through spells and attacking, it's losing, and considering how velocity-dependent it is, that lost turn can be fatal.

So, what are the pros and cons of keeping Lurrus for Prowess?

Pros

• Guaranteed prowess triggers every turn Lurrus is in play with upsides
• Long-game threat recursion
• Additional threat
• Flood insurance
• Requires no mainboard concessions (Bedlam Reveler doesn't fit into Lurrus builds anyway)

Cons

• Requires multiple turns of mana to cast
• Delays lethal attack
• Tutoring generates no prowess trigger
• Reduced damage-per-turn output

What About Jund?

Jund specifically and GBx in general is a much harder call. Before, Lurrus Jund had proven itself to be unequivocally better than normal Jund. Despite having an objectively worse maindeck, Lurrus' power, efficiency, and utility more than made up for dropping Bloodbraid Elf and Liliana of the Veil. Lurrus's power hasn't been diminished, but its efficiency has significantly, and with that its utility also falls. Thus, the upside of giving up Bloodbraid and Liliana is lower. The decision isn't clear to me, but I'm certain that the pros and cons are much closer than before.

Pros

• Guaranteed threat which plays well in attrition strategy
• Additional utility from card types for Tarmogoyf
• Deckbuilding constraint rewards preexisting mana utilization strategy
• Extra information from Bauble can be utilized in midrange decks

Cons

• Individual card power is lower (no Liliana or Bloodbraid)
• More vulnerable to graveyard hate
• Additional mana reduces recursion utility
• Engine card now more vulnerable to disruption

Obosh is Unaffected

On the opposite side, I don't think Ponza will really notice Obosh, the Preypiercer getting worse. Obosh wasn't excatly integral but neither was it an incidental companion. It was more of a combo piece, being cast to ensure lethal and/or awaken Klothys, God of Destiny. This is harder to make happen than before, but the catch is that Ponza can make it work. Between mana dorks, Utopia Sprawl, and Klothys, Ponza often ends up with lots of extra mana and nothing to do. Obosh's additional cost will therefore be less burdensome in context than Lurrus.

Yorion is Ambiguous

The final widely played companion is the most difficult to evaluate. Most commentary says that Yorion will be unaffected by the errata. Yorion is a late-game card already, and being played slightly later won't really affect its utility to control decks. After all, control deck hate win conditions taking up room that could be used for answers, and Yorion is still a win condition that doesn't take up maindeck slots. Therefore, there's little reason for Yorion to fall off.

At least as long as analysis is limited to Standard. in Modern, Yorion was frequently not a late-game win condition, but a mid-game stabilization tool. Yorion decks would spend their first few turns playing cantrip permanents and using leftover mana to interact. On turn 4-5 they'd play Yorion, draw cards and/or gain life and have a big blocker against aggressive decks.

Now, they'll have to take a turn off of buildup/interaction to tutor up the Yorion. This means they're much more vulnerable to being overrun by aggro than before. In turn, the statistical weakness from playing more than 60 cards will become more pronounced. The advantages against non-aggro are still intact, so I think that Yorion's value will be decided by the metagame rather than its own merits.

Pros

• Deck's late-game remains intact

Cons

• Increased vulnerability to aggro
• Mid-game power diminished

Where Does Modern Go?

Companions are still around, much to the chagrin of their most vocal critics. However, now there are actual decisions to be made about their inclusion. How this will shake out will take time to determine. The rule change doesn't go into effect until tomorrow, so I won't have data to work with for several weeks. Then the impact will become clear.