Modern Allegiances: Fire and Phoenix

Last weekend, Grand Prix Toronto gave us a first glimpse of Ravnica Allegiance in action in Modern. I’ll look into a couple of the biggest additions. First, though, let’s turn our attention to the metagame makeup in this new world without Krark-Clan Ironworks.

Toronto’s Day 1 Metagame

I know what decks 786 players chose to run at Grand Prix Toronto, which equals about 60% of the total attendance. 60% isn’t 100%, but it’s a large enough sample for us to assume that the full field largely followed the percentages listed below.

For example, it is very likely that Burn and Izzet Phoenix were indeed the number one and number two most popular archetypes, respectively. This already is major news. For several months, Spirits was locked into first place, with no less than a 2% lead. The departure of KCI and the rise of Arclight Phoenix reshuffled everything, it appears. Now Spirits barely broke into the Top 8. Instead, Dredge, Tron, control, and Death’s Shadow all rivaled the former supreme ruler Supreme Phantom in numbers. Taken together, White-Blue and Jeskai Control even made for one macro archetype with more followers than everything but Burn.

Why Burn took the pole position is unclear. I’ll investigate the impact of the new toys Burn got from Ravnica Allegiance further down. Why Izzet Phoenix didn’t make it past 8% is just as unclear. The deck’s metagame share increased by 3.6% between the two previous Modern GPs, but inexplicably the development didn’t continue. Izzet Phoenix even received a new tool too.

In other news, Amulet Titan almost doubled its metagame share since the GPs in Port- and Oakland. Arclight Phoenix decks without blue also made a bit more of an impression than they did at earlier events.

Day 2 Conversion Rates

193 pilots of the known 786 decks qualified for the second day, which equals 24.55%. You’ll find this base line highlighted below. Some archetypes put a higher percentage of their players into Day 2, some fared worse, and the table is sorted accordingly.

Several of the best conversion rates are clearly the result of a low sample size. A larger population couldn’t possibly replicate such success. Whir of Invention in particular had a great weekend, and its previous performance at multiple smaller events suggested that a big break was about due. But 50% of eight isn’t as impressive as 45% of 22 or, come to that, 41% of 27.

If we’re looking for an even more reliable sample, Izzet Phoenix delivers—and one of the best credible performances to boot. For additional reliability we can further check its Day 2 rate against the same at GPs Port- and Oakland. First 39.6%, then 35.6%, and now, squarely in the middle, 37.5%. So, yes, Izzet Phoenix continues to rake in some of the very best results in all of Modern, and it does so with remarkable consistency.

Among the archetypes most popular in Toronto, only Dredge’s performance even came close. Tron, Jeskai Control, White-Blue Control, Spirits, and Death’s Shadow all did below average. Admittedly, most of them didn’t fall short of 25% by too much, and their base numbers were on the smaller side as well. The same wasn’t true for Burn, though. Toronto’s 90 Burn players made for a huge data set. That only 18 of them advanced to Day 2 turned a huge showcase into a huge disappointment in turn.

A Skewed Stage

Ravnica Allegiance brought Burn two potential improvements: Skewer the Critics and Light Up the Stage. Let’s try to figure out whether these cards actually constitute an improvement or not, and if so, how big of an improvement.

Skewer the CriticsLight Up the Stage

The problem is that we don’t have a proper control group for one of the two cards. Virtually none of Toronto’s Burn players failed to update their decks with Skewer the Critics. A grand total of five players ran none, and all of them failed to get to Sunday. That may make Burn decks with Skewer better, but it could just as well mean that players who don’t keep up with trends are generally more prone to losing.

Across the GPs in Port- and Oakland, 17.9% of the then still Skewerless Burn players qualified for the second day. 20% is an improvement but falls within the range of natural fluctuation. I’m inclined to take the almost universal adoption of Skewer as evidence of the card’s merits. Though that is all we have to go by, really. The data tells us no more than that it wasn’t any kind of game changer.

Light Up the Stage is a more interesting case, if only because we find the perfect control group within our sample. Among the Burn players whose lists are available to me, 45 ran Light Up the Stage and 45 didn’t. This success story is one the card itself couldn’t live up to, unfortunately. Only six players who lit up the stage made it to the second stage of the tournament, whereas 12 of those without Light did.

Once there, light Burn players again fared way worse than heavy Burn players. None of the versions with Light Up the Stage placed anyone within the Top 100, whereas five of the versions without it did. For reference, here are the lists of the highest finisher with a full playset and of the highest finisher with zero copies:

Light Up the Stage Burn

José Luis Echeverría, 134th place (9-6)

Unlit Stage Burn

Richard Yam, 12th Place (12-3)

In the end, neither group’s size is large enough for any of this to count as conclusive proof. Yet it is rare for tournament results to show a trend this clearly.


Speaking of trends: Izzet Phoenix’s climb in popularity may have petered out at 8%, but the more exciting development is just how many of Toronto’s top finishers ptered. Pteramander wasn’t on my radar last week, but over the course of the weekend, Simic’s Salamander made a strong case to become a regular cast member of the Izzet Phoenix show. It even made a prime time appearance in the hands of semifinalist Kale Thompson.

Once again for reference, here are the decks each with and without Pteramander that reached the highest position in the final standings after Top 8:

Izzet Phoenix Salamander Drake

Kale Thompson, 3rd Place (14-3)

Izzet Phoenix

Peter Rawlings, 5th Place (13-2-1)

The functional differences between the two main decks—that is, excluding the Flooded Strand/Misty Rainforest split—amounts to: one Steam Vents and two Pteramander for Thompson in the place where Rawlings ran one Island and two Pyromancer Ascension. That is very little difference, and allows us to concentrate fully on the Salamander.

  • 39 Izzet Phoenix players registered a list without Pteramander. Nine of them, about 23%, advanced to Day 2.
  • 25 Izzet Phoenix players registered a list with Pteramander, overwhelmingly two copies. 15 of these players, also known as 60%, advanced to Day 2.
  • Ten of the Pteramander players finished the tournament with a record better than 10-5, whereas three of the players without Pteramander did. This equals two thirds and one third of each subset who reached the second day.

I’m not usually in the business of calling the results of my data gathering into question myself—that’s what the comment section is for. At this point, I’m tempted just to repeat slowly, “Sixty. Per. Cent,” and to call this case closed. But of course it isn’t as easy as that. We can’t attribute this outrageous difference to the presence of Pteramander alone. It’s unrealistic that a 2-of influenced the deck’s performance to such a degree.

While people who earned better results were more willing to adapt in literal rules terminology, they also exhibited a higher tendency to adapt figuratively. I believe both must be a factor in this success story: Pteramander is great and players who pick up on a new trend early are great too.


Izzet Phoenix is still one of the strongest decks in Modern—possibly the best. Burn still isn’t. Skewer the Critics didn’t improve the archetype noticeably, and Light Up the Stage actually lowered people’s win percentage. Pteramander, on the other hand, turned out to be the real deal in Izzet.

Scroll to Top