Buckle up. The Star City Games Regional event decklists are here, and all of a sudden we have a ton of Modern results to sift through. Today’s goal is to speculate as little as possible—instead, I plan on taking an analytical, systematic look at the data we have available, to provide a framework upon which we can base future speculations. Without further ado, let’s get into it!
Twelve Regionals events, 8 decklists per event. For those counting at home, that’s 96 decklists to sift through. Today I’ll be focusing primarily on metagame breakdown, market share and archetype diversity, and doing my best to leave specific deck composition discussion to future articles. Believe me when I say we can quickly get lost in the details if we dig too deep into specific lists. There will be plenty of time for that later.
Acton, Duluth, Gurnee, Raleigh, Catskill, Baltimore, Minneapolis, Plano, Lenexa, Columbus, Orlando, San Diego. For the purposes of this article, I’m weighting all the Regionals Events equally. I guess you could make a case that the level of competition in San Diego was higher than that in Catskill, but I’m sure the Catskill players would disagree. Also note that suburbs of major metropolitan areas often get billing over the actual city center, even if the same flock of players from the area will be in attendance. So equal they shall be. Archetype representation, aggregated across all events, is as follows:
Top 8 Archetypes
|BR Pack Rat||1||1.04%|
For purposes of metagame representation, most of the Tron variants can be lumped together under the “Tron” umbrella, as they play relatively similar strategies and a lot of the same cards can be used to fight them. Yes, I know GB Tron and GR Tron play differently—thank the lord that Mono-Blue Tron didn’t podium or we’d have that argument as well—but for the most part, they are the same on a macro level in terms of what they are trying to do in the format.
At first glance, the Regionals results appear to show a relatively normal distribution for Modern; 4-6 decks taking a majority of the slots, a clear wide middle ground, and then a distinct group of rogue archetypes at the bottom of the pile. Were I to draw the line around “Tiers” today, I would call Tier 1 the top four (Jund, Affinity, Bant Eldrazi, and Burn) as they hold a full 37.5% of the metagame between them. Tier 2 would be any archetype that placed more than two copies in the collective Top 8: Abzan Company, Grixis Delver, Elves, RG Breach, Ad Nauseam, GB Tron, and Scapeshift. Everything else is Tier 2.5 to Fringe.
The same can be said for the Eldrazi lists, including Eldrazi Tron, which plays a lot of the same cards and just reaches higher to Ulamog and Karn Liberated. Similarly, Sultai Delver is still a Delver deck and can be grouped with the Grixis variant. Abzan Evolution, on the other hand, is built and plays fundamentally different from the Abzan Company deck, so it remains its own entity. No Viscera Seer to be found here.
Grouping variations, and combining all archetypes that only boasted a single top finish into “Fringe,” here is another look at the Regionals results:
Top 8 Archetypes (Consolidated)
Conversion Rates in Top 8
The big takeaways from filtering the data this way is, in my mind, a clearer perception of Eldrazi and Delver’s places in the metagame. Yes, Eldrazi only gained two points, and Delver only gained one, but in doing so they place a noticeable gap between themselves and the rest of the format. Jund, Eldrazi, Affinity and Burn occupy a defined top tier, while Abzan Company and Delver remain slightly out of the spotlight at “Tier 1.5.” Elves and below occupy a definite Tier 2, but all are capable of strong results, as seen by RG Breach’s dual first-place finishes.
Speaking of finishes, what archetypes were able to convert a respectable Top 8 finish into a finals appearance? First, let’s take a look at each archetype’s total finishes distributed by position.
Finish Distribution by Archetype
This table might look weird, but it’s really just a visual representation of a similar statistic we’ve been looking at for a long time in the form of the “Day Two Conversion Percentage” that Wizards’ likes to put out in their GP coverage. Before looking at percentages, I want to discuss a few things we can gain from looking at the information this way.
First, if we order our archetypes by representation (most played at the top, least played at the bottom) it’s relatively easy to determine a pattern, and any possible outliers from that pattern. For example, if Jund is the most played deck, it makes sense to see it take numbers in all categories (1st through 8th place). This type of result fits the narrative of “played by many, won by few”. The fact that the first column is blank should be of interest to you, as Jund was the most played archetype yet failed to place first in any of the events.
As we move down the chart, it makes sense that the results should get more sparse, as there are fewer entries to fill the row with each successive archetype. According to the “fringe decks are clearly worse” narrative, we should expect to see a general curve down and to the right, as the lesser-played decks are “expected” to finish worse.
In some cases this is true; looking at the table its easy to notice Ad Nauseam putting up middle-of-the-pack results, along with Abzan and Jeskai Control. The outliers, of course, are Bant Company, Goryo’s Vengeance, and RW Prison. The first two put up two 2nd-place finishes each, while the latter placed pilots in 1st and 5th place. For “fringe” decks, these results are pretty impressive, especially when paired against Jund and Abzan Company’s more numerous, yet comparatively worse finishes.
A Note on Statistical Significance
Note that the n we’re working with here is super low, which means we have to be very careful not to extrapolate too much. While the decks that made Top 8 presumably came from reasonable numbers in the tournaments at large, once we get into the Top 8 itself most archetypes have single-digit representation.
Take 8Rack as an example. With two pilots in the Top 8, both managed to win their first round, and then lose their semi-finals round. That’s a 1-1 record in Top 8, which really tells us very little. Similarly, both Jeskai Control pilots lost their first Top 8 match. If we assume an exactly even 50% matchup in both cases, the odds of this happening are 25%—far too high to assume that it’s out of the norm.
The higher finishes, however, do tell us more. Looking to Bant Company and Goryo’s Vengeance, all four pilots between these two archetypes made it to the finals. Again, if we were looking at exact 50% matchups, the odds of this happening would be 6.25%. That said, we also know these decks lost in the Finals. So they collectively two-thirds of their matches in Top 8—nothing out of the ordinary, to be sure, but for the purposes of evaluating this tournament, more significant than the 8Rack and Jeskai numbers.
The Finals Table
With those caveats, here are the Finals table conversion rates.
|Archetype||Final Table Conversion Rate|
Rather than sort by conversion rate, in my opinion it’s much more helpful to keep things in order of archetype representation. If our goal here is to challenge our assumptions, and the most popular assumption of all is that popularity denotes strength, then continuity across our analysis can provide clarity towards proving or disproving our hypothesis.
We’ve already discussed the strong performances of Bant Company and Goryo’s Vengeance, so we don’t need to re-tread that area, other than to say that all of the Bant Company and Goryo’s Vengeance decks that made the Top 8 made the final table. Sometimes it sounds better when you say it. Without forgetting the caveats mentioned above, we can see that 100% conversion rate really sticks out. Only two other decks (RG Breach and RW Prison) managed a conversion rate above 50%.
At this point, I’m interested in the top of the chart, and what takeaways we can find there. Yes, RW Prison did well in one event, but it did poorly in 11 others. On the other hand, Affinity was one of the most represented archetypes and converted 44.4% of its Top 8s to Finals appearances. There is another thing to take into account here: when looking at the most represented decks some amount of “infighting” occurs, as some of the Affinity players may knock each other out. We don’t know for sure if that happened here, as we’re looking across 12 events and we don’t have matchup information to go off of. But keep in mind that the rates near the top are likely to be slightly underestimated compared to those at the bottom.
Finally, your champions:
Still listed by overall representation, Burn and RG Breach are the standouts here, not only because they took home more than one victory, but also because they were not the most represented archetypes on the day. Burn was the fourth most represented, true, but RG Breach was only the eighth most represented archetype in the Top 8. Of the five Breach decks to Top 8, seeing two win is impressive, almost as impressive as the 8-to-3 ratio for Burn. Yes, Breach’s Finals conversion was “technically” 40% to Burn’s 37.5%, but with numbers that close I’m more interested in that third win. With numbers these small we can’t draw extreme conclusions about the best deck or anything, but Burn looks to have been one of the big winners on the weekend.
It might not be fair to lump all the fringe decks together—I’m sure the guy that did well with Abzan Evolution doesn’t appreciate being grouped alongside Skred Red guy, or that one dude that played Infect as a statement. The one Abzan Evolution list that top-eighted spiking the win is definitely interesting as well, and the same goes for Living End. Still, more weight would be given to such a result if more players played the deck, but therein lies the issue: if more players played the deck, it’s doubtful they all would have won, so wouldn’t we then be treating this hypothetical result the same way we treated Grixis Delver? That archetype also took home a trophy, and had many more players play the deck. Which is the more impressive performance?
Personally, I place more weight in representation than I do for finish, within certain bounds. If the most played deck doesn’t take home two trophies, I call it a disappointment, as I did with Jund. If a lesser-played deck takes home multiple victories, I sit up and take notice. If you feel the same, I’d be interested to hear why, and obviously let me know if you feel differently. I will say I’m more interested in the Abzan Evolution finish than the Living End result; one is a known entity, while the other can be considered “rogue.” I know what I’m testing this week…
Thanks for reading,