The Truth About Mythic Championship Cleveland

The GAM Podcast sent Kat Light to Mythic Championship Cleveland on a data collection effort. She managed to get the vast majority of data, which helps paint a clearer picture of what was successful during the event and why.

The Numbers, at a Glance

Here’s a quick glance at the biggest archetypes, their relative metagame shares, and their win percentages. There’s nothing too crazy, except that Nexus decks were a larger metagame share that I anticipated.

This next chart is probably what you’re here for though. It shows each archetype and how they performed during the entirety of MC Cleveland.

The first thing to note is that Sultai felt like a safe choice for many in Cleveland, myself included. As it turns out, the players at the MCs are a strong, clever group, and they were well-prepared for Sultai, both in deck selection, specific cards, and the ways they approached the matchup.

Even though Sultai seemed safe, this was one of those instances where the massive target on its head was too much. Of decks with a large metagame share, only Izzet Drakes, Rakdos Midrange, and Selesnya Tokens did worse than Sultai. Matchups that were supposed to be close for Sultai, like Simic Nexus, Mono-Blue Aggro, and Esper Control proved to be anything but. Team ChannelFireball’s take on Izzet Phoenix was particularly brutal.

Going forward, Sultai needs to massively adapt or fade out of existence entirely. It might beat up on fringe archetypes, but that’s about all it has going for it at the moment. One of Sultai’s strengths is how adaptable it is, but nearly everyone was a week behind where they needed to be.

The biggest winners on the weekend were Mono-Blue Aggro, Esper Control, and Azorius Aggro. Aside from Esper, those deck propped themselves up by preying on the Nexus of Fate decks. In fact, it was surprising to me how many people chose to play a Nexus of Fate deck given how volatile the format seemed toward it going into the MC. Kenta Harane’s success at GP Memphis, thanks to his transformational sideboard, was likely the catalyst that had many people switching to the deck at the last minute, including Top 8er Michael Bonde.

By the numbers, Azorius Aggro outperformed Mono-White Aggro by a fair margin. Oddly enough, Azorius outperformed Mono-White in matchups where you wouldn’t expect the blue splash to help in the matchups, such as the mirror, Izzet Drakes, Mono-Red, and Gruul. That can likely be chalked up to a small sample size for white aggro decks.

Obviously the blue splash came in handy against both versions of Wilderness Reclamation decks. Given that Mono-White had a much stronger win percentage against Esper Control, it leads me to believe the blue isn’t worth it unless you expect a strong presence from Wilderness Reclamation in general.

Stop playing Izzet Drakes. It doesn’t beat anyone. CFB’s Izzet Phoenix deck had a polarizing matchup spread, but at least there are matchups where it’s good.

As expected, the “other” category didn’t fare particularly well. There’s a reason the most popular decks are the most popular. Props for trying to break the mold, but this wasn’t the MC to do it.

The Numbers, by ELO

One of the things I was most curious about was whether or not the elite players qualified for the Mythic Championship would be able dramatically shift how some of the matchups typically play out thanks to their playskill advantage. Next, we looked at players with a 1700 or higher ELO rating (courtesy of against the field as a whole.

Oddly enough, the numbers don’t seem to change much. The biggest difference was Izzet Drakes and Team FacetoFace’s take on Temur Reclamation getting a boost in win rate, which likely indicates how much the pilot matters with those particular decks.

I’m always searching for the truth. Data itself doesn’t lie, but how you acquire that data can. If two inexperienced players play a matchup 100 times, the data from that set can’t tell you much, if at all. At the MC level, this typically isn’t the case, since each player is playing what deck and version they expect gives them the best chance to win the tournament. They have been testing and perfecting their decks for weeks, so this seemed like a good chance to filter by ELO rating in order to find the “true” matchup data.

The win percentages for Mono-Blue, Rakdos Midrange, and Esper Control absolutely tanked when put up against stronger competition. On the flip side, Selesnya Tokens, Izzet Drakes, Gruul, and Temur Reclamation ballooned.

You could argue that the 1700+ ELO players will have better decks on average, so their takes on these fringe archetypes was much stronger than the rest of the fields. You could also argue that when two excellent players are facing each other, the bonus of playing a fringe deck that your opponent may not have tested against, nor build their deck to beat is how you get an edge.

The Future

Based on this data, I would expect Sultai to be a smaller percentage of the field. Mono-Blue Aggro, Esper Control, and Azorius Aggro appear to be the best performing decks, but for Mono-Blue and Esper, that was in an MC world where they were able to prey on Sultai Midrange.

Esper Midrange, although a small sample size, was the deck that put up the strongest results against those three. It even had fine matchups against Simic Nexus and Sultai! It’s probably a deck to revisit. Both Gruul and Mono-Red have the tools to fight the best performing decks from MC Cleveland, but it’s a matter of finding the right configuration.

This Standard format is far from solved.

3 Replies to “The Truth About Mythic Championship Cleveland”

  1. This is like the holy-grail of GP data. How did data collection work on the ground to actually get this? Huge props to Kat Light for actually doing the leg work though!

  2. This is an awesome article. The work being done in these data gathering efforts has been fantastic the last year. One thing i kept assuming was most of the wins for Izzet Phoenix were due to some top tier players piloting the deck. Now i can see that overall the deck went 24-18 among 1700+ players and only 9-11 among the rest of the field. Small sample size and 9-11 is probably an average record for sub 1700 ELO players, so maybe the deck is a decent choice either way and wasn’t “just” Luis and company, possibly.

Leave a Reply