This was written prior to the Top 8 taking place. Normally when I do these articles I have full information. This time, however, I do not. I do not know who won. In a way that's beneficial, because I'm not clouded in my perception of that deck. In reality, seven more matches aren't going to make a big difference in an archetype's overall performance. Of course, those seven matches carry a little more weight than any of the other ones and will be the matches remembered by history. And they probably mean something to the competitors.

In 1994, Zak Dolan took home the first World Championship trophy for Magic. It's far to say that things have changed since then, wouldn't you say? Back then there were 512 people. In 2011 there were 365. And there were three more Hall of Famers inducted this year than there were in 2004. This year the three formats a Planeswalker must conquer in order to battle for the title are Standard, Innistrad Booster Draft, and Modern. We're going to investigate the constructed portions, digging into things like how many people played each deck, how each archetype did and how they did against each other.

This is going to be kept light. I'm not going to go crazy with inferential statistics; this is going to be almost entirely descriptive. And, of course, there are plenty of variables. Obviously not every deck is built with the same 75, nor do they all sideboard the same way. I'm also quite confident that the play skill of the 365 people slinging these spells varies. (If it didn't, then I'm as good as LSV and I'm just unlucky.) Long story short, this is just showing what happened. So if I tell you that Tempered Steel crushes Green-White Tokens and your testing says otherwise, that's cool. There are plenty of variables involved and reasons why both statements are true.

Let's start with the easy stuff: What did people play? I say easy, because I didn't have to do anything. Rashad Miller did all of the legwork on this one, and in fact already wrote up a nice little summary here. For ease of reference I'm including his chart here.

What you see here is a diverse metagame. There are no boogeymen in this format (yet, at least) and no dominant you-must-be-able-to-beat-this-deck-or-you-will-never-win-a-tournament decks taking up 25-30% of the metagame. Just good, wholesome, solid decks. For our purposes we're going to focus on the five decks that made up 10% or more of the field. And Tempered Steel, for obvious reasons. OK, we'll do Blue-Black Control too—I'm feeling generous. But first, of course, the summary. Remember, all mirror-matches have been excluded (math fact: all mirror matches are 50%, barring a double DQ).

At first glance, Tempered Steel crushed it. Of the top seven most popular decks, Illusions is closest in performance, and it was 8.5% worse. You could call the deck dominant in that respect. I'd caution, though, since Tempered Steel did have something going for it: about two-thirds of those playing it are on the dominant ChannelFireball.com team. After Tempered Steel there is a grouping of decks in the 52-53% range with Illusions, Mono-Red and Wolf Run all having similar results. Solar Flare and Green-White Tokens ended just under 50%, and Blue-Black Control tanked entirely. Let's look at how all of this happened.

Green-White Tokens

The tokens proved problematic for Blue-Black Control and difficult for Mono-Red to power through. Otherwise, though, "Token Town" gave up an unspectacular performance, including three sub-40% matchups against the top seven most popular decks. Given that, getting within 1.4% of breaking even is actually kind of impressive.

Illusions

It's never bad to have multiple pairings where you can get above 60%, and that's what Illusions did. Their aggression was too fast for the control decks to handle. Green decks, too, had trouble (though to a lesser extent). If it weren't for really rough pairings with Mono-Red and Tempered Steel things would have gone quite well for those illusions.

Mono-Red

You should be used to seeing Blue-Black Control as a good matchup by now, so that's not news. But 81% is pretty crazy. With only 16 matches of that, I have to think the number would regress some over a larger amount of matches. But maybe not; it could just be that good. What do I know? Small creatures and a lot of burn easily handled Illusions. Solar Flare was not a great result, but still above average. Tempered Steel and Green-White Tokens weren't disasters but were still very dangerous. Steel, of course, could have gone either way with more matches; 17 is not a lot to go on. Wolf Run, however, was a disaster.

Solar Flare

Blue-Black Control at the top again. Weird. Very strong showing against Green-White Tokens too, at least in part due to Day of Judgment. Wolf Run was also very positive—though, again, with only 19 matches there's room for some change there. Mono-Red was near the break-even point but still a negative, and Tempered Steel was 8-2 against Solar Flare. Of course, 10 matches is hardly enough to say anything, but it's a pretty bad start. Illusions were a very negative pairing, and 29 matches is getting pretty close to "statistically relevant."

Wolf Run

Two of the three most popular decks were above 63%. Now that's a good sign for a deck. The bad news is that after that there are two coin flips and two bad pairings, resulting in only being slightly above 50%.

Tempered Steel

Boom. Dominance. I'm not sure if that's a commentary on Tempered Steel or on the names of the pilots. Paulo Vitor Damo da Rosa, Luis Scott-Vargas, Conley Woods, Brian Kibler, Martin Juza, David Ochoa, Ben Stark, Owen Turtenwald, Josh Utter-Leyton, Matt Nass, Shuhei Nakamura, Eric Froehlich... You've probably heard of some of those names. So, yeah, it is fair to say that the deck had above average play skill. You could give those guys 56 Forests, 3 Mountains, and 1 Devil's Play and they'd probably break even.

Blue-Black Control

Boom. Anti-dominance. Poor performance all over, nothing really to discuss. Just bad.

Here's a quick summary of the top seven decks and how they did against each other. Opponent deck is across the top:

That's quite a lot to take in. And we're only halfway through! Can you imagine how the competitors felt?! Let's talk about Modern. First, as usual, what as in the field? And, as usual, Rashad Miller has done the hard work on that and even explained himself here. Recent bannings had many people wondering what would happen to the format. Now we know.

2 themes are apparent when looking at this. Zoo. Cheap creatures that hit hard backed up with powerful spells across 3, 4 or 5 colors is as tempting as it gets. Zoo is a classic deck, a default fallback for people who either haven't tested or just don't know what to play. And, of course, for those who have tested a lot and found nothing better. The 2nd theme I see is that after Zoo and Twin, the field is about experimentation, pet decks and diversity. There were 36 different archetypes, compared to only 23 in Standard. That's even more impressive when you consider that 89% of the field had dropped out by day 3. So a 57% increase in archetypes when there was an 11% decrease in players. With so many available cards in the Modern format, people want to play with all of them. We'll focus on the top 4, or those that were 5% or higher (or very close to 5%) of the field.

The problem with such a diverse field is that it means fewer matches for each archetype. We'll work with what we have though. Melira Combo came out as the leader for Modern, something I expect is aided by its inability to be played online which makes it harder for people to test against it. Death Cloud and Jund both put up good numbers. The top performer from our four most popular decks was the least popular one. This isn't surprising, since fewer people playing means fewer matches which means higher variance. Zoo came out almost 2% better than average, which is pretty impressive when taking in account that just under three out of every ten players ran with it. Affinity was just shy of 50% and Splinter Twin came in at about 40%.

Zoo

Zoo's performance was actually a letdown from Philly, where it won 56% of its matches. A lot of that looks like it is because of Snapcaster Mage. Here's how the variants for Zoo did:

There's a clear underperformer. Removing Snapcaster variants from the mix brings Zoo back up to 56%. At any rate, Zoo beat up on Splinter Twin and Affinity pretty well. It looks like people's miscellaneous concoctions were the downfall of Zoo. Strangely, the Snapcaster variants had very real issues with Affinity, winning only 39% of those matches. Non-Snapcasters won 71% of their Affinity matches.

Splinter Twin

Zoo is just hell for Splinter Twin players, and those matches were over a third of their non-mirror matches. Removing Zoo would bring Splinter Twin up to 43%, which is still pretty bad. There are just too many decks with either speed or disruption for the Twin decks, it would seem. Splinter Twin, albeit in a very different form, was significantly better in Philly. It dropped almost 12% from event to event, though you can't really say that was because of how the metagame shifted but rather how the deck had to adapt to the loss of key cards.

Affinity

Affinity had a lot of pairings in the positive, but they constituted a smaller portion of the metagame. Those positive matches didn't help Affinity enough since it was in the red against the two most popular decks. Affinity did better at Philly, winning 52% of the time. Some of that delta is a result of its 56% against Twelve-Post in Philly, an archetype that's no longer viable after the bannings.

Mystical Teachings

A control deck! With a winning record! In Modern! Progress, indeed. Mystical Teachings decks made a comeback in large part due to a very strong showing against Splinter Twin (in only 14 matches) and Zoo. Their strength, it seems, is in handling many of the various other decks that comprise the metagame. Classic control deck strategy.

I love that Modern is as diverse as it is from a player and fan perspective. From an analysis perspective, however, it is a downer. There simply aren't enough matches for each archetype to do any meaningful analysis on. Really, it's annoying.

See you after the next Pro Tour!