Rent This Deck

tl;dr - Affinity was the clear winner of the Modern MTGO scene last weekend, as shown by analyzing win-loss record, instead of merely popularity. Jund and Melira-Pod were played far more often... but Affinity put up unreal numbers head-to-head against all other 3-1 or better decks, and also had MORE 4-0 decks than 3-1 decks (6 vs 4), which also implies metagame excellence.

-----

I'm hard at work creating what I expect will be the best, most valuable metagame breakdowns of Modern and Standard, through the lens of Magic Online tournament results. Obviously, compiling data and/or analyzing MTGO tournament results is old hat by now, so what makes this different?

In short, win-loss record. I look not only at the metagame %'s of 4-0 decks, but also compare with 3-1 decks, to estimate which decks are winning. Plus, I am able to compile actual head-to-head wins and losses, to get an even clearer picture!

Coming soon, I will be able to actually 'solve'* a metagame as though it were a Rock-Paper-Scissors game theory puzzle (by looking at head-to-head matchups for all the major archetypes) to give which decks are being overplayed, underplayed, predictions of metagame shifts, and "what to play now?" recommendations.

Note that the accuracy of the predictions will heavily depend on not just the amount of data available, but also the number of viable archetypes. The more splintered the metagame, the less accurate the observed winning percentages between any given two decks. For that reason, the Standard metagame analysis is likely to be a lot better than the Modern one, as there's a lot more people playing Standard, and fewer archetypes being played.

Here's just a small taste of what's in store! I got data from 15 of the last 16 Modern MTGO events, including two Premier events, from Fri-Mon. With only 3 days, it's only a small sample. Also, I haven't finished refining my list of archetypes, so there's some "Unknowns" in there.

The format of these numbers is:
1. Name of archetype
2. The plain number = number of decks (that got 4-0, or 3-1, or top8 in a Premier event)
3. In parentheses: the head-to-head record against OTHER archetypes. (E.g. Melira-Pod went 6-6 in 6 games against itself, so it's 'real' record was 20-21, but the more relevant record was 14-15 against other archetypes.
4. In brackets, # of 4-0 decks vs one fourth of the # of 3-1 decks. The divide by 4 is because there are about 4 times as many 3-1s as 4-0s. So this normalizes it, so it's obvious whether the archetype is doing well or poorly. Basically, it gives a rough estimate of likelihood to win a 3-0 vs 3-0 game. (In reality, the proper number to divide by is ~3.9, as with Byes and pair-ups / pair-downs, there tends to be just slightly more than one 4-0 for every four 3-1s.) #3 is probably a better indicator than #4, but both are included.

All Decks: 171
Melira-Pod 32 (14-15) [5, 6.75]
Jund 22 (19-13) [4, 4.0]
WU Tron 17 (6-12) [1, 3.5]
Unknown 16 (19-11) [6, 2.0]
UR Storm 12 (5-7) [3, 2.25]
Affinity 12 (17-5) [6, 1.0]
Goblin Guide Decks** 15 (6-13) [0, 3.5]
Splinter Twin 11 (7-9) [1, 2.0]
Caw-Blade 10 (6-6) [2, 2.0]
Smaller Archetypes (Everything below combined) 24 (15-23) [3,4]
***
White-Martyr 5 (1-5) [1, 0.75]
Scepter Aggro 5 (7-7) [0, 0.75]
Bant Hexproof 4 (4-2) [1, 0.5]
RUG Aggro 4 (1-2) [0, 1.0]
Teachings Control 1 (1-0) [1, 0.0]
Melira-Pod 1 (0-1) [0, 0.25]
Caw-Blade 1 (0-1) [0, 0.25]
UB Aggro 1 (0-1) [0, 0.25]
Merfolk 1 (1-1) [0, 0.25]
Fairies 1 (0-3) [0, 0.0]

Analysis:
Affinity is the HUGE winner! If you were just looking at 4-0s, you'd see Affinity, Melira-Pod, Jund, and others all pretty splintered and even. But the win-loss indicators show that Affinity has been dominating the top tables, whereas Melira-Pod and Jund have fewer 4-0s in spite of likely having FAR more players running those decks! Melira-Pod had 32 players manage 3-1 or better... only 5 of the 32 managed 4-0. Affinity actually managed MORE 4-0s than they had 3-1s! (6 vs 4).

My 'Unknown' grouping also did surprisingly well, so I'll have to take a closer look to see if there's one or two new or new-ish breakout decks hiding in that grouping. And Jund was really the only other winner, at least in the W-L record, as Jund was still only average in 4-0s vs 3-1s.

Here's more info on how Affinity did against the field:

Affinity W-L:
3-0 vs Jund
3-0 vs Unknown
2-0 vs UR Storm
1-0 vs Scepter Aggro
1-0 vs White-Martyr
4-1 vs Melira-Pod
2-1 vs Goblin Guide Deck
1-1 vs Splinter Twin
0-1 vs Bant Hexproof
0-1 vs RUG Aggro

Remember that all of this data is from a pretty small sample size, so take it all with a grain of salt. Still, it's pretty likely that until and unless the metagame shifts again, Affinity is one of if not THE best-positioned deck right now.

- Robert Hatch

Footnote:
*[As far as 'solving' a metagame is concerned, it's obviously a lot more complex than can be captured statistically. For instance, rather than switching decks to have a better win % against deck X... you could just switch 4 cards in the sideboard, or switch 4 cards between main and side. However, simply by looking at what the statistical solution IS, you get a step ahead of people without that data.]

**[Goblin Guide decks includes mostly 'Bump in the Night' decks right now, but also Boros and straight RDW.]

In the post just above this one, I use the phrase "very soon". The phrase "coming soon" also shows up on this website. But what does that REALLY mean?

Well, here at Rent This Deck, we believe in transparency and erring on the side of OVER-explaining, rather than complete silence. That's why I want to let you know as much about the amorphous time range "soon" as I can. Now, of course, if I actually KNEW the day we'll be ready, I'd tell you! I don't. Here's what I do know:

- We might be ready as quickly as around Feb 7th.
- We hope and intend to be ready by the end of Feb, if not much sooner.
- The more people that join the mailing list, leave comments, and/or PM RentThisDeck on MTGO, the more incentive we will have to put in more hours on this, and get it done faster. TRANSLATION: Join the mailing list, and invite your friends to do the same!
- If almost no one expresses interest, and/or we run into other major obstacles, we may not make the Feb deadline. If that looks likely to happen, we will give plenty of forewarning to anyone on the mailing list.

Hope that helps clear that up!

- Robert Hatch
Founder, RentThisDeck
MTGO username: RentThisDeck

Rent This Deck

Pages

Feb 13, 2012

A Small Taste of the Modern Metagame

Jan 31, 2012

Rent This Deck website in progress

What Does 'Very Soon' REALLY Mean?