The Road to DC, Part II: National Metagame Analysis

Hipsters is proudly sponsored by:

When I did my metagame analysis for the Mid-Atlantic region, two weeks ago, several people had asked for a similar analysis on a national level. Ever the crowd-pleaser, I started dumping data into Excel in hopes of helping my readers identify some meaningful patterns and trends in what the DC (*cough* Chantilly *cough*) meta might look like. For those that are too lazy to click the above link (though I highly recommend it), here is the TL;DR explanation of how these percentages are being calculated:

…we will assign each archetype two points for finishing 9th-16th, three points for 5th-8th, four points for 3rd-4th, five points for 2nd, and six points for 1st, then average out the data from the eight 32 events to arrive at what the expected metagame should look like.

The method described above was developed by Frank Karsten for analyzing the “winner’s circle” metagame. That is to say, we are not looking at the expected meta of the entire event. Many people misunderstood this last time, and were under the impression that these results could help predict what they might see in every round throughout the tournament. I cannot stress enough, these results are meant to show the likeliness of seeing a given archetype at the top tables. The information presented here will be far more useful for someone preparing to win the event, than someone preparing to hope to make day two. If you have two or three byes, though, your early rounds will more closely resemble the winner’s meta than the meta for the entire tournament. I just don’t want to mislead anyone and think that they can safely forget about the plethora of decks that exist in this great format and assume that they will be playing against nothing but Delver and Show and Tell decks.

In addition to Karsten’s method, I added a time-weighted version. From my original article:

…I decided to set up my model to give 100% weight to the most recent event (9/8/2013 10/27/2013) and a 10% weight to the farthest back event (12/2/2012) with a linear rate of decay.

This time around, I created two separate T-Scores. The first, T-Score¹, is calculated for the period of time described above. The exact formula used to determine the percentage weight is as follows:

=100% – c * t

t is the number of days that have elapsed since the date of a given finish (remember, we are using 10/27/2013 as our t-zero)

c is a constant that we determine by setting t to our maximum value (329) and solving for c if the weight is 10%; the result is .2736%

For T-Score², we are simply changing our cutoff date so that we are looking strictly at M14 and beyond (7/28/2013). The relevance of this date is twofold in that Young Pyromancer (and I suppose Chandra, Pyromaster) was added to the format and we got the M14 Legend/Planeswalker rules update. There hasn’t been a whole lot to change the format since then, so it’s a good cutoff point to see the recent metagame trends. Our calculated value for c, here is .9890%

To spice things up a bit, I added in a distance-weighted score, as well, called the D-Score (I know, it’s a super clever name). The D-Score is calculated so that anything within a 100 mile drive of Chantilly, VA gets a weight of 100%, and as you get further away, there is a linear rate of decay until distance is equal to 1000 miles, at which point, everything stays at a flat weight of 10%. The exact formula is as follows:

=100% – c * (d – 100); if d-100 < 0, then the weight is 100%

If we want the weight to be equal to 10% when d equals 1000 miles, then c is .1%

Just as we could use the ratio of the T-Score over the K-Score to determine which decks have increased their relative share of winner’s circle in more recent events, we can use D-Score over K-Score to determine which archetypes are more popular in closer proximity to the event site.

The events we used for our calculations include EVERY Star City Games Open between Indianapolis, last weekend, and Baltimore on 12/2/2012). Without further ado, here’s the first table, broken down by archetype:

Look at that drop off on Esper Stoneblade when we use T-Score²! On the flipside, Reanimator, UWR Delver, Elves, and Shardless BUG have been trending up, lately. Format stalwarts, RUG Delver and Sneak and Show have shown very little deviation between their T-Score² and their K-Score (and even the D-Score; these mumbers all fall in a tight range). Those archetypes have consistently been steady performers, regardless of time or place. Esper Deathblade, The Epic Storm, and Death and Taxes all have notably higher scores when we take distance from the DC Metro area into account. There are a bunch of different ways you can sort this data, but I’m just going to leave it sorted by K-Score, here; you should be able to cleanly copy paste this into a spreadsheet program and sort it any which way your heart desires.

Let’s collapse some of these into Macro-archetypes. You may notice that I grouped some of these a bit differently than last time. One glaring mistake that I’ll own up to was lumping all of the non-Delver BUG decks together. Shardless plays much differently than typical BUG Midrange/Control. Anyways, here’s the chart, sorted by K-Score:

The raw data that I used can be found here. You can see how I grouped all of the Archetypes into their macro forms on that spreadsheet, as well. Again, we see that Graveyard combo decks, like Reanimator and Tin Fins have been putting up results, lately. Storm combo decks have a significantly higher D-Score, indicating regional popularity, but they also have a lower T-Score in the post-M14 period. I’m not particularly sure how to weigh this, whether proximity is more important with regards to distance or time. These results, combined with my regional results, do hammer home the point that if you make it to day two, you’re going to be dealing with an assault on your manabase combined with an assault of giant insect men or giant, game-ending bombs dropping in from out of the blue. Expect much of the same if you plan on battling for the Wasteland painting, this weekend!

“Evil” Tim Akpinar is one of Brooklyn’s finest durdlers. If there’s a top-tier control deck in the meta, you can bet he’s spent a minute taking it apart to see what makes it tick. If it wraths and draws cards, “Evil” Tim Akpinar approves. You can find Tim on Twitter/Twitch @efil4zaknupome or on MODO under the username ziggy_stardust.