Chatter of the Squirrel — Polarity

Read Zac Hill every Wednesday... at StarCityGames.com!
If you can successfully stop the fastest deck in the format from killing you while simultaneously not packing to the optimally-positioned control deck’s late game, chances are you can handle a less streamlined version of either strategy. This doesn’t necessarily mean, either, that every deck that isn’t positioned along the poles is “sub-optimal” to take to a tournament strictly speaking; it just means it’s less unilateral.

On Friday night, usually around six o’clock, I always send a text to Aaron White that contains a variant on the same word. “Draught?” means the Flying Saucer, where we’re working on our 200-beer plate-on-the-wall party-on-the-house, but “Draft” in last week’s case meant Masters/Tenth/Tenth. The less educated might insinuate that given our history we wind up drinking even more beer than we would at the bar, but “Bud Light” is hardly beer and M/T/T is hardly draft. Nice Oubliette, Wizards. While it turns out that Banding is actually awesome, I’m not going to bore y’all with an article on that format.

This time.

No, the important revelation from that night was that the marathon Valencia playtesting session I scheduled on Saturday was going to have to be postponed due to Aaron’s wife’s high school reunion. Insert amused Internet slang here (ROFLOLOLOLOL etc.) I was all frowns, but at long length we eventually did get to confirm what everybody knows anyway:

Dredge is nuts, you can’t just try and beat it with a Mogg Fanatic, you can’t just board in Leylines, and unless you’re going out of your way I promise that you don’t have a “good” matchup.

What was most interesting though was the conversation said session led to a couple of days later with Adrian Sullivan. We were trying to figure out how on earth to narrow down the field and, in passing, I remarked, “Well, the two poles of the format right now are Dredge on one hand and Top-a-tog on the other.”

Loose deck nicknames aside – and believe me, I have several – he asked me what I meant by “poles” and I decided that rather than respond directly I’d just write an article about it. It turns out that there are at least five poles in the format as I understand it to be developing, but the factual accuracy of the statement isn’t what’s important. I know this idea isn’t original, and I think it’s Mike Flores who coined the term, but the concept is invaluable for figuring out a wide-open field and inferring based on a relatively narrow gauntlet what your matchups are going to be like more broadly. Moreover, even if you’re able to playtest each and every matchup to a level that you’re comfortable with, setting up a format’s “poles” can give you an idea of what you’re probably going to face at a given tournament.

First, though: What Are They?

Broadly defined, the poles are the reasonable strategic extremes you can take to a tournament and still expect to win. They are composed of either the fastest, most blisteringly quick decks in the format, or the control decks most frequently able to seize the initiative and convert it into a win. Assuming there aren’t any specific cards – and this is a big assumption – that give your hypothetical deck a problem, you should be able to beat the field if you also have a good matchup against those two poles.

From an intuitive perspective this certainly makes sense. If you can successfully stop the fastest deck in the format from killing you while simultaneously not packing to the optimally-positioned control deck’s late game, chances are you can handle a less streamlined version of either strategy. This doesn’t necessarily mean, either, that every deck that isn’t positioned along the poles is “sub-optimal” to take to a tournament strictly speaking; it just means it’s less unilateral. A hypothetical Zoo deck from two seasons ago may have wanted to run Soltari Priests and/or Silver Knights to gain an advantage in the mirror, and once the field is X percent mirror matches it might be a numerically “better” choice to run. Nevertheless, if I am testing Tron, I’d much rather see Silver Knights and Soltari Priests than Watchwolves and Scab-Clan Maulers, and so if I can figure out a way to beat the Watchwolf deck it’s extremely likely I can beat the Silver Knight one, too.

Similarly, Richard and I employed this principle when we built Hermit.dec for Grand Prix: Columbus. Our poles were Flash and Fish, and we devoted all our time to maximizing our chances in those two matchups. Other decks like Threshold that fell somewhere in the middle (vulnerable to Chalice, Masticore, and Poacher while still being vulnerable to graveyard hate in the form of Leyline) got taken care of inadvertently.

Notice that Goblins and Landstill were not a part of our polar system. This was because of the qualification embedded in the definition earlier: the strategic extremes you can take to a tournament and still expect to win. Landstill is certainly “more controlling” from an objective standpoint than Fish is, but we were of the opinion that taking the entire field into consideration playing Landstill wasn’t a high-EV proposition. Similarly, Goblins (while slower than Flash) was an aggressive deck that required a different angle of attack than either Flash or Fish and thus didn’t fall between the two poles at all, but we assumed the deck would be unplayable. Obviously Owen Turtenwald proved that assumption false, but the fact remains that generally the deck fell outside our accepted range of decks to prepare for and thus could successfully be considered a nonentity.

This provides a nice segue way into an implicit point I made before: a given format can have several (theoretically as high as the number of cards legal in that format) poles, not just two. But how do we go about determining that? Well, as mentioned earlier, I originally assumed Extended to be bipolar, but now I’ve modified that hypothesis. The reasoning goes as follows. I started out with the absolute unarguable hyper-fast combination deck, Dredge. Clearly, as the defining deck of the format, it deserves polar status somewhere. It kills quickly, has several angles of attack, and is surprisingly resilient against all but the most devoted hate-plans.

Once I’d established it as a benchmark, I knew any control deck that could beat it would reside at the opposite pole. My first thought was obviously Tron, but I couldn’t come up with a list that had better than a 30%-ish matchup against Dredge without sacrificing something against literally the rest of the field. Here was my best attempt, if you’re interested:

The four maindeck Chalices were an attempt to mise free wins off a turn 1 Chalice at one, under the assumption that even if you didn’t know you were up against Dredge that’s a pretty high-EV play against the format as a whole. Replacing additional artifact copies with Fabricates was a similar concession, since they allowed me to counterspell a relevant card and tutor up a Tormod’s Crypt on turn 3 (or sit behind my Moment’s Peaces until I drew one, at any rate). The problem was that this was about as far as I was willing to go to improve the Dredge matchup, but it didn’t get my numbers high enough to consider it even. Considering that you don’t just auto-win with this deck against the rest of the format, the sacrifices didn’t seem worth it to make.

With CounterbalanceTog, though, it’s possible to get a good enough matchup against Dredge without sacrificing very much against the rest of the format at large – indeed, it turns out that the technology that we used to up the Dredge matchup actually improved our percentages slightly against two other prominent decks. Thus, given that it was a defensible choice against the field, it earned a polar position as well.

That’s not all, though.

Imagine, if you will, a gradient between these two polar extremes. You’re constructing this model to predict the field, and so for it to be valid every available strategy would have to fall somewhere within that range. If there’s another strategy, though, that has a reasonable chance against your existing poles while requiring attack from a different angle in order to beat, you’ve got to accommodate your model to include it. If not, your “gradient” isn’t representative of the actual field and thus is not going to get you anywhere. What our testing revealed was that aggressive Red decks, while posting positive matchups against neither our Tog list nor our Ichorid lists, was posting high enough numbers that to play it wouldn’t be a slaughter. Furthermore, it was having decent luck against what we considered to be “inferior” lists of those two decks and was winning more games the more the rest of our gauntlet focused on our existing gradient. Thus, Red aggro (but not specifically either Levy or Turtenwald lists) had to be introduced as a pole.

Moreover, once we realized that Affinity could basically maindeck four Tormod’s Crypt no problem – and were hearing similar thoughts from other testing groups – it had to be added to the model. While its matchups against Tog and Ichorid were both very good, its weak matchups against the aggressive Red decks that were now added to the gauntlet didn’t necessitate the removal of either Ichorid or Tog like it would have if we were only measuring along one gradient. More importantly, though, the fact that Affinity requires yet another angle of attack – you can’t nuke the graveyard or out-control the long game or kill a bunch of weenie creatures, but really need to have a way to destroy multiple artifacts generally – meant that it deserved a place as yet another pole. It introduced another variable to consider.

Finally, as the metagame broadened, both Loam and TEPS were re-introduced as possibilities. The aggregate blend of four very different strategies meant that, even though both decks didn’t possess blowout matchups against any of the poles specifically, they could gain a good bit of mileage by attacking formats from different – yes, I’m going to say it – vectors than are currently being exploited. Moreover, these decks contain enough raw power that they could positively maul hybrid or mid-range strategies that attempt to take a less linear approach to the metagame – in other words, decks that exist within our gradient and not on the poles. I don’t know that each one of these decks deserves its own pole just yet, but at least one of them does and at any rate quantifying the difference between five or six poles doesn’t really lead us to a useful conclusion.

Does this model break entire formats open by itself? No. It’s simply a useful testing tool – but it does let me know that unless I’ve missed the memo about some major determinate factor, I’m not going to get totally blindsided by a particular strategy. Furthermore, if I’m approaching the field from outside the angle of the several established decks, it knows what angles of attack I am going to have to contend with – that is, I’m going to have to be able to kill creatures, deal with artifacts, control the graveyard, and avoid falling under Counterbalance lock or dying to a quickly-lethal Tog backed up by countermagic. We’ve taken a wide-open metagame and reduced it to factors that we can work with and easily assess when building decks.