Potential new deck categorization for metagame analyses

After receiving some constructive criticism, Matt and I are considering changing our general approach to categorization.

Decks such as Doomsday, and more recently Brian Kelly's Gush Oath lists, have forced us to shoehorn hybrid decks into a single archetype. Our proposed solution is to label each deck with any number of "tags". Possible tags would include familiar archetypes such as gush, combo, and shops, but also include cross archetype tags such as tokens, storm, thorn, and null rod. We would then publish metagame saturation and winrates for each tag.

If there are any tags you think would be particularly useful, let us know by replying or PM'ing.

Before implementing this, we want to know if this potential change is worth the additional complexity to you .

last edited by diophan

I find this a great idea as I have even named decks by the category I thought they were such as "mentor storm" or "mentor control"

@diophan Id be happy to have whatever you decide to do, but I do like the matchup charts that you make with the archetypes

Pitch counters might be a reasonable thing to tag to refer to non-blue decks playing mental misstep and maybe mind break trap. Hell, if we want to go all in, just tag mental misstep across the board (only half joking here). This isn't limited to dredge as mental missteps being played in shops happen as well so I wouldn't say it's too narrow, and notably this deck feature was brought up with much curiosity in the EW analysis.

With this, dark depths, or no tags as an option, dredge gets intelligently categorized with more ability to have a dialogue about mental misstep.

It also does feel like we're moving toward a card inclusion based win rate analysis if you try this tag idea if tags are 'gush', 'oath', or 'Doomsday'. Not entirely, but food for thought.

I was going to say that once you move toward tags, you might end up devolving into 'card tagging.'

Would it make sense that tags would be categorical? Like "Win Condition: Mentor/Planeswalker" or "Win Condition: Oath into Griselbrand" or "Card Engine: Thoughtcast/Paradoxial Outcome" or "Card Engine: Gush".

totally coincidentally, I've been working toward some features to make topic-tags a little easier to work with on this site ... it could be pretty neat if we came up with a categorization system that worked both for the TMD topics and for your analysis. (but don't let the site's needs limit yours)

Okay here's the first draft using the decklists from NA Champs:

A couple notes on the tags:
Thorn doesn't include null rod only Eldrazi decks ("Jacodrazi")
Blue Control includes the very controlling Nahiri Gush lists
Decks were tagged with Null Rod only when they had a Null Rod or Stony Silence in the maindeck
The dividing line for tagging a big blue style deck as combo was that I labelled Academy Combo and X City Vaults as Combo but not Tezz or Painter.

I tried to be as "hands off" in tagging as possible. For instance there were some off-the-wall Green/White Oath of Druids lists, legacy reanimator, and 2-Card Monte. Even though it didn't quite feel right to label them as Oath, Combo, and Shops respectively, I did in the interest of using the tags without putting my own bias into them.

However, if you want to see these archetypes without these lists, you can use the "Tag Search" feature. To see shops lists that aren't 2-Card Combo you include the Shops tag but exclude the Combo tag. To do the same for Oath or Combo you exclude the Other tag. To play around with these tags save a copy to your google drive so that you have edit privileges.

I thought it was interesting just how many vault/key decks were at Champs, which was obscured by the Archetype/Subarchetype classification scheme. Note that the Thorn winrate being higher than the Shops winrate does not necessarily mean that Thorn without Shops did better. The reason is that there are more Thorn mirror than Shops mirrors, which are excluded from the winrate calculations.

Let me know what you guys think!

last edited by diophan


I'll take a closer look in the near future, but at first pass I think things gives a more nuanced picture..

I like it.

  • 8
  • 4980