last edited by Guest
Ok, in an effort to get wheels back on the cart here, I have a question for the other members of the peanut gallery who aren't myself. How do you use this data, if at all? And subquestion, to what extent do you consider the above archetypes to be 'decks' and to what extent are they just individual cards?
It seems to me that this new BrianKellyesque hybridization of decks has made the above statistical work (which is in fact awesome) more and more difficult as the venn-diagram of decks gets blurred. I heard Kelly himself say the other day that he considers Shops decks to now be a subset of the Thorn of Amethyst archetype... He was half joking, and the joking half is funny... but it was only half. As primarily an Oath player, this is pretty tough for me, since decks that just contain Oath of Druids are all labeled Oath... yet they vary incredibly within that category, from Control style Oath that I just love... all the way to that Burning/Omni/FaceSmashing Oath that Kai was on during the VSL play-in... Thoughts?
The data is a tool - I use it along with my own personal experiences to shape my opinions on the metagame and influence deck construction. It is not the end-all-be-all and it never was meant to be. The true value is looking at the results over time - for the reasons @vaughnbros pointed out, the data from one event is limited and best viewed as a range rather than a discrete point value (for the same reason that if you flip 4 coins and have 3 heads, you don't assume that the odds of flipping heads is 0.75). More data equals more certainty.
Regarding your subquestion, decks are never just about the individual cards but about the synergies and interactions between them and with the cards in the opponent's deck. When I brew something up, I try to think about every possible interaction, from the obvious - like Gitaxian Probe + Cabal Therapy - to the obscure - like taking advantage of Thing in the Ice flipping and Containment Priest having flash to effectively "counter" a Barbarian Ring trigger from Dredge. This doesn't change on the archetype level - synergies still exist. Just having Gush in your deck enables you to play a control game - I think Steve on the VSL stated he plays Doomsday like a control deck. You have virtual card advantage from a lack of mana sources or you have other cards in your deck like Dack or either Jace to convert the returned lands into raw card advantage or further manipulate your deck. One of the most common mistakes I see players make with Gush decks is that they play too aggressively and lose the advantage their deck has in the late game, like tapping out against Storm to land a Mentor instead of keeping mana up for Flusterstorm. Now, individual cards can change certain dynamics but to a limited degree. I don't want Sudden Shock if my opponent is playing Thing in the Ice instead of Mentor, but Thorn of Amethyst is going to be great against almost any Gush deck. Does that answer your question - it's a complex topic?
Hybridization is another limitation of data-collection. Deck classifications are messy. I'm reminded of the Uncertainty Principle in quantum physics - the more you know about a particle's position, the less you can know about the particle's motion. The more detailed you make your classifications, the less of a sample size you have from which to draw conclusions. So what we have done is establish two levels of classification: a kind of macroscopic view that lumps decks into rather generic archetypes, and a more detailed "subarchetype" view that for instance breaks "decks with Oath of Druids" into "decks with Oath of Druids and Omniscience" or Salvagers, or just Griselbrand, or Tendrils of Agony. These views tell us different things but it is a challenge to balance them appropriately.