Deriving a metagame

The “metagame,” broadly speaking, is the set of factors resting outside of individual game play that nonetheless have a big impact on your success in any gaming environment. Succinctly, the Magic metagame amounts to “What kind of decks will other people bring to [this PTQ/this local tournament/this Pro Tour/my house tomorrow], and how should I prepare for them?” This is an intelligence exercise — in the sense of information gathering. If you’re going to a Pro Tour, you watch what decks are being played in the Magic Online Premier events, and you even listen to what people say the night before the PT starts (since you don’t have to register your deck until the morning of). With the rain delay of PT Valencia, the participants had a whole extra day to check in with the other players and decided to change their decks, or not. Notably, event winner Remi Fortier was going to run a U/W Tron build, but in the extra day decided to instead run Manuel B’s “Chase Rare Control” build (I do like that name for it).
My current bit of pondering is “Can you, from the card pool, predict the likely metagame?” This is a bit of Bayesian reasoning problem — we have the evidence of successful builds from events (as well as the unsuccessful builds that the winners had to beat) and the card pool…can we integrate all that information and understand why certain card pools lead to certain builds?
More in the extended.


SeedbornMuse.jpg
I’d like to play this card in a deck. “Untap all your permanents” just seems like such a nicely abusable option, it’d be great to put together a deck that works up the Muse synergy. Consider, for example, Muse and Arcanis, and drawing six cards for each of your turns.
However, my recent tournament experience saw me hitting a combination of aggro and counterspell-based control builds. The only post-Lorwyn Standard major event, GP Krakow, also saw two of its eight top eight slots go to very similar “wall of counterspells” control builds (you can see the Krakow top eight decks by clicking here). My preference for “tapout control” approaches means that I need to counter the wall with my own wall of disruption, which pretty much rules out the Muse deck…assuming that I think I’m going to hit faeries and counterspells next time I play, too.
This brings us back to the opening question — how do you predict a metagame? How indicative are the top eights at major events? In the coverage for GP Krakow, Rich Hagon said something I’ve heard elsewhere, that in events soon after a set rollover in Standard, people tend to play slightly updated versions of known builds. Certainly, the Krakow top eight saw a Pickles deck and a mana-ramp build, but it also had Faeries and Mannequin builds. The powerful mono-blue builds also feel natural given the new Cryptic Command, even though people were vigorously declaring the death of control with the rollout of Rewind and Remand from Standard.
Now, I think these people just weren’t paying attention, what with Cancel and Rune Snag staying in, and with such an array of powerful card advantage tools available in Standard.
However…if I were to look back at the Standard environment going into Worlds 2005, I think I would have expected more counterspell-based control there, as well. Consider this subset of the available tools…
Decent counterspells
Remand
Hinder
Disrupting Shoal
Mana Leak
Rewind
Decent CA spells
Greater Good
Phyrexian Arena
Tidings
Jushi Apprentice
Compulsive Research
Even with all these tools, the standout deck of that event was Ghazi-Glare, a build that functions by kicking out tons of tokens and then tapping down opposing creatures and nibbling for the win. Ghazi-Glare builds have basically no counterspell cover. Similarly, Frank Karsten’s Gifts-Greater Good deck (note that use of two superlative CA tools here) has nearly no direct opposition to counterspells.
So what gives? Why did the early Kam/Rav/Ninth environment lead to the types of decks it did, while Krakow and my local scene suggests that early CSP/TSP/LOR/Tenth is all about the counterspells?
Or is it? Certainly, the tournament scene at Krakow saw a major presence of Red Deck Wins, Tarmorack, Rock, and Elf builds…and different States and Champs events saw quite varied mixes in their top eights. Consider some of the biggest environments from the US and elsewhere…
California
1. Mana-ramp into Overrun, sort of (Aggro)
2. Teachings with Haakon
3. Skred Red
4. U/W Blink
5. G/B Rock
6. Teachings
7. G/B Rock
8. U/B Faeries
New York
1. Teachings
2. Mannequin
3. RDW with Tarmogoyf
4. G/B Elves with Garruk
5. U/B Faeries
6. W/U/G Blink
7. G/B Rock
8. W/U/G Blink
Hokkaido
1. R/G Mana-ramp with Garruk
2. R/G Mana-ramp with Garruk
3. B/R aggro
4. U/G Shifter with Garruk
5. R/G aggro with Garruk
6. W Kithkin
7. W/U/R Blink
8. W/g Martyr
London
1. G/B Elves-Rock
2. U/G Tempo-Elves
3. Pickles
4. B/R aggro (but quite different from the Hokkaido take on B/R aggro)
5. G/B Elves-Rock
6. Pickles
7. W/u/R Control
8. B/R aggro (very much like the number four slot)
I have no clear take-home message from the available information. The conventional wisdom is that early in a given environment, people roll out their aggro builds because that’s easier to figure out. The corollary conventional wisdom is that if you just up and bring your control, you’re to some degree more likely to make it higher in an event — be it placing well at your local tournament or making it to day two of a GP. California, New York, and London saw a majority of control and tempo-control builds, but Hokkaido had a significantly stronger aggro presence, just at a glance.
If anyone really wants data for their own personal Bayesian reasoner (either a program or the one in your head), you can get all the top eights from 2007 States in the US and 2007 Champs in Canada, the UK, and Japan, by clicking here.
In the absence of any clear conclusion for myself, I’m left back in the position of just trying to pick a plan that is either stellar against aggro or control, or reasonably robust against both, with a solid plan for the sideboard.
This puts me onto a completely different set of questions, like, “Why was Wrath of God good as a one-of in Mori’s Worlds 2005 deck?” Genius, or quirkiness?
More on that later.