How to Metagame - Redefinition of "Best deck"
Pt 1 - "Play the Best deck" - Redefined
If you have much experience in competitive magic, you've probably stumbled upon the saying "Just play the best deck" many times.
This saying is one of my biggest pet peeves in all of magic. I believe it is over used, misleading, lazy, and often.. not true.
The intention behind this article is to explore some insights of successful meta-gaming and hopefully provide a more useful definition of what a "best deck" is.
Part of the issue with using the term "best deck" to refer to the "best recently performing deck" is the language of "best" is connected to a deep intuition of what it means to be the best at something. When we think of the "best player" in hockey, we think of the best performing, highest skill , scored the most points etc. Attributes that are fairly static. If you were the best player in the world last week, the chances you will be this week is extremely high.
Most people throw the term "best deck" around when a deck puts up the most impressive performance at one or two events, early in the format. Generally these decks are what the best players at a tournament are playing.
The issue with applying that to a magic deck, is that the best performing deck at a recent tournament, is only a mild indicator that it will be the next week.. Many factors can lead to a deck performing great over a couple tournaments, such as ; pilot skill, an inefficient meta-game and variance.
Magic has a very complex rock/paper/scissors type dynamic to it, where one strategy nearly always has an effective counter strategy to some degree. The performance of a deck in one tournament is highly influenced by the composition of that particular tournament.
I see the term "best deck" thrown around in contexts, where it would be equivalent of saying "Rock" is the best in rock/paper/scissors, because it won the last round.
While I'm simplifying the narrative to get my point across, the point is, a decks recent performance is often over-valued and it's conflated with its inherent power level. This is especially true, early in a format, as information is still being uncovered.
I blame part of this confusion on the ambiguous definition of what it means to be "the best deck".
I would want a definition of a best deck to not only be a useful indicator of what to play, but have it be something independent of time, closer to the intuition of what it means to be the best athlete in the world.
My definition, "The Best deck" in a format is a deck that when perfectly piloted would have the highest win-rate against an exhaustive sample of random decks ranging in quality from failed FNM brews, to the tier 1 tournament decks. Weighted by deck quality.
theory aside:
It is essential that the frequency of the comparative decks be weighted according to their power level. IE the second best deck would be played against 10 more times than a fishy brew with a low winrate.
This is important to capture an important dynamic that came up at the start of Throne of Eldraine standard. In my opinion, even during the field of the dead/golos reign, Food decks built around Oko were the best deck when played against a wide field of FNM decks. However, the Golos decks were a very close second place AND they had a noteable favourable matchup against the Oko decks. Therefore if you only consider how the decks do against a random deck, the specific matchups against top tier decks are undervalued. If there were only two tier 1 decks (Golos + Food) , and one tier 1 deck has a clear advantage heads up against the other (Golos > Food), I think the classification should have Golos being the "best deck" in that metagame.
/end theory aside
Under this definition, every format has a theoretical "best deck" which is static and independent of a particular tournament composition. However, figuring out what this deck is , and figuring out if you should play it at the next tournament are very very different things.
The adage "play your estimation of the best deck" is a much better heuristic under my definition than "play the best performing deck from last week". The second statement is seductive in it's simplicity, but it's also not a good way to metagame in todays information rich environment...
Early in the format, your guess of what the best deck is (as is everyones) is very limited, so other factors should be driving your deck choice anyway. The best performing deck in early tournaments rarely turn out to be the actual best deck in the format and rarely succeed multiple weeks in a row.
Aside on - Resiliency
Perhaps my definition of best deck is closer to "the most powerful deck". A more complex definition of the best deck, should likely take into consideration a concept I'd like to call the decks "Resiliency".
A decks Resiliency is dependent upon - How bad are it's hypothetical worst case scenario matchups? How effectively can it counter-exploit it's exploitation's? How good against the rest of the format, are it's worst matchups? Is there sideboard cards that are extremely good against it?
An example of a very resilient deck would be The Oko Food decks of MC VI.
First, there weren't any decks with a very high winrate against the deck. Second, the decks that beat it didn't fair well in an open field of random decks. Third, the Oko decks could counter exploit a bad matchup if it focused enough sideboard space on it. An example is splashing black for massacre girl/noxious grasp against GW adventures. The splash barely effected the decks inherent power, yet turned a bad matchup to a good one.
An example of a medium-low resilient deck would be the BW Vampire decks of M20. Vampires were dominant as a strong counter measure to the Breeding pool ramp decks and Bant Scapeshift decks that were dominating prior. There were a few weeks where vampires were considered the best deck and boogie man of the format. However, it had a very weak matchup against both Grixis and Esper control, neither of which Vampires could do anything about, no matter how it sideboarded. Over time, as sideboards and metagames shifted, vampires had a few poor performances and quickly became another one of many decks. It was no longer percieved as the best deck.
In general, linear decks have a low resiliency and midrange decks have a high resiliency.
If you want the adage "Just play the best deck" to hold up as a useful heuristic, the definition of best deck should consider a decks resiliency.
Pt 2 - Theory of Metagaming
In theory, every format has a top tier of decks that are composed of decks analogous to a complex game of rock paper scissors. There is the best deck (under my definition above), a counter deck to the best deck and a deck that counters the counter deck. Rock, paper, scissors.
Equilibrium can allow for much more complex top tiers than that, composing of 5-10 decks. But the basic A > B > C > A , dynamic holds.
The "Optimal Deck" in a format is the deck that should be played at the highest frequency among tier 1 decks, against game theoretically optimal opponents.
If every player were playing game theory optimal, they would come to each tournament with a top tier deck, at a frequency governed by the optimal solution of a format.
However, given how many competitors in a tournament are making their deck choice based on irrational justifications, incomplete information or other non-competitive factors,
choosing a deck based on exploitative factors is always the correct approach.
Hence, why my definition of best deck considers how it performs against the jank of the format, whereas in game theory, that shouldn't matter.
Below are some of the higher level exploitative factors I consider when selecting a deck.
Rule #1-
Don't try to play deck to counter the public's perception of "best deck" unless your deck is also a very good deck.
The ONLY exceptions are for small tournaments with top tier players (Arena MC, Players championships)
Grand Prix's and MCQ's have WAY too many people playing a deck that they personally enjoy. Except under the most extreme circumstances, the best deck will not be >40% of the field. Even if the best deck is played at a frequency of 50%, unless you've broke the format, it is unlikely your winrate against that deck is higher than 60-65%. Your match-up against the field matters a lot.
"Play the best deck" is lazy. "Play a tier 1-1.5 deck" is a commandment.
Rule #2-
Prize structure and Winner's metagame matters.
If you do try to counter the best deck for whatever reason, make a conscious effort to figure out how your deck lines up against other decks in the field also trying to counter the best deck. The other counter decks should be disproportionate in the winners metagame if your read on the public's perception of the best deck was correct.
Most tournaments have sharp prize payout shifts, such that getting first may be 20x better than getting 9th. In top heavy structures, absolute winrate against the field starts becoming less important than your winrate against the best 8 remaining decks.
As a rule of thumb, if I'm targeting the public's perception of the best deck, i would prioritize
1) Overall deck quality
2) Matchup against the other decks that beat the best deck
As high as maximizing my winrate against the best deck. One exception to that, is if you think the other counter decks are "bad decks" that won't get past the jank in the early rounds at a high enough rate to worry about.
Rule #3-
If the deck I'm considering has a spotlight on it's recent strong performance, it needs to be resilient. If writers at Starcity games or Channelfireball have articles saying it's the best deck, you better believe people will be targeting it to some extent.. either with sideboard cards, or decks that directly counter them. Basically , never play WR tokens if it won the previous tournament.. or just never play it period. ;)
Hogaak, Eldrazi (winter), UGx Food were all examples of recent dominant decks that excelled despite aggressive hate. There were more examples than not that faltered.
In a metagame with a perceived best deck, "How well can it withstand aggressive hate?" would be the first question I'd try to answer during tournament prep.
Rule #4 -
Identical copies of decks that performed well the previous week, will show up in larger numbers the next week. Part of this is misunderstanding the concept "play the best deck", another part of it is rational. A lot of people can't/won't devote time to understanding a format for an upcoming event. As far as heuristics go, playing the best performing deck of a previous week is much better than picking a deck at random.
Rule #5 -
Seek and exploit information asymmetry.
If from my personal play experience, I believe I've identified an interaction or idea that has not gone public or is counter-intuitive to the public perception, I've gained an edge. If that edge can be converted into deck building, that's a huge deal.
An example might be, you discover that a matchup the public thinks is unfavorable is in fact favorable.
Recently, in my MCQ top 16, I identified mass manipulation to be a key mirror breaker in the Oko food mirrors, that I believe wasn't getting enough attention. I heard such criticism such as "Veil of summer and mystical dispute makes the card bad", card is win more etc. I saw very few lists running multiple copies at the previous tournament. Me focusing on Mass manipulation with others not respecting it was how I got my biggest deck building edge.
If that information was common knowledge, people would be playing the card themselves and preparing to counter it better after sideboard... eliminating my advantage.
At MCVII, I thought Great Henge was an extremely powerful card that had been ignored due to it's terrible matchup vs Oko (who had just gotten banned). Early in testing I was hoping that people would not catch on to this card, hoping they assumed it wasn't seeing play before, because it wasn't that good. Sure enough, the card was rarely seen on the mythic arena ladders.
It didn't perform quite as well as I'd hoped, as it received some indirect artifact hate from the power of Oven/Trail/Fires of invention, but it still was good.
Favourable information assyemtry is the largest theoretical advantage you can obtain from deckbuilding in a tournament.
Conclusion
"Play the best deck" is lazy. "Play a tier 1-1.5 deck" is a commandment. "Play a tier 1 deck that people don't know is tier 1" is how you break it.
If you have much experience in competitive magic, you've probably stumbled upon the saying "Just play the best deck" many times.
This saying is one of my biggest pet peeves in all of magic. I believe it is over used, misleading, lazy, and often.. not true.
The intention behind this article is to explore some insights of successful meta-gaming and hopefully provide a more useful definition of what a "best deck" is.
Part of the issue with using the term "best deck" to refer to the "best recently performing deck" is the language of "best" is connected to a deep intuition of what it means to be the best at something. When we think of the "best player" in hockey, we think of the best performing, highest skill , scored the most points etc. Attributes that are fairly static. If you were the best player in the world last week, the chances you will be this week is extremely high.
Most people throw the term "best deck" around when a deck puts up the most impressive performance at one or two events, early in the format. Generally these decks are what the best players at a tournament are playing.
The issue with applying that to a magic deck, is that the best performing deck at a recent tournament, is only a mild indicator that it will be the next week.. Many factors can lead to a deck performing great over a couple tournaments, such as ; pilot skill, an inefficient meta-game and variance.
Magic has a very complex rock/paper/scissors type dynamic to it, where one strategy nearly always has an effective counter strategy to some degree. The performance of a deck in one tournament is highly influenced by the composition of that particular tournament.
I see the term "best deck" thrown around in contexts, where it would be equivalent of saying "Rock" is the best in rock/paper/scissors, because it won the last round.
While I'm simplifying the narrative to get my point across, the point is, a decks recent performance is often over-valued and it's conflated with its inherent power level. This is especially true, early in a format, as information is still being uncovered.
I blame part of this confusion on the ambiguous definition of what it means to be "the best deck".
I would want a definition of a best deck to not only be a useful indicator of what to play, but have it be something independent of time, closer to the intuition of what it means to be the best athlete in the world.
My definition, "The Best deck" in a format is a deck that when perfectly piloted would have the highest win-rate against an exhaustive sample of random decks ranging in quality from failed FNM brews, to the tier 1 tournament decks. Weighted by deck quality.
theory aside:
It is essential that the frequency of the comparative decks be weighted according to their power level. IE the second best deck would be played against 10 more times than a fishy brew with a low winrate.
This is important to capture an important dynamic that came up at the start of Throne of Eldraine standard. In my opinion, even during the field of the dead/golos reign, Food decks built around Oko were the best deck when played against a wide field of FNM decks. However, the Golos decks were a very close second place AND they had a noteable favourable matchup against the Oko decks. Therefore if you only consider how the decks do against a random deck, the specific matchups against top tier decks are undervalued. If there were only two tier 1 decks (Golos + Food) , and one tier 1 deck has a clear advantage heads up against the other (Golos > Food), I think the classification should have Golos being the "best deck" in that metagame.
/end theory aside
Under this definition, every format has a theoretical "best deck" which is static and independent of a particular tournament composition. However, figuring out what this deck is , and figuring out if you should play it at the next tournament are very very different things.
The adage "play your estimation of the best deck" is a much better heuristic under my definition than "play the best performing deck from last week". The second statement is seductive in it's simplicity, but it's also not a good way to metagame in todays information rich environment...
Early in the format, your guess of what the best deck is (as is everyones) is very limited, so other factors should be driving your deck choice anyway. The best performing deck in early tournaments rarely turn out to be the actual best deck in the format and rarely succeed multiple weeks in a row.
Aside on - Resiliency
Perhaps my definition of best deck is closer to "the most powerful deck". A more complex definition of the best deck, should likely take into consideration a concept I'd like to call the decks "Resiliency".
A decks Resiliency is dependent upon - How bad are it's hypothetical worst case scenario matchups? How effectively can it counter-exploit it's exploitation's? How good against the rest of the format, are it's worst matchups? Is there sideboard cards that are extremely good against it?
An example of a very resilient deck would be The Oko Food decks of MC VI.
First, there weren't any decks with a very high winrate against the deck. Second, the decks that beat it didn't fair well in an open field of random decks. Third, the Oko decks could counter exploit a bad matchup if it focused enough sideboard space on it. An example is splashing black for massacre girl/noxious grasp against GW adventures. The splash barely effected the decks inherent power, yet turned a bad matchup to a good one.
An example of a medium-low resilient deck would be the BW Vampire decks of M20. Vampires were dominant as a strong counter measure to the Breeding pool ramp decks and Bant Scapeshift decks that were dominating prior. There were a few weeks where vampires were considered the best deck and boogie man of the format. However, it had a very weak matchup against both Grixis and Esper control, neither of which Vampires could do anything about, no matter how it sideboarded. Over time, as sideboards and metagames shifted, vampires had a few poor performances and quickly became another one of many decks. It was no longer percieved as the best deck.
In general, linear decks have a low resiliency and midrange decks have a high resiliency.
If you want the adage "Just play the best deck" to hold up as a useful heuristic, the definition of best deck should consider a decks resiliency.
Pt 2 - Theory of Metagaming
In theory, every format has a top tier of decks that are composed of decks analogous to a complex game of rock paper scissors. There is the best deck (under my definition above), a counter deck to the best deck and a deck that counters the counter deck. Rock, paper, scissors.
Equilibrium can allow for much more complex top tiers than that, composing of 5-10 decks. But the basic A > B > C > A , dynamic holds.
The "Optimal Deck" in a format is the deck that should be played at the highest frequency among tier 1 decks, against game theoretically optimal opponents.
If every player were playing game theory optimal, they would come to each tournament with a top tier deck, at a frequency governed by the optimal solution of a format.
However, given how many competitors in a tournament are making their deck choice based on irrational justifications, incomplete information or other non-competitive factors,
choosing a deck based on exploitative factors is always the correct approach.
Hence, why my definition of best deck considers how it performs against the jank of the format, whereas in game theory, that shouldn't matter.
Below are some of the higher level exploitative factors I consider when selecting a deck.
Rule #1-
Don't try to play deck to counter the public's perception of "best deck" unless your deck is also a very good deck.
The ONLY exceptions are for small tournaments with top tier players (Arena MC, Players championships)
Grand Prix's and MCQ's have WAY too many people playing a deck that they personally enjoy. Except under the most extreme circumstances, the best deck will not be >40% of the field. Even if the best deck is played at a frequency of 50%, unless you've broke the format, it is unlikely your winrate against that deck is higher than 60-65%. Your match-up against the field matters a lot.
"Play the best deck" is lazy. "Play a tier 1-1.5 deck" is a commandment.
Rule #2-
Prize structure and Winner's metagame matters.
If you do try to counter the best deck for whatever reason, make a conscious effort to figure out how your deck lines up against other decks in the field also trying to counter the best deck. The other counter decks should be disproportionate in the winners metagame if your read on the public's perception of the best deck was correct.
Most tournaments have sharp prize payout shifts, such that getting first may be 20x better than getting 9th. In top heavy structures, absolute winrate against the field starts becoming less important than your winrate against the best 8 remaining decks.
As a rule of thumb, if I'm targeting the public's perception of the best deck, i would prioritize
1) Overall deck quality
2) Matchup against the other decks that beat the best deck
As high as maximizing my winrate against the best deck. One exception to that, is if you think the other counter decks are "bad decks" that won't get past the jank in the early rounds at a high enough rate to worry about.
Rule #3-
If the deck I'm considering has a spotlight on it's recent strong performance, it needs to be resilient. If writers at Starcity games or Channelfireball have articles saying it's the best deck, you better believe people will be targeting it to some extent.. either with sideboard cards, or decks that directly counter them. Basically , never play WR tokens if it won the previous tournament.. or just never play it period. ;)
Hogaak, Eldrazi (winter), UGx Food were all examples of recent dominant decks that excelled despite aggressive hate. There were more examples than not that faltered.
In a metagame with a perceived best deck, "How well can it withstand aggressive hate?" would be the first question I'd try to answer during tournament prep.
Rule #4 -
Identical copies of decks that performed well the previous week, will show up in larger numbers the next week. Part of this is misunderstanding the concept "play the best deck", another part of it is rational. A lot of people can't/won't devote time to understanding a format for an upcoming event. As far as heuristics go, playing the best performing deck of a previous week is much better than picking a deck at random.
Rule #5 -
Seek and exploit information asymmetry.
If from my personal play experience, I believe I've identified an interaction or idea that has not gone public or is counter-intuitive to the public perception, I've gained an edge. If that edge can be converted into deck building, that's a huge deal.
An example might be, you discover that a matchup the public thinks is unfavorable is in fact favorable.
Recently, in my MCQ top 16, I identified mass manipulation to be a key mirror breaker in the Oko food mirrors, that I believe wasn't getting enough attention. I heard such criticism such as "Veil of summer and mystical dispute makes the card bad", card is win more etc. I saw very few lists running multiple copies at the previous tournament. Me focusing on Mass manipulation with others not respecting it was how I got my biggest deck building edge.
If that information was common knowledge, people would be playing the card themselves and preparing to counter it better after sideboard... eliminating my advantage.
At MCVII, I thought Great Henge was an extremely powerful card that had been ignored due to it's terrible matchup vs Oko (who had just gotten banned). Early in testing I was hoping that people would not catch on to this card, hoping they assumed it wasn't seeing play before, because it wasn't that good. Sure enough, the card was rarely seen on the mythic arena ladders.
It didn't perform quite as well as I'd hoped, as it received some indirect artifact hate from the power of Oven/Trail/Fires of invention, but it still was good.
Favourable information assyemtry is the largest theoretical advantage you can obtain from deckbuilding in a tournament.
Conclusion
"Play the best deck" is lazy. "Play a tier 1-1.5 deck" is a commandment. "Play a tier 1 deck that people don't know is tier 1" is how you break it.
Comments
Post a Comment