Magic players are consumed with the quest to find the perfect deck. But what if the perfect deck was actually an imperfect deck? That’s a possibility that this article explores. This is my contribution to a very narrow, but important, gap in magic theory.
Every writer and most players have their own opinion about what sort of decks you should play and why. Zvi Mowshowitz has written his own diatribe on the matter. Here is what he says (I’ve bolded the important parts):
Players are playing what they like to play, both in general concept and in the specific cards, and it’s hard to point to their choices and call them unplayable. Allow me to do it anyway. A lot of what is being played is unplayable, because it’s just not on the same level as the top decks, but with so many decks out there it becomes hard to see that. The bulk of the field still believes in playing Magic "the way it was meant to be played" with decks that, at their heart, are fair. As I’ve said many times before, that’s crazy talk! Never ever EVER play fair when you have a choice, and if there is one thing you have in Extended right now it is a choice. With that choice, players are using decks like Rock and Red Deck Wins as some of their most popular choices. They’re attacking for two. I have nothing against attacking for two (or drafting the booster), but neither will give you strategies as strong as combination decks can be when they’re working properly.
After engaging Flores’ article on interactivity, Zvi continues his rant:
Interactivity is also highly dangerous on principle because you’re counting on your opponent to care. He might not care, and if he doesn’t you are dead. The more diverse the field, the more it will reward those who don’t bother interacting. The narrower the field, the more you can pinpoint the right interactions that matter to enough of your opponents. Right now, interacting is silly. It brings all sorts of elements of your opponents’ deck into play, often rewarding them for cards they have no business playing. By ignoring them, you force them to rely only on their clock and on the few things that can stop or slow you down. Force them to play your game. When there are lots of games, what’s the chance they’re prepared to play yours?
If you want to qualify, you want to be on one of the most powerful decks in the format.
If you are a hardcore spike, this philosophy of play will seem to articulate a coherent view of Magic. And yet, it seems to strike a notable contrast to the philosophy of Mike Flores. In a sense, Zvi is having a metalevel dialogue with Mike Flores. Here is an excerpt of Flores talking about his approach to deck selection:
I’ve always tried to think of myself as a tournament Spike because I love my edges. I’ll take every edge that I can, and more and more, have been studying the gamesmanship of the best players rather than the technical plays, or even the decks they play… But I think that after more than ten years, it’s time to face facts: I’m just not ruthless enough to be the Spike I fancy myself, proclaim myself. I like to win, but winning itself is not why I enjoy playing. My friends have chided me time and again that I lose Feature Matches not just because I am up against better players (hence matches being "featured"), but because I’d rather play to the crowd than against the other guy. I love tournament Magic because, frankly, what other terminus can there be when one directs so much of his time and energies towards this game? What other terminus than breaking a worthy rival against all the power and preparation within you, or challenging him – begging him – to break you? What other barometer than objective victory?
So I am a Johnny; a lusty, feral, angry Johnny. I can accept that. It’s better than being a Timmy, I suppose… any kind of Timmy. I must be a Jonny because I have literally no capacity for playing The Deck to Beat. It’s not just that I don’t do it, but that I can’t bring myself to do it. I’ve actually isolated what it is that I lust for in Magic, why I am always playing these insane long-game decks with intricate mid-range plans that don’t win for a million turns. All I want is to feel like I’m smarter than the other guy. I could stomach playing Vial Affinity in Mirrodin because, logically, there was no smarter choice.
I bolded what Flores wrote that was in stark contrast to the theory of deck selection presented by Zvi Mowshowitz. Specifically, Zvi says that you should play the non-interactive, broken deck. Zvi was, in part, reacting to one of Flores other articles on interactivity and denouncing it. Flores is arguing that he likes long-game decks with intricate mid-range plans that are by necessity highly interactive. Flores couches his language in the argument that he’s being a Johnny, but I don’t buy it. Flores plays fair decks, but he also plays to win. His decks do what Zvi decries: they have solutions to everything relevant.
While I think it is a real boon to the game that we can read the personal deck selection philosophies of Mike Flores, Zvi Mowshowitz, and other Magic greats, what is left absent and unspecified in this analysis is exactly what is meant by “winning.”
You might find that a strange or startling comment or observation. The assumption, explicit in Zvi’s case, is that he is interested only in winning the tournament. I think that this is wrong. It’s not because I don’t love winning. I do. It’s also not because I have anything against winning tournaments. I don’t.
In my view, he’s being imprecise – perhaps even sloppy. The goal isn’t to win the tournament; the goal is to maximize your expected prize earnings. Winning for the sake of winning is a Johnny trait – it admits that fun is derived from primarily from the honor of winning, instead of the loot that it entails. Naturally, in all cases, the first place tournament winner gets the largest prize. Thus, we rarely make the distinction between winning and maximizing your prize payout. However, if your goal is solely to win the tournament, it could paradoxically have the effect of decreasing your expected prize earnings.
Why?
Simple. The goal of finishing in the money is sometimes (and often is, as this article will elaborate) in tension, if not outright contradiction, with the goal of winning a tournament. I can show how this is the could be the case with a simple hypothetical:
Assume you are going to participate in a 80-person tournament with the following prize breakdown (this prize payout resembles a Vintage P9 prize payout):
1st Place: $500
2nd: $400
3rd: $350
4th: $ 325
5th: $300
6th: $300
7th: $280
8th: $280
9th: $200
The tournament has a cut to Top 8, and then single-elimination playoffs until a winner is declared.
Now, suppose you have two deck options (and only these two decks – perhaps because of card constraints or experience). You can play deck Alpha or deck Beta. You calculate that Deck Alpha has a less than 40% chance of making Top 8, but if it makes Top 8 you are almost guaranteed to win the tournament (let’s just assume for the hypothetical that you have a 100% chance of winning of you make Top 8). The reason is that this deck is extremely effective against Tier 1 opponents, but struggles in the swiss rounds against random opponents. Your expected prize value is then:
E= 40%*500 = $200
Deck Beta has a much better chance of making Top 8. You have about a 80% chance of winning enough rounds in the swiss to make Top 8, but you probably won’t go very far in the Top 8, because of your decks underpowered nature or inconsistency or some other reason. For the purpose of simplifying this hypothetical, let’s assume that your deck will not win a match if it makes Top 8. Your expected prize value of playing Deck Beta is:
E=80%*280 = $224
Now, I know that I’ve simplified matters here for the purposes of making this hypothetical, but this demonstrates that you have a better chance at winning more money by not trying to win the tournament (contrary to Zvi’s assertion), but by playing Deck Beta instead.
However, the simplification underestimates the gains to be made from selecting deck Beta. We should calculate the chance that Deck Beta could advance in the Top 8 into these calculations, which would further raise the expected prize value of playing deck Beta.
Many of you will fight the premise and argue that there is no difference between your chances of winning in Top 8 and making it to the Top 8 in terms of deck choice. I believe that this is clearly wrong and has helped perpetuate inattention to this important nuance. So much of our game is deconstructed and analyzed: matchups, deck choice, sideboarding, the mulligan, even things such as how to use the stack to your advantage and timing. Yet so little is dedicated to a serious analysis of the differences between swiss play and Top 8 play.
Last July, as I began preparation efforts for the Vintage Championship, my team began the deck selection process, debating and discussing which decks we should play at Vintage’s highest tournament. We were immediately struck by something that you very rarely see in Vintage tournaments these days: a prize structure so heavily weighted toward winning over making Top 8, that anything less than first place was not really worth winning. The prize structure in the hypothetical above resembles the prize structure awarded at Vintage’s major events. The prize structure at the Vintage Championship looked something like this:
1st place: $9,000
2nd place: $80
3rd place: $80
4th place: $80
5th place: $80
6th place: $80
7th place: $80
8th place: $80
A more tilted prize payout I cannot imagine. The reason is simple: first place is given the Mox artwork, which can be sold for close to ten grand. The rest of the Top 8 is given 48 booster packs and some Italian legends. Whereas the regular Vintage event maximizes the expected payout by trying to make Top 8 first and foremost, this event demands that players play to win the whole thing if they want to maximize their expected payout.
When confronted with that prize structure, I actually began to re-evaluate my normal deck selection process. Making Top 8 was worthless, beyond the honor of it.
In a short message to my team, I stated, without fully explaining, my view that I felt that a fierce combo deck, like Grim Long, would have the best chance of making Top 8. However, Meandeck Gifts, a control deck I created, would have the best chance of winning the tournament if you made Top 8. The trick would be getting into the Top 8. Turns out that I was right – Travis Spero playing Meandeck Gifts ended up taking home the painting.
No one on my team objected to that analysis or even asked why that would be the case. Vintage players, on my team at least, seemed to grasp intuitively the truth of my assertion.
Part of what I was reflecting was the historical fact that Mana Drain decks are the most likely decks to win Vintage tournaments. A Mana Drain deck had won two of the previous three Vintage championships, and have now won three of four. But there was more to it than that.
To understand why some decks are more likely to win tournaments and others more likely to make Top 8, we have to begin analyzing what it is that helps earn a deck a Top 8 slot.
Let’s proceed with a logical extreme. Let’s assume that you do not care whether you win the tournament and your entire goal is to make Top 8. In an article I wrote about making Top 8, I argued that surprise decks are hugely advantageous if your goal is to make Top 8. A surprise deck will enable you to play off your opponent’s inexperience in the match and induce them to make play mistakes they would not make if they had familiarity in the matchup. A great example is when my team took Oath of Druids, just after Forbidden Orchard was printed, to SCG Richmond – we played some of our biggest rivals who lost to our Oath list, with half of my team making Top 8. We knew that many of our strongest opponents would be playing Workshop Aggro using 4 Trinisphere. After extensive testing, we discovered that our the Workshop deck actually had a decent matchup despite appearing terrible on the surface. But we also knew that they wouldn’t figure that out until it was too late.
I remember hearing Shane Stoots remark that if only he had played his Goblin Welder, he would have won that match. The fear of Oath kept him from playing the Welder that would have recurred Memory Jar to kill the Oath player while Gaea’s Blessing was on the stack. Later, a fallen Shane Stoots could only lament that he could have won if he had played differently. That’s the advantage of playing a deck where no one knows exactly what’s in it or testing against it. Hindsight is 20/20.
Also, to make top 8, playing the most powerful deck is a huge boon. This is where Zvi’s analysis is point-on. You want to destroy your opponents as best as you can. Playing underpowered cards and decks makes you susceptible to getting overpowered and “outbrokened” – in Vintage or any format. But also, your opponent may or may not be able to stop you. Better to make them play your game than vice versa.
The downside to both of these points is exactly the problems that surprise and most powerful decks have in Top 8. This is exactly where Flores’ analysis comes in. First, the most powerful decks often lose if they are forced to interact. This is the entire point that Flores makes in his interactivity article. Second, playing a surprise new deck also has costs. You could easily find yourself in a position where you don’t have enough experience with this new deck yourself to really advance once you are in Top 8. You, yourself, are more prone to make errors. You can’t make errors if you want to win tournaments. When I played Ichorid at SCG Richmond last year, being the first to Top 8 with Ichorid at a major tournament. I was doing great, but I got into a tight situation in the Top 8 where I hadn’t encountered this type of threat before, and I miscalculated and lost.
In a sense, I’m suggesting that both Zvi and Flores are right, but at different levels. Flores’s analysis is most powerful in the context of the playoff structure of a Top 8, whereas Zvi’s analysis is most persuasive in the context of the swiss rounds.
And now, I’m going to present some anecdotal evidence that I’ve been picking up, as this theory has coalesced in my mind, to help explain this odd dichotomy.
Let me begin with an analogy to Baseball. For those of you who aren’t aware, Baseball has a rather intense playoff structure that very much resembles a Magic Top 8. It looks just like a cutoff to Top 8. Four teams from each league are paired up first in a division playoff, then a league championship, and then put into the World Series (think Top 8, Top 4, Finals).
Last year, in the National League side, the Cardinals slipped into the playoffs in what is known as a "Wild Card" spot for the team that has the best record in the league but didn’t win its division. The Cardinals had a 51% game win percentage in the season. Compare that with the Mets, the Yankees, the Tigers, and the Athletics whose game-win percentage approached 60%. And yet, the Cardinals ended up winning the World Series. In fact, the 2006 season of Baseball is notable for having a World Series champion with the lowest regular season victory total in history. How in the world did the Cardinals beat the Tigers and the Mets – teams who surpassed them by huge margins in the relevant categories? The Cardinals slipped into the Top 8 as a 6-1-1 8th seed, to borrow a Magic analogy, and yet ended up winning the tournament.
This pattern is very common in Vintage Magic.
Consider the 2004 Vintage World Championship. Mark Biller won the tournament with Control Slaver, taking home Mark Tedin’s recent rendition of Timetwister. What is not so well known is that as the final round of the swiss started, Biller appeared to be locked out of Top 8 contention. Most of the Top 8 had drawn in. Two matches played on with one winner emerging soon. The final match for Top 8 was locked in a heated battle. The combo player made a mistake and gave away a crucial game. The emotions were running high and the clock was running down. The match ended up drawing and being prideful, neither player was willing to concede the other into Top 8. Mark Biller slipped his way into the Top 8 as the last seed and, a few hours later, was photographed with that trophy.
Fast forward to 2006. Tommy Kolowith and Paul Mastriano were undefeated in the swiss, drawing in the last two rounds to guarantee 2nd and 3rd place in the Top 8. The only person who leapfrogged them was Nick Tradeau, who had to win six rounds of swiss and faced one loss to place first in his swiss bracket. Down at the very bottom, a Meandeck Gifts player named Travis Spero had managed to go 5-1-1 before the final round of the tournament. He faces the guy that took me out the previous round and quickly dispatches him, ending up 6-1-1 and the lowest 19 pointer. Travis snuck into Top 8 as the last seed and ended up going home with the most amazing painting they’ve given away yet.
And to me, this was totally predictable. This pattern recurs throughout sports and competitive enterprises of all sorts. How many teams that have the best record in football end up winning the Super Bowl? The Colts, who were dominant in seasons past, had a solid season, not dominant, and ended up with their first playoff success. The Steelers the year before – same story. And college football?
Can you possibly feel my pain at my precious Buckeyes getting annihilated by 4th ranked Florida? Florida wasn’t even supposed to be in the NCAA championship Fiesta Bowl. Michigan was ousted from contention by a spurious coach’s poll by one vote, and USC unexpectedly was taken out by their pathetic rival UCLA. 4th ranked Florida snuck into the championship game and annihilated the undefeated and thought-to-be unstoppable buckeyes with their (and my) Heisman trophy winner quarterback. Troy Smith was Chalice of the Voided, to borrow a Magic analogy.
Rich Shay is one of the most successful players in Vintage magic. Rich is hit or miss making Top 8s. He makes Top 8s at about half of the events he attends, never having made Top 8 at a Vintage Championship, but having won several Waterbury’s and at least one StarCityGames.com Power 9 tournament. What is so striking about Rich is that although he makes Top 8 less than half the time, he always wins a tournament when he makes Top 8. I can’t remember ever seeing a Top 8 that didn’t list Rich as the winner if he was in it.
The deck that is most likely to make you Top 8 is not the deck that is most likely to win the tournament. It’s the reason why the Cardinals won the World Series, why Travis Spero and Mark Biller won the Vintage Championship after slipping in as the 8th seed. It’s the reason why I remarked to my teammates on our team boards before the last Vintage Championship that I was torn on what to play. It’s also the reason that Rich Shay wins every single tournament he makes Top 8, but also the reason why he makes Top 8 less than 50% of the time.
It’s because the decks that make Top 8 most frequently are the most robust along the most relevant indicators: speed and power, tempo and synergy with decent doses of flexibility and resilience. But it’s also the reason that the decks that Win Tournaments are decks that are less absolutely dominant in their matchups, in their speed, and in their power, but overall present a more solid team with no serious weaknesses. Control Slaver has less favorable matchup win percentages across the board than lots of decks. Yet Control Slaver, once it makes Top 8, is a favorite to win the tournament. It’s the same reason why the Cardinals won the World Series. The Tigers had just dominant pitching when it came to strikes and hits. But the Tigers had a few sad weaknesses: lots of in fielding errors. This is something that just didn’t come up that often. The Tigers were so dominant offensively both at pitching and at bat that the small weakness was apparent until it was too late. Similarly, Grim Long or Fish are probably the two decks that right now best guarantee you a slot in the Top 8, but they are the least likely to win the tournament. Grim Long because of its manabase weaknesses, vulnerability to hate cards like Null Rod and Chalice of the Void, and mulligan problems that surface once in a while (consistency problems). Fish because, well, it’s underpowered. Both decks are fantastic choices if you want to make Top 8, but not so good for winning the tournament.
So now that you’ve read all this – what does this have to do with Finding the Flores deck? I look at Mike Flores creations and I see decks that are “Fishy.” That is, decks that are designed not to pwn and obliterate your opponent, but are supposed to be solid and force interaction – in other words, they are good, but imperfect. I don’t actually advocate playing Mike Flores type decks (I couldn’t believe the deck he submitted to the Invitational last year), nor am I advocating that you play imperfect decks.
What I’m offering is an analytic tool rather than strategic advice.
I’m saying is that if your goal is to Top 8, you play the 8-0 deck (more specifically, the deck that goes 6-0 and then draws the last two rounds) or the deck that loses one match.
But if you want to win the tournament, you want to be Travis Spero going 5-1-1, sneaking into Top 8 and then cleaning up.
Of course, Magic is complex and anything can happen. It is arguable that this pattern might reverse itself in formats with a much smaller card pool because there are fewer ways to force interaction. It is possible that you could play the monster deck, make Top 8, and never have to interact because you dodge the problem matchups. I saw this happen recently in a Legacy tournament. One goblins deck made Top 8. All of the combo decks were knocked out by aggro control decks, which in turn were knocked out by the Goblins deck that ended up winning the tournament.
But more likely than not, you want to play the solid and consistent deck with no glaring weaknesses in the Top 8 rather than the most powerful and dominating deck. You can have it both ways and sometimes you will, but to win tournaments, sometimes you want the perfect imperfect deck.
Have a great week,
Stephen