Shooting From The Hip: Doing the Math and Playing to Win

When preparing for a tournament, do you do all you can to ensure you take home the prize? Do you approach each match with the correct mindset? Or have you lost before you’ve laid your first land? With extensive theoretical examples based on the Team Standard format, Mike shows us how doing the math – and playing to win – will help us improve our game tenfold.

The best book I read last year was Freakonomics, by "rogue economist" Steven D. Levitt and co-author Stephen J. Dubner. Freakonomics is about the strange conclusions Levitt uncovered answering unusual riddles about everyday life (not your everyday life, probably, unless you are a sumo wrestler willing to take a dive, or a penniless inner city drug dealer); for a numbers-intensive economics text, Freakonomics is anything but dry… It’s quite brisk in fact, and surprisingly meaningful. For example, take my reading:

In a very fundamental way, Freakonomics ruined my life, or at least my self-image. I have long thought of myself as someone who makes very economical decisions. My good friend Tim McKenna and I used to laugh at people who made emotional or superstitious decisions rather than trying to maximize their personal utility, flying in the face of what was clearly the best path on the numbers. After I read Freakonomics, I realized that I was in the large and bovine camp of those who saw life the way I wanted to see it, or perhaps suspected it should be based on what I may have overheard some arcane or long-dead guru thought, rather than the way the world actually works. Jarring as it may have been, this realization didn’t offend me the way that – if they came to it in the same way – it probably offends most of my fellows… I just want to disassociate myself as much as possible with "witch"-burners, et al!

In which camp do you live? Do you make decisions – in Magic or otherwise – based on valuation and probability… Or do make them based on how they make you feel? I am not here to pass judgment on why you play Magic – hopefully the core of everyone’s experience is "fun" at some level – but if you want to win in the tournament setting, you are short changing yourself if you work from the wrong paradigm… Believe me, I’ve certainly made that mistake enough times!

That One Time Jonny Lost to Trinity

One of the strangest moments in my Magic life happened at the 2000 US National Championships. I had worked with the really tight squad of Jon Finkel and the OMS brothers, and with only one night of Constructed testing total, we elected to focus the remainder of our preparation time entirely on draft. The fact that “terrible terrible I” demolished “very very good Jon Finkel” in matchup after matchup made our eventual deck decision fairly academic, but to appreciate our truly insane amount of Limited preparation, you have to understand that we were keeping Neutral Ground open until at least 2am maybe four nights a week (this turned out to be awesome, because Jon ran the tables on Day 1, setting himself up for the eventual victory).

Anyway, none of that is the strange bit. The first round of Standard on Day 2 was like a Twilight Zone moment. I walked into the Feature Match area and found myself surrounded by… me. Jonny, SteveO, Bob Maher, Casey McCarrel… wall-to-wall Pro Tour Champions were playing Napster. I had won my match against Alex Shvartsman in seven minutes, so I got there in time to see… everyone lose.

Jon’s loss was really inexplicable, because he went down to Trinity Green, a storied Plow Under deck that was the buzz of the European Championship Season… but nevertheless a brutal underdog when faced with Napster. I was trying to explain how my deck was so good to Randy [Buehler] and [Brian] Hacker, but the caliber of players behind the Swamps and the conflicting tidal wave of ink in the "L" columns of the top tables completely drowned out anything I might have had to say.

"I know you believe in your deck," said Hacker. "But you know what I believe in? Math."

We all know the story, and ultimately, I was vindicated. McCarrel… Who knew where Casey had gotten his list? The Internet, perhaps? Bobby’s version was some "Corrupted" abortion that had taken out most of the uniquely powerful bullets to play Dragon Mask and extra lands in the sideboard (it made me quite sad at the end of the day, as I had no tools with which to advise Bob in his must-win battle with Mike Long for Top 8 other than "hope he gets manascrewed"). Steve bounced back to bang his adamantium shield around some enemy agents and ended up Top 16, including making the best play of all time against Dave Humpherys in the final round (you know a play is tight when you can make a 20% worse play against a reasonably developing Pro Tour Champion and still win in a blowout… but don’t make the mistake). Jonny was… well… Jonny.

But back to Brian.

Finkel’s victory in 2000 notwithstanding, Hacker’s assessment of the Feature Match area rewrote how I think about deck valuation (it took five years, but eventually I came around). I don’t care who made the deck or whatever baggage is attached to its lineage. Really, all that matters is one thing: How well a deck holds up in the metagame it has to play in. It didn’t matter to Brian that I had run through my little Regionals with one Swiss game loss… Jonny, SteveO, and Bob had to win in the Feature Match area against the best players in the world: Same Standard, completely different metagame, apparently.

That much wasn’t hard to grasp, but the idea that multiple metagames can exist in the same tournament is a relatively new one. You probably don’t remember, but it was noted liar Osyp Lebedowicz, I think, who first talked about the notion of deck choice for making Top 8 versus actually winning a PTQ (choosing Life versus Deep Dog). I didn’t buy it on 28 December 2004, but I do now. Now I think if you look at stuff like the Swimming With Sharks PTQ stats it’s pretty obvious: some decks are good at advancing to the Top 8 [against ostensibly weaker opponents], but those aren’t necessarily the best decks, because different ones, sometimes, are the ones that actually win.

Top 8 Winner
Ghost Husk
Heezy Street
B/W Control
U/R/W Weirding
Ghost Dad
B/W Hand
G/W Greater Good
Greater Gifts
Tron Wildfire
B/G/W Control
Critical Mass
Greater (no) Gifts
G/W Beatdown
Rogue Vore

In Team Trios Top 4 competition, Wafo-Tapa (the "beneficiary" of a tiny population of adherents) wins almost twice as much as Ghazi-Chord, which wins about three times as often as Ghost Dad.

Advancement is an interesting thing to look at, too. The Swimming With Sharks Top 8 (Top 4 in this case) lists don’t tell us how many people played which decks, only what decks made the elimination rounds and how they did. It’s by no means an exact science (because the Top 4s include almost no dead money decks, which you will invariably see in the Swiss), but you can also use the Top 4 totals tallies for decks as a faux read on the metagame. Critical Mass is super unpopular… Next to nobody plays it. Maybe only Ken Ho, ever, played the wacky Rogue Vore deck. Break down the rows in total, and you can probably put together a reasonable snapshot of the metagame as it develops. Once you are in that universe, the actual win rate makes a lot more sense: Heartbeat is super popular and super good. B/W Hand is super popular… and not particularly good. Tron Wildfire… can’t… win… ever… Why would you ever play it?

Sure, questions like "why would you ever want to play [awful deck Tron Wildfire]" seem loaded… but they’re based on the numbers; eight Top 4 appearances with no wins is the worst performance by any deck we tracked in the PTQs. On the flip side, the main thing that holds back the win percentages of best decks Heartbeat and Vore are that at some point as the best teams suck it up and all play these accepted decks, a maximum of 25% of them will be able to win. Ghost Husk is just awesome; it has fewer adherents than either of the other well-known B/W aggro decks… and posts three times the win percentage. The opportunity cost of playing either B/W Hand or Ghost Dad is not playing Ghost Husk

Given the statistics, doesn’t this seem like a violation of the Prime Rule?

I don’t care what your opinion is about which decks are good or not. If I start spouting things that are verifiably untrue, you should cease caring about my opinions. Opinions are all well and good; sometimes they are enlightening. They are completely irrelevant when you can argue the same point using math, facts, and logic.

Do the math.

Which decks are the most popular? Which decks win the most? Which decks do the "experts" say to play? When given the opportunity, which decks do they actually play? You have the data: Just. Do. The. Math.

Remember… That Time… Somebody… Netdecked?

Imagine, if you will, a New Yorker-style cartoon. You know the ones, the kind of scratchy black-and-white political squiggles that lampoon local politicians that are sometimes found in old newspapers or poke fun of impotent old men in the word-plagued pages of Playboy magazine.

Imagine this scene: A darkened subway tunnel. A little boy playing robots on the tracks. Twin lights casting the single-framed image in high contrast. The caption "… But Mommy, I like playing on the tracks!"

A quote:

"… I think all of this ‘if you aren’t playing this deck or this deck hate, then you are going to lose’ stuff needs to dial down a click or two. Despite a lot of new archetypes, it seems so many players are afraid to step outside of the safe arena of netdecking. Any that were on the fence would probably now fall the netdeck side of things, and whilst that might be strategically safer, its a shame. Moreso, as Flores genuinely seems to enjoy new deck archetypes – so how about a little less fear-mongering against them[?]"

I don’t mean to single out this poster because he seems really positive, and his position is a constructive one. However, it is basically the opposite of mine in terms of tournament preparation.

Imagine the cartoon we described above. Imagine that the oncoming train is labelled "Heartbeat of Spring" and the little boy is called "your deck choice." Wouldn’t you want to give that little boy every possible chance to not get trampled by several tons of oncoming steel? The whole point of keeping abreast of the metagame is to do just that: not get killed by a train that you can hear coming. In the current PTQ environment, Heartbeat is the best deck. You can look at the numbers! The past couple of weeks, it was not only 2/2 on wins, but 7/8 on Top 4 population. That means that if you want to win a PTQ, you should probably have a Heartbeat deck on your squad. At the very least, you have to be prepared that the other guys sitting across the table from you are going to have a Heartbeat deck on their squad. If you don’t approach the tournament with these things in mind, your chances of winning will be diminished. It’s not really up for debate.

The Opportunity Cost of Team Trios

Recently, a player sent out a request for team-mates for his upcoming PTQ. He declared that he wanted to play a terrible version of Owling Mine that splashed Green over Red (before you waste your time clicking, no, Storm Seeker is not in Ninth Edition), and that the players lucky enough to join forces with him would be allowed to choose from non-Remand, non-Recollect, non-Boomerang cards.

What is the problem with this (I mean other than the fact that he is robbing the team of the best deck in the format)?

When we started in the first Team Trios PTQs, my squad assumed out default opponents would be Heezy Street, Ghost Dad, and URzaTron. Our setup of Heartbeat, Vore, and Combo Deck had projected matchups of =,+,+; -,+,+; and +,+,+; presumably we were heavy favorites on the defaults (note that over two PTQs, our only loss occurred when facing a team whose configuration differed by arguably three-thirds, certainly two-thirds).

As the metagame has progressed, the best teams – at least those who are not introducing any new "tech" or odd decks – are projected to play Heartbeat, Vore, and Ghost Husk; against these, Heartbeat is pretty even… The mirror is the mirror, Vore is a slight favorite over many games, and Ghost Husk with Mindslicers is favored, especially in sideboard games over Heartbeat.

Should a player switch out of Heartbeat for U/G Owling Mine, he is trading =,+,- for -,-,-… And that doesn’t even suggest a very possible matchup – Swiss or Top 4 – where the opponent fields at least one R/W aggressive, removal-heavy Ghazi-Chord, or Heezy Street deck (and Heezy Street is still among the most successful decks on the Envelopes)!

Now, am I saying that no one should ever deviate from Heartbeat, Vore, and Ghost Husk? Certainly not! I take to perhaps too extreme the position that one should differentiate himself from the decks to beat in Constructed PTQ play. Sometimes that means playing "rogue" (or just bad) decks… in the case of Team Trios this time around, it meant playing good versions of fairly stock decks, but being a week or two ahead of the curve (and notice the direction that curve went). However, deviation from the statistical best decks – decks that any informed teams in contention are going to have (Heartbeat/Vore/Husk and Heartbeat/Vore/Hand won this week) – has to be done in favor of beating the best decks in order to be the best decks at any given PTQ.

That said, I still think that swapping R/W in for one of the Big Three (probably Vore, unless your Vore player is awesome) is the right way to go. R/W is one of the only decks that beats both Heartbeat and Ghost Husk, and it can run with Vore. R/W was built to play next to Ghazi-Chord, which according to Jeroen, beats Vore (I never beat Steve in testing), is a slight favorite against Heartbeat (most people take the opposite position), and crushes all the B/W aggro decks and most of the aggro and Tier 2 decks in the format (everyone agrees); it’s either +,-,+ or -,+,+ (i.e. it gives you a competitive advantage against the defaults). There is opportunity cost here, too… You can’t very well play R/W, Ghazi-Chord, and Ghost Husk… In order to play one or both of these rogue decks, you have to cut one of the acknowledged best decks from your squad.

Think Like Finkel

When I wrote "Who’s the Beatdown" seven years ago, I was under the impression that the most common mistakes players make is mis-assignment of role in a game. I think that a competing – or even more important – error is making decisions based on non-relevant information. For example consider this playtest game:

After reading a Richard Feldman article, I am inspired to try an Enduring Ideal deck. My opponent is B/W Rats, and the game goes fairly long, with him on Bob and a Shrieking Grotesque. I fire the Ideal, but Robert Maher immediately flips a Mortify and he cans my first enchantment, preventing him from the hard lock in a single turn. I then go for Zur’s Weirding; he attacks me to five but I figure I’m going to be at five anyway, and grab Form of the Dragon.

On my first Form trigger, I move to target his Shrieking Grotesque… after all, it is his only creature with flying.

This would be terrible!

I have Zur’s Weirding in play. He has a 2/1 flyer. I can take a maximum of four points per turn, with my life returning to five every turn. The Shrieking Grotesque, therefore, is completely irrelevant! Not that it matters because the opponent is essentially in the hard lock, but that move would have simply wasted a turn (i.e. it would have been the wrong play).

You’re awful, you’re probably thinking. Of course I would have targeted my opponent immediately.. Don’t lie to me, man. Better yet, don’t lie to yourself and say that you would not have considered taking out the 2/1. Lying to yourself is the worst! How do you plan on getting any better, lying to yourself? How do you plan on evaluating your plays correctly? You probably would have clicked it as the target, won in four more turns, and never realized – let alone remembered – you had even made the mistake.

It is sometimes difficult to get information out of Jon, but during a recent interview for our new book, I now think that the distinguishing thing about his thought process – more even than his superior strategy – is his ability to ignore or even forget non-relevant information. He can have complete understanding of how to run a complicated deck like Napster one day, and then literally forget the one-ofs that are in it the day after the tournament. Jon is therefore not distracted by the extraneous garbage that causes players like myself to make Magic play errors.

In a recent article, I made the assertion that Heartbeat is the best permission deck in Standard, or at least Team Trios. My reasons are, surprisingly, based on analysis edt did on Vintage some years ago. Basically, Vintage players can play huge volumes of great permission cards, but they don’t. The reason is that, between the artifact mana acceleration sported by threat decks and the nature of reactive cards, Vintage players don’t generally have enough mana to cast them all in a relevant window.

Conversely, here are some examples of dissenting opinion regarding Heartbeat’s permission suite:

"Heartbeat isn’t the best permission deck, I have no idea what that statement is supposed to mean, but it does make the best tempo use of Remand for sure."


"I think Flores is overestimating Heartbeat. Best permission deck? WTF? Has Mike ever actually seen a decklist for Counter-Angel or Wafo-Tapa?" … "So Heartbeat runs no hard counters. I think Hinders and Leaks (a better situational counter) in the two decks above are better than this pile."

These pundits seem to believe that the goal of permission is to stop threats. Permission was [primarily] used for the purpose of immediate threat suppression only during a short window from about 1997-1999. At the dawn of time, Brian Weissman’s Mana Drains were used to protect his Scepter lock; at the first Pro Tour, Mike Loconto did not run a lot of permission – only four Counterspells and one Deflection – which he judiciously aimed only at game-winning cards. Both of these players used their White cards to control the board and their counters, ideally, to protect their game plans. If you want to look at really dangerous permission decks, High Tide and Trix used Force of Will to force through their combinations against other permission decks. They tapped their opponents out, or forced down Necropotence on the first turn. Generally speaking, they were loath to use counters to stop their opponent’s stupid creatures! By the same token, Heartbeat could care less about "hard" counters most of the time. Muddle the Mixture might not be able to counter a six-mana Dragon, but who cares? It needs to counter a Scour, or a Cranial Extraction, or a game-ending Flames of the Blood Hand… A glacial 5/5 is, in most cases, relevant only in games that have already been lost. Remand is not a "hard" counter either, but it is still the best counter in Standard, in Heartbeat as in URzaTron. The powerhouse decks – specifically those with non-interactive elements – don’t need to answer every threat, because if they establish their minimum games, the majority of the opponent’s cards become completely meaningless. Do I have to answer your three Grizzly Bears if I can set up some sort of lethal Blaze?

If you ask a player why he may have just countered a non-essential card (probably some creature), his answer may well be that he wants some breathing room. Why do you want breathing room? Are you expecting to make mistakes so you need additional margin? Ideally, we aim for the sky and fall short. Where do you plan to end up if you are aiming for a half-assed level of performance?

The reason that Heartbeat is the best permission deck is that it can cast all of the counters it draws; whether it wins or loses a permission war is usually moot, because most permission-packing opponents will have a lot of mana tapped and counters stranded in their hands if things go right. Ultimately, the notion that certain decks should care about certain classes of an opponent’s cards – except in the context of how many turns they have before they lose to them – is one of the main things that differentiates Pro Tour Champions from us mere mortals.

In summary… play better, choose better decks, and stop lying to yourself, and you will win more.