1. Pick a side. Have a purpose. Make a decision.
The word “decide” comes from the Latin “to cut off” … In English, the word is closely related to decision and decisive. What we are going to work on today is best accomplished—only accomplished, really—when we begin from a particular point
Really worthless deck playtesting comes from a notion of exploration. “How does this match play out?” … This is the opposite of the
correct way of approaching playtesting if we’re actually trying to make good sideboards or solve real problems. Did you know that 1 + 2 = 3, and that 2 + 1 also = 3? It’s insane, right!
Our playtesting, and the information of our innovation, should come from a place of advocacy, not objectivity, or even exploration. It is perfectly
okay for us to want one side to win… In fact, the whole rest of it is our laying out the path to get to the place we want to get. Who cares
if 1 + 2 = 3 and that 2 + 1 = 3 if we are trying to get to five? Remember: we control the inputs. Passivity and acceptance is a recipe for a mediocre
What we want to do is to figure out how to turn a 10/90 matchup into a 90/10 one after sideboarding… As has been done, time and again, by
sideboarding masters by eating up lots of space, killing their darlings, and rejecting—after first trying—bad idea after bad idea, but
snapping the inputs in and out like LEGO pieces, ourselves.
2. Approach the process with a clear understanding of expectation.
You: Argh! You’re so lucky.
Him: Yeah, because your Wild Nacatl was so going to get me on turn three.
It happens every day.
You can substitute whatever unplayable deck you want with whatever bad matchup you want, and it’s ultimately all the same. You have a horrendous
matchup in Game One, with a negligible win percentage. How can you proceed?
Ignoring it is a perfectly reasonable solution… Sometimes.
Can you win a Star City Games dot Com Legacy Open without ever facing [Dredge] (or whatever your horrible matchup is)? Yes. There are like 40
decks in Legacy, and the fact that Goblins, Zoo, and other Zoo so often make Top 16 is a testament to the fact that tons and tons of players are
willing to play unplayable decks; and you can hit each other all day in a merry-go-round of nobody having Force of Will or a turn three kill. Hooray!
Can you play through a Standard Open or even a Standard Grand Prix without ever playing Caw-Blade? Also yes! Can you win one without playing
against Caw-Blade? Um… no.
So if your bad Game One matchup is Caw-Blade, and you plan to win the tournament, you have to have some powerful—and repeatable—mathematics
if you have any hope of winning the tournament.
How good is “good enough?”
Sometimes you’ll see deck designers write things like “50% against the field” or that they can get to 50% after sideboarding.
This is awful.
Never listen to these people. They don’t understand math.
A deck that is 50% against the field is a ticket to a 3-3. Do you think Edgar Flores with his one-mana / 25-land Caw-Blade is looking for a fair
mano-a-mano with everyone he sits against every week? No way! He is angle shooting and looking for an edge.
So how much of an edge?
Let’s pretend you are the Zoo deck that isn’t very good against the Dredge deck (or substitute, again, whatever decks you want, with the
understanding that some decks just never beat other decks in Game One, unless something ooey-gooey is going on… Like Lands against Cephalid Breakfast
When you have a negligible Game One—and this should be obvious—you need to win both Game Two and Game Three in order to win the match.
Let’s say you win Game Two half the time. That would mean that you lose the match half the time after Game Two, and that you lose the whole thing
75% of the time.
See why getting the match to 50% is pointless and a stupid thing no one should really even ever talk about? BTW if you can “get the match to
50%” I am skeptical that your 50% even holds in Game Three when you are on the draw.
So… How much is good enough again?
Without going into every scrap of math (though it isn’t that hard, actually), if you can win about two-thirds of sideboarded games, you are still
going to lose the matchup more than half the time, and if you can win seventy-five percent of sideboarded games, you are just barely beating a
Do you know why we worked so hard, back in the day, to build sideboards like the one in Kuroda-style Red (which couldn’t be conventionally
interacted with, had all those Boseiju, Who Shelters All, and anticipated and stepped around known sideboard cards)? Because you actually have to win
like 80% of the sideboarded games to have any kind of reasonable expectation of winning a matchup with a negligible Game One.
Here’s a cheat: To get to 80+%, you almost always have to utilize an overload strategy or repositioning / transformation, or both.
When your matchup isn’t that bad (or even pretty good, main), you obviously have to improve much less in order to close out wins with your
sideboard. Remember, the opponent is a moving target though; keep that in mind when deciding how much space to devote to a matchup, as well as how
important that matchup is.
Corollary Takeaway: It’s very difficult to win a tournament, even with the best sideboard, when you are behind in Game One to The Deck to Beat, if for
no other reason than that you are drawing in Game Three… and that’s if you’re lucky and not already dead.
3. Bias practice games by starting with key cards in hand.
“GP Orlando was one of the few tournaments in the ‘Era of Affinity’ where Affinity was not considered the absolute stone-cold nuts. The main
reason for this blasphemy (aside from sheer stupidity) was that people hadn’t been playing Aether Vial in their Affinity lists up until this
point, so decks like Mono Red and U/G Shards had relatively decent matchups against Affinity. Orlando changed all that.
“I cruised through the Swiss with an Aether Vial list I got from Josh Ravitz (via Jelger Wiegersma). However, the best example of how
unprepared for Affinity people were came in the Top 4, when I was paired against yet another Mono Red deck. I had managed to win game one rather
quickly, yet my opponent didn’t seem too concerned as he stated, ‘I have a great sideboard against Affinity.’ The second game played out as
“Turn 1: (Him) Land Go (Me) Land Aether Vial
“Turn 2: (Him) Land Go (Me) Land put a Worker into play
“Turn 3: (Him) Land Go (Me) Attack for one and put an Atog into play
“Turn 4: (Him – with a smile on his face from ear to ear) Plays a fourth land and taps out to play Granulate! I sacrifice the Worker to
a Shrapnel Blast, untap, play a fourth, a Frogmite, and a Disciple, and kill him with the Atog.
“Lesson: Make sure your sideboard cards actually do something!”
Remember what I said back in #1 about picking a side?
One thing you can do to really jazz up your side is to cheat in playtesting.
Paul Jordan likes to make fun of me for cheating in playtesting; in public, I have mostly denied this, but I actually cheat quite vehemently… but I
do it with very good reason! Especially back in the days when I used to play both sides of matchups for twenty hours a week on Apprentice, I would
pre-load the deck I wanted to win with [sideboard card x] in its opening hand.
This does a lot of things.
First of all, it should help “your” deck win.
Doesn’t it feel great to win?
Well what if you don’t win?
Paul is in particular right—or at least looks funnier—if I lose in testing, while cheating.
Biasing opening hands in testing not only gives you a potential sense of accomplishment that can be vital in order to “keep going” in your
advocacy of a deck against the oppressive shadow of that hard-to-beat Deck to Beat, but it tells you whether your sideboard cards are good or not.
How awful would it be to show up to the tournament, overvalue an opening hand based on the presence of a sideboard card… and then lose when you not
only draw it but play it?
Isn’t it better to learn that it sucks ahead of time?
If you’re biasing your hands and not getting the results you want, that can be a signal to abandon a deck, or at least pursue a different
line. If you are winning, that can be a signal to figure out how to consistently get the kinds of hands that you have in front of you as you
win sideboard games.
4. Know what hands you can keep.
At the 1999 US Open, I was surprised to watch my friend John Shuler mulligan so aggressively. John was playing Malka Death, Sol Malka’s Survival
of the Fittest deck, years before he coined the term “The Rock” … John always seemed behind going into sideboard games. In
particular, I was surprised to see him go to five cards, on the play, against a blue combo deck… while down a game. John further went to
four cards by playing his Vampiric Tutor for Duress.
Yes, he won.
In fact, he won all of them.
John won the first US Open of 1999, splitting with Jamie Parke in the finals.
I asked him about his aggressive mulligans (such were almost unheard of at the time), and he responded that you can never beat a blue combo deck with
[The Rock] unless you Duress them, so you can only accept hands with Duress, or some proxy for Duress, viz. Vampiric Tutor.
Long story short—know what kinds of hands you can keep.
You have very different criteria for keeping a hand in a sideboarded game than you do in a Game One. Presumably in a Game One, you don’t know
what your opponent is playing, so your algorithm is probably around whether you have enough lands and spells to operate at a basic level. In a
sideboarded game, your algorithm is actually simpler: can I win this game?
I have become absolutely fearless around mulligans in recent years, due in large part to the influence of Gerry Thompson. I’ll mulligan to four, or
even three, knowing that there is no point griping about the loss of yet another card, if my five-card hand wasn’t going to get there
Remember just one bullet ago when we were cheating in playtesting, starting out with key cards? If you can only win the games where a key card is
present, that is probably a signal for the strategic mulligan.
5. Write down all the cards, where they were, and what they were doing.
It’s one thing to lose games, and in particular Game One.
It’s another thing to be able to win sideboarded games.
What contributes to losing Game One, though?
How do you know which cards were effective in sideboarded games?
If you write down all the cards you drew in Game One and what zone they were in when you lost, you can gain a great deal of information.
Is there a card that you decided, pre-game or pre-testing, was important… But was stuck in your hand as the tenth poison counter festered up your
neck? That is a signal.
How about a flagship creature? Where was it and how effective was it at what it was supposed to do?
It turns out that there are a great many matchups where Goblin Guide—specifically on the draw—isn’t actually very good. But if you
don’t isolate Goblin Guide’s nonperformance, you might never realize that your best first-turn play on the play is actually kind
of a dud going second (at least in this matchup).
6. Play both Game Two and Game Three.
This is probably hella not obvious to most people.
You play a handful of games and declare a sideboarded win percentage. Hooray!
That’s not actually how it works.
Typically Game Two is a completely different animal from Game Three. For one thing, you’re typically on the play in one and typically on the draw in
the other if you have lost Game One.
That can change what cards are good. For example, Mana Leak is usually pretty good when you’re on the play and usually loses much of its value when you
are on the draw. So if your winning is contingent on having Mana Leak, very likely your performance will wane considerably in Game Three.
Your opening hand evaluations will also fluctuate depending on whether you are on the play or the draw.
For example in the Faeries mirror, Broken Ambitions could serve as a passable Bitterblossom proxy in terms of opening hand evaluation if and only if
you were on the play. On the draw, no Bitterblossom usually meant a snap-mulligan (especially if you don’t have Thoughtseize), whereas on the
play, you might keep an otherwise strong no-Bitterblossom hand if you had an answer to the opponent’s Bitterblossom in Broken Ambitions.
Your Game Three sideboarding strategy might be completely different, your numbers might mess up your #2 evaluation, or you might even realize, over the
course of many trials, that going second is where you want to be!
7. Make your opponent a moving target… but don’t overcomplicate things.
You have to keep it real.
That’s why #6 suggested you practice games going second… Biasing your opening hands is a fine tool, but it has a specific goal, and
that’s not an implementation you can probably carry with you in tournament play.
What does it mean to be a moving target?
Your opponent, very likely, will have the tools to make some kind of improvement via sideboarding. Acknowledge those tools in playtesting, but
don’t overthink the problem. Remember, as an advocate, you’re more concerned with the other deck’s success than this one.
The “moving target” assumptions you make in playtesting can become opportunities, not just threats.
For example in Critical Mass (sideboarding against Mono-Blue), it came out that Critical Mass had only one target for Threads of Disloyalty
(Sakura-Tribe Elder). It is more-or-less pointless to leave Threads in for that decidedly terrible target.
Knowing that our opponents had no Threads targets, we could side in Jushi Apprentice with impunity… Our opponents would have no good answer.
If the opponent has left in Threads of Disloyalty, likely we aren’t losing to him, anyway… Then again, he might have just leveled us.
As a final note—but not really a bullet—I would challenge all of you to seek uncertainty. So much of what we do in Magic is to try to
establish a position of certainty… But the fact is being certain of a loss is a kind of certainty, too. Great sideboarding is just as often about
breaking up the expectations of others as it is about getting your own plan on. Being able to look at the process from both perspectives, therefore,
can expand the horizons of your possible plans.
So go out there and make some awesome sideboards already!