After some observation, it has occurred to me that people don't know how to sideboard. No, really, they don't. Whenever an article with a decklist comes up, the most frequent question is always, "how do you sideboard versus deck X?" When people PM me questions, it's also generally on how to sideboard. So, today I'll try to cover sideboarding a little bit. Of course it's a topic that has existed since the dawn of time, and I don't mean to reinvent the wheel with this article, but I feel like a piece summing up all my thoughts about it would be useful.
Sideboarding is really the sum of two processes - building a sideboard and actually applying it. It would be a mistake to separate them, since one does not exist without the other - you build a sideboard with what you're sideboarding in mind, you don't just get to 15 cards and then realize you have 11 cards to side in against UB but only 3 to side out.
So, how do you build a good sideboard?
Some people say you should practice a lot of sideboarded matches - I disagree. Of course, given infinite time you would, but I believe your time is much better spent practicing game ones and then applying your knowledge to the sideboarded games. That is because:
If you don't know how the matchup plays out, you will sideboard incorrectly. It is very time-consuming to test all the possible sideboard choices to identify what you would maybe want to board in - it's better to play game ones so you narrow down your choices a lot. By the time you can make this choice, you don't know your sideboard yet, and you don't want to spend a lot of time figuring out how the game plays with a certain card if you don't even know you'll play it.
- Playtesting with your game one configuration will help you in more games than playtesting with sideboard. In a tournament you actually play more games with a sideboard than without, but for every match your sideboard configuration is different - your sideboarded games against Mono-Red will not help you much against Valakut, since the sideboard will be different, but your deck will be the same in game one and, though the match won't be played the exact same way, it'll overlap a lot more.
- Knowledge about a game one will help you in sideboarding, but knowledge of how a sideboarded game plays out will rarely help you game one.
- You don't know what they will be sideboarding. You might playtest against a different sideboard strategy than the one you will face in the tournament, and then it'll actually hurt you instead of helping. When you play game ones, it's much easier to deviate from your accepted sideboard guide in face of the unknown.
Of course, that is not to say playtesting sideboarded games is useless, far from it - I just think you get a lot more out of playing pre-boarded games if you have to choose between them. There are two specific points where I feel it's worth it to playtest sideboarded games, though:
1- You will only play a deck if it beats another deck after board. Imagine, for example, that you want to try a deck that loses game one to the most popular deck, say, Caw-Blade. If it keeps losing after board, then you'll not play the deck, but you might if it starts winning games 2-3. So, by testing the sideboarded game first, you actually save time if it doesn't work.
2- You believe something changes the dynamics of the match to a point where you can't theorize it, for example, Koth/Thrun in Valakut. It is hard to theorize how the matchup will be with this kind of sideboard, because they change the focus of the game and attack an angle that was irrelevant before. It is possible that they're the best card ever for the matchup, and it's possible that they don't do enough, and you have no way of knowing without actually playing. For PT Paris, we were concerned about Thrun against our Caw-Blade deck and if we needed to change something in our board to combat it, but after only a couple games we realized it was not a big deal because it couldn't beat what we already had (Sword). If the Valakut players had done the same testing (which is a bit unfair because we knew their deck would be popular but they didn't know about ours), then they would probably not even have had Thrun in their boards.
You cannot test every card. If you're like me and decides on your deck the day before the event, then you cannot test most cards. If that is the case, then you need to have a quick "guide", something that filters through what is available and tells you what could possibly be a good sideboard. I will again follow my favorite method, which is to split it in the three major archetypes, control, combo and aggro. I understand some decks are all of those at the same time or they change from match to match, but I still think it's the better method for me to write something that is understandable.
Control decks have the best sideboard in Magic. People often say pro players prefer control because you get to make a lot more decisions during the game, and though that is true to some point, since games will generally last longer and you'll rarely be powerless, one of the things that really drives me towards control decks is their powerful sideboard.
Taking out cards in a control match is relatively easy. In a control deck, your cards rarely all work for the same purpose - some of them are good to win the game, some of them stall early rushes, some draw cards, some stop powerful spells. Once you identify which kind of card you do not need, you just take out those. You will generally be able to identify them - removal is worse versus control, counters versus aggro, etc. If you can't, then, well, SHOULD HAVE PLAYED MORE GAME 1s!!!!11!!!!1
The fact that all your cards have different roles means you get to side out all the ones that are bad against a certain kind of deck, improving your deck tremendously. Where before board you can always draw "the wrong half" of your deck, after you sideboard your entire deck will be a machine to beat what you're facing.
I generally like cheap answers in my control boards. Against aggro decks, cheap answers are what you need to stay alive to play your more powerful cards. The thing with aggro decks is that each card gets exponentially better with every other card in the deck - for example, a Goblin Guide is not very threatening, but two Goblin Guides is more than twice as good as one Goblin Guide. By having cheap removal, you make sure they do not get to the mass of cards they need to make all their cards effective - i.e. if you kill the creatures, you automatically kill the burn as well. For this reason, cards like Disfigure, Oust, Lightning Bolt, Pyroclasm, Doom Blade, and Condemn will generally be in my control sideboards.
Against control decks, you also want cheap answers because you want to be able to play many of them in what will be the "critical turn." Cards like Negate, Duress, and Spell Pierce are excellent in most control mirrors, and so are cheap threats that you can stick early in the game without tapping out when facing something big or that let you leave mana up to defend them (Jace Beleren, Vendilion Clique). In general, control does not need specific cards against combo because all the control stuff already overlaps.
Nowadays, the best control deck in Standard is not really a "control" deck, and the best combo deck is based on a creature, so the normal rules would not actually apply. Negate, for example, would be really bad when the key cards in your matches are Stoneforge Mystic and Primeval Titan, which forces you to play Flashfreeze and something else instead.
Another kind of card I generally want in my control boards is a way to "go even bigger.” Sometimes you will take out a lot of expensive cards for a lot of cheap answers but then you won't have anything to do with all the time you're buying, or with all the disruption you're presenting - it doesn't help to Duress and Thoughtseize their entire hand if you have no follow up. If both players do that, then it's going to become a big attrition war and suddenly both will be left with not much, and then having something big will trump all.
In a game one, you sometimes don't care how you kill them - it'll eventually happen - but in games 2 and 3 you generally want the game to end faster so you don't give them time to recover from your onslaught of cheap disruption, be that for creatures or for spells. Imagine the two scenarios:
That is not really optimal - you've left them with nothing, all their burn is useless, but you take the risk of simply losing a topdeck war. What if they now play Koth and you start drawing removal? It is much better if your follow up is Grave Titan, Wurmcoil Engine or something that means they will really not get out of their hole.
Those cards also get better postboard because the players will generally take out most stuff that deals with them, because it's not good against your deck. There are many cards that play this role - my deck in Paris had 2 Baneslayers, for example, for when you boarded in Ratchet Bombs and Ousts, but people also play Sun Titan, Precursor Golem and even Geth.
Another important thing is that sideboarding doesn't have to be the same on the play or on the draw. Sometimes you need fast answers on the draw, but on the play they are not necessary. Mana Leak is the card that changes the most - in the UW mirror, for example, Mana Leak on the play counters t2 Mystic, but on the draw it doesn't do anything, so you can take some more out, and the same happens when facing a card like Bitterblossom. Against a Boros deck, you really want Mana Leak on the play, but on the draw it might be too slow and you'll be overwhelmed.
Take, for example, our UW list from Paris (which I would not play today, but just for the sake of the example):
In this sideboard, we have a bunch of general cards against aggro, because we don't want counters against them and there are many here. You can definitely take out 4 Spell Pierce, 1 Mana Leak, 1 Deprive and 1 Stoic Rebuttal against every one of them, and against anything but Boros on the play I was also taking out the other two Mana Leaks. This means that against WW, Kuldotha Red, Vampires, Elves and Boros on the draw we had at least 9 slots. If our board against aggro was specific, say a bunch of Celestial Purges over Ousts, then we'd be left without cards to board in against three of those decks. For this reason, we went with versatile, and all our 9 - 4 Oust, 2 Baneslayer, 3 Bomb - could be boarded against all of those. We could have played Condemn over Oust, which is much better against Boros, but we also wanted Oust for Valakuts with Cobras, so again versatility won.
For control decks, though, we did not need versatility. We were taking out one Day of Judgment and one Lifestaff against something like UB, and we were boarding in 1 Deprive and 1 Sword of Body and Mind. The other 2 slots could be specific - Flashfreezes, for example. Sure if we had another hard counter instead - say Rebuttal - we would have boarded it in against UB too, but we wanted stuff to deal with Grave Titan anyway, so we felt like we didn't need anything else to board against the other control decks so we could afford the actual best card against Valakut, even if it was only good against Valakut.
Now take the new version, UWb. UWb has no spell pierces in the maindeck, instead having Inquisitions, a much more versatile card, and it's good against both aggro and control. They also play less hard counters, which means that suddenly there are not many cards you want to take out against the aggro decks. Because of that, you no longer need a bunch of generic cards that you bring in for a lot of matchups - you can specialize more in your sideboard, with cards like Flashfreeze, Celestial Purge, Divine Offering (which I dislike by the way, but that's neither here nor there). The moral of the story is, in general, the more specific your maindeck, the most versatile your sideboard should be, and vice-versa.
You also have to watch out for the times where you end up sideboarding out a card for something that is only marginally better. Imagine you had 4 Flashfreezes in your sideboard and, after boarding against Valakut, you ran out of things to take out and took out a Mana Leak instead. Sure, Flashfreeze is better than Mana Leak, but it is not that much better for you to spend a sideboard slot with it - if you get to this point, play 3 Flashfreezes and leave the Mana Leak in there (unless you actually need 4 for a different deck, then make the swap).
Let's take a look at Edgar Flores' UW list:
After a quick count, it comes to mind that you have too many cards you want to side in against both the mirror and Boros, in comparison to what you have to take out. You end up taking out cards that would actually be good in the matchup (say Mana Leak on the play versus Boros) so that you can accommodate your sideboarded Divine Offerings, which are not even that good. If Divine Offering is making you take out something that is good in the matchup for something that is only slightly better, if better, against the two big decks you'd want to bring it in, then why is it in the sideboard? It's better to use that slot for something very specific, like more Flashfreezes or another Sun Titan or a Baneslayer Angel or Ratchet Bomb or or...
Aggro's sideboard is a lot harder than control's most of the time. The reason for that is that all the cards in an aggro deck will generally do the same - they kill the player very efficiently. You cannot take out a one-drop for a one-drop that kills them better, because if such a card existed, it would be on your deck instead.
In general, the easiest card to take out is the removal, because it is apparent when the removal is bad - when they don't have creatures you want to kill. It is also easy to know when a card is too slow - if you're playing against a turn 4 combo kill, a card that costs 5 is not good. It is very complicated, though, to take something out when removal is what you want - you'll likely have even more in the board, because you don't have much else. If that is the case, you can try to go big.
Going big is a term for increasing the power of your deck at the expense of speed. It is, for example, what LSV did with his Boros deck against Vampires - he took out Goblin Guides and Plated Geopedess for Cunning Sparkmages, Gideons and Inferno Titan. What he does here is assume the control role, with a ton of removal, so that the big bombs can win. In fact, it seemed LSV took out Goblin Guides every match. Is that because he dislikes Goblin Guide? Not necessarily (though he does). Sometimes, you just take out the same card in every match.
Whenever that happens, there is always someone ready to jump and point out that "if you take the card out in every matchup why do you play it?” The answer is that it's good in game ones, but it is not good in game twos because the opponent will bring cards to combat it that they didn't have before. Seems very simple to me, yet people often refuse to understand it, just like a card could be good in every game two but not good in game ones because people take out their answers to them that they would originally have in the main - such as bringing Baneslayer in against everything out of a deck with no creatures, if it's there game one it'll just die, but game two it'll often live because they took out their removal.
In the case of Goblin Guide, it goes back to what I said before - they're exponentially better with your other aggressive cards. Once the opponent has a lot more cheap removal, your guys are going to die a lot more often, and Goblin Guide will have no buddies to help him deal the 20 damage. In the end, you'll be left with a 2/2 that cannot win the game by itself, and your opponent will have the time to use the extra cards he drew, something he didn't have in game one because you just overwhelmed him. So you take him out for cards that will be good in the situation where the opponent will deal with most of your other threats, like Gideon or Koth.
This solves the problem when you play against other creature decks or other decks against which it's profitable to go big, but how about the others? For the others, you don't really have as many cards to take out - generally the removal. What happens then is that, instead of having 8 or so "generic cards against aggro", like control does, you turn to more specific cards. You only have 4 slots against each deck, so you play the 4 best cards against those, regardless of how good they are against anything else because you have nothing to take out to bring them in against the other decks anyway. As an example, assuming red aggro and black aggro are the two big decks, then a control deck would probably play 4 Condemns 4 Celestial Purges, but an aggro deck would play 4 Circle of Protection: Red and 4 Circle of Protection: Black.
In an aggro deck, you also have to make sure you don't mess with your curve too much if you don't plan on going big. Sometimes you will have a card that costs 4 and is good against something, but then you might have to take out the other card that costs 4, even if it's not necessarily bad against that deck, because you cannot support this many fours. Sometimes people will say "I have eight one drops, so I can take out some right?" You might even be able to take them out (i.e. Goblin Guide in Boros), but it is not because you have 8 - you have 8 because your deck wants 8, otherwise you wouldn't play this many. You have to make sure your sideboarded deck doesn't also want 8.
One thing you can do when you do mess with your curve, and that's not only with aggro decks, is to also change your land configuration. It is not uncommon for a deck to have a land in the board, for matches where they think it's important to hit early land drops or for an aggro deck when it goes big (can't really play 6 drops in your 22 land deck), but it is very rare for someone to actually take out a land, and it's something people should do a lot more often, in my opinion. When I played UB at Worlds, for example, I was very comfortable taking out a Tectonic Edge against Mono-Red Aggro for a sideboard card, since I was lowering my curve with Disfigures and Doom Blades, and if I was going to have mana problems they would likely be colored anyway (because I'd want to play multiple Inquisition/Preordain/Disfigure/Leak/Blade in the same turn). That doesn't mean you should just go around taking out lands when you don't know what to do, but keep in mind it's an option.
Even more than that, you have to make sure you don't mess with your plan. You can't forget that you're still an aggro deck, and sometimes boarding in something strictly reactive is not the best. Imagine, for example, Zoo vs. Counterbalance in Legacy. If Zoo boards in too many Krosan Grips, the game might get to a point where CB is not even going to need a Counterbalance to win, because Zoo will have drawn too many slow and reactive cards. As a rule, don't side in a card that means your opponent is not going to need whatever you're dealing with through that card because you drew that card instead of a threat.
Unlike control, aggro decks usually attack control and combo in different ways. Against control, you want something hard to deal with, resilience, but that is also a threat if preferable. Against combo, you want something very specific to deal with them. Luckily that goes perfectly with the "plan" of not having much to board out in other matches anyway, which means you can afford some slots for cards like Mindbreak Trap, Ethersworn Canonist and Tormod's Crypt.
Take the Boros decklist that LSV uses in his videos, for example:
Here, we see the "going big" package - four Sparkmage, one Collar, one Titan, two Gideon and sometimes 4 Firewalker, sometimes 1 Koth, that you actually side in against both control decks and aggro decks.
Now take, for example, Patrick Sullivan's Legacy list:
This deck cannot support "going big", because it is too dedicated to aggression (and it doesn't really need to go big anyway because it's fine versus creatures). Therefore, it utilizes its sideboard for very specific cards - eight against blue combo decks, four against creatures and three against artifacts. It simply takes the best at doing whatever and plays those, not caring if they overlap or not. He only has 4 cards against creatures, but that is irrelevant because he only has four cards to take out anyway, which means he doesn't have to play anything versatile in the other slots, going for maximum efficiency instead.
Combo is the hardest to sideboard with, because every card in your deck had to fight for a precious slot. You can't take out combo pieces, you can't take out disruption, and you can't take out card drawing, what can you take out?
Generally, against aggro, you will take very few cards out for very few cards. You do not need something to be faster than they are (and if you did it wouldn't exist), so what you need is something that deals with what they're bringing for you. You have to be careful not to overload on answers, just like aggro, because if you do they might not even need their answer. Combo decks usually have a ton of card drawing and tutoring, so you don't really need that many answers anyway - you can often go by with say one bounce spell or one removal, or even just one bounce. If you're going for one slot, you want to make sure that it's the one that will solve the most scenarios - for example in Flash Hulk the best card for your one bounce spell was Rushing River, since it dealt with any two of Leyline of the void, Tormod's Crypt or Engineered Plague. Echoing Truth was a better card to draw when they only had one, but this is your only slot - what are you going to do when they play two Tormod's Crypts, or one and a Leyline, just lose?
Against control you want disruption, cards like discard, Orim's Chant or Mana Short effects. The game will go longer, so you don't have to be very fast, which means you can generally afford to take out mana producers - Lotus Bloom was a very frequent side-out when I played combo decks. For combo vs. combo, it's going to depend on whether it's just a race or if you can disrupt them - if it's just a race, board like aggro, if you can disrupt them, board like control.
Combo decks have some peculiarities, for example a transformational sideboard, which is a very potent weapon if executed correctly but one that does need previous testing. The idea of the transformational sideboard is that you attack them in an angle they were not expecting, rendering all their own sideboarding useless. You do not care about weakening your combo this way - in fact, you usually take it all out. It is very important, though, that your sideboard does not use the same resource as your combo and does not get splash-hated by it. It is futile, for example, to board in creatures when you're playing Cephalid Breakfast, because they will leave their creature removal in anyway. It is just as bad to bring in Tarmogoyfs with Dredge, because your graveyard is the resource they'll go after. It is also important to figure out if you have enough cards to bring in - a Dredge deck would have to have a 30 card sideboard or you'd be left with Careful Studies, Stinkweed Imps and Narcomoebas in your beatdown deck after board, for example.
In recent memory, I can think of one combo deck that had a transformational sideboard in Standard and Extended - Pyromancer Ascension. Of the possibilities, the most interesting was the Polymorph one - it used Khalni Garden and Spawning Breath to Polymorph into Emrakul. People would board in enchantment removal, and even if they didn't, they would at least board out their creature removal. Even if they didn't, they would likely not wait on their Bolt for your Spawning Breath token, because they wouldn't know what was going on. After they see it game two, if there is a game 3, it becomes really interesting - do you bring it back to their original configuration? Are they boarding in cheap removal or leaving in enchantment removal? From my experience, the player playing against the combo deck will generally try to do both - whereas in game one he has four Bolts and in game two four Disenchants, for games 3 he will have 2 of each in his deck, because he doesn't want to simply be leveled and completely unprepared - he usually cannot afford 4 Bolts and 4 Disenchants, because his deck gets too reactive. Then, you figure out whichever you can beat.
The disadvantage of this strategy is that, when they know what you're doing, your deck becomes much worse. At Worlds, I saw Guillaume Matignon with 5cc play against an Ascension guy, in extended. Guillaume expected the guy to bring in Pestermite/Splinter Twin, so he left his Lightning Bolts in. At some point I saw Matignon's hand and it had 3 Bolts, and his opponent's had a Pestermite. By using the transformational sideboard, Matignon's opponent turned a match in which he was perhaps a favorite into one that he could not really win, by giving his opponent cheap disruption and an edge in counter wars, since his threat now costs 3 + 4 instead of 2.
That does not mean you cannot board into the transformational sideboard when they know you have it - often it will be correct anyway, because you'll be playing a match that you have playtested (after all you'd not walk in with a deck that sideboards in 12 cards against everyone if you hadn't tested) but they haven't, so they might not know what to do.
Another example of combo with a transformational sideboard would be Cedric Phillips’ Aluren deck:
The sideboard here has some answers to answers - Proctor, Shredder - and then an alternate plan of Natural Order and Progenitus, which is your to-go against aggro. The theory behind that is that aggro will bring in either enchantment removal that can't be countered or something like Ethersworn Canonist, and since they're fast you don't have much time to disrupt them and then combo, so you try to go around it by attacking them from an angle they did not expect, that does not get hated out by the hate they are already bringing. Against control, there is time to disrupt them with Duress and Grip, so you don't need the plan, though you could board it in depending on what they have.
It is also important to note that, on this list, you take out 4 Aluren and a kill condition for 4 Natural Order and 1 Progenitus, which leaves you with, if not the best deck ever, at least a functional one - it's not like you randomly still have Emrakuls, Lion's Eye Diamonds or Donates in your deck after boarding, every card can be played.
To wrap this up, I'll offer one piece of advice that has served me well in the past years. Pay attention, because it is probably the most important thing in the article. Ready?
YOU DO NOT HAVE TO SIDEBOARD A CARD IN JUST BECAUSE IT IS IN YOUR SIDEBOARD.
What this means is basically that your sideboard cards, much like Mistbind Clique, don't have provoke - if you have nothing to side out, then don't bring them in. You might feel like you're wasting your sideboard if you don't side in a removal against aggro, but maybe that removal is there for a different deck, and it is actually worse than anything that you could take out for it. Sometimes I'll see "but I have to bring this in, what do I take out?", but you don't have to bring them in. In fact, there are very few things you have to do - you have to eat, sleep, breathe, drink, play four Cryptic Commands in Faeries and mulligan 1-landers on the play; everything else is optional.
I hope you've enjoyed this, and I hope it helped a little when you find yourself without a clue on how to build a sideboard or on how to sideboard. See you next week, or this weekend if you're going to Dallas!