People don’t pay anywhere near enough attention to sideboarding in any format. There are nowhere near enough articles on the subject, and articles about specific decks don’t address it sufficiently. I’m not sure that I’m qualified to write the defining piece on the subject, but more needs to be said.
This article was primarily inspired by Evan Erwin coverage of the PTQ finals match between James Rogers and Sean McKeown. That coverage featured the line, “Sean takes out Ponder, Meloku the Clouded Mirror, an Ancestral Vision, and Vendilion Clique for three Threads of Disloyalty and one more Engineered Explosives.” That looked pretty weird to me. I’m not trying to hate on Sean here, those are reasonable cards to cut, and I’m not entirely sure it’s wrong, but it doesn’t look right to me, and it at least looks like a fairly common way of sideboarding that is generally indicative of not really even trying to sideboard correctly.
The problem here goes much deeper than “Sean didn’t sideboard right.” I believe the problem is with the way people test. Preparing for the top 8 of U.S. Nationals is when it really all clicked for me. I knew exactly what I was playing and what I was playing against, and I had a night to prepare for the matchup, and I’d never played against the deck before, as it was a rogue deck. The matchup looked pretty bad, but there’s always a chance, so we set about trying to figure out how to sideboard. I talked to everyone I could find to ask their opinions, sat down and played a bunch of sideboarded games with different configurations of both decks, and then talked to everyone else I could find. Throughout our testing we completely changed how we were siding both decks because our initial guesses were just wrong. Meanwhile, Shaheen wasn’t testing, and in game 2 when he played a Rune Snag, I knew that I had suddenly gone from unfavored to favored, as I was winning in our testing before we figured out that he needed to cut Rune Snag.
This was when I came to the conclusion that almost all testing should be done post sideboard. You learn so much more, and you get more relevant data. Unfortunately, it seems extremely difficult to convince people to test this way. This is what led to my shifting primarily to testing on Magic Online, where I can be sure that I’ll always play matches.
The most common excuses for not testing sideboarded games that I hear are “I’m not sure how I would sideboard,” and “I’d rather just see how my deck plays game 1.” The first is the exact reason you should be testing sideboarded games. The second is reasonable if you’re playing a deck you’re not familiar with, but even then I don’t think it outweighs the advantage of gaining experience in sideboarding.
Enough ranting about testing methods, let’s get to specifics.
In games 2 and 3 you’re presenting a deck. What I mean is that the stack of cards you put in front of your opponent should look like a deck. Personally, I can’t think of any decks I’ve seen with 3 Ancestral Visions and 3 Ponders, and this is the problem with the trimming approach. I feel that, if people really tested post sideboard games, they would find cards that they didn’t want in their decks, and cut all of them. Specifically, I think there’s a reasonable chance that Sean should have just been siding out all of his Vendilion Cliques.
More generally, you should think about sideboarding as an opportunity to rebuild your deck. You should build it to look like the deck you would build if you were building a deck specifically to beat your opponent. This might work best as a thought exercise on what to side out, think about it like this: If you were to build a different main deck version of your deck designed to perform better against the deck you’re playing against, what card wouldn’t you include?
Another thing people don’t properly consider is what a sideboard plan does to their mana curve. Everyone pays a lot of attention to exactly how many cards of various casting costs they have in their deck when building, but when sideboarding, that often all gets ignored. In general, this is fine, as you’re bringing in cards that are powerful enough that it doesn’t matter, and taking out cards that are bad enough that it doesn’t matter, but it’s something to pay attention to when building a sideboard and deciding which cards you want access to. Think not only about which cards are best against the deck you’re playing against, but which of the cards that are good against them fit best in your deck, or possibly more specifically, which best fill the slot that will be opening up when you cut the card that is worst against them.
When building a 75-card deck, it is also important to consider when you’ll want to change roles in a matchup. I’m not talking about a full transformation, but I am talking about a significant shift in curve and game plan. The best recent example I have of this is my approach to building Reveillark decks. When I build a Reveillark deck, I want to be able to cut my Reveillarks, lower my curve, and try to be aggressive against Faeries. I don’t think a Reveillark-based game plan can beat Faeries, so I want to be able to abandon that plan and focus on doing something useful with my deck. Often this means sideboarding 2/2s for two mana.
When building a deck it’s important to consider how many cards you want to take out in the most important matchups, and to make sure you have the right number of cards to bring in for each of those matchups.
For example, my current sideboard plans with Faeries in Standard:
Creatures (12)
Planeswalkers (2)
Lands (25)
Spells (21)
There are a lot of approaches to the mirror match, but my goal is to lower my curve as much as possible to maximize my chances of winning the tempo war and to take full advantage of Jace. As such, my sideboard plan is -4 Mistbind Clique, -1 Vendilion Clique, -1 Agony Warp, +1 Peppersmoke, +2 Thoughtseize, +1 Jace Beleren, +1 Countersquall, +1 Negate on the play, Remove Soul on the draw (Negate is better if and only if it can counter turn 2 Bitterblossom). The reason for the 1 Countersquall 1 Negate split is that I think Countersquall is better against Faeries and Negate is better against Five-Color Control (based on how important the 2 life is), and I want the 1 that is always going to be in my deck against Faeries to be Countersquall.
Against Five-Color Control, I want to be able to cut all of my Peppersmokes, Scion of Oonas, and Agony Warps. That means I need 9 cards to bring in. Fortunately, 2 Glen Elendra Archmage, 1 Countersquall, 1 Negate, 1 Flashfreeze, 1 Remove Soul, 1 Jace Beleren, and 2 Thoughtseize adds up to exactly 9. Because I’m cutting cheaper spells, I don’t want to bring in more Glen Elandra Archmages so I settle for less powerful cards that are still good against them in the form of two-mana counterspells.
Against R/W Lark (Boat Brew) I fear Volcanic Fallout and Mogg Fanatic, so I want to cut my 3 mana creatures. I also fear Banefire, so I want to leave Vendilion Clique in anyway. I don’t like Thoughtseize because my plan is to take advantage of their high curve and try to tempo/howling mine them out with Jace Beleren. Spot removal is relatively ineffective because of the nature of their threats. The plan is -2 Thoughtseize, -2 Agony Warp, -3 Scion of Oona, +1 Remove Soul, +1 Flashfreeze, +3 Infest, +1 Jace Beleren, +1 Peppersmoke. I don’t want to bring in Sower of Temptation because any one of their creatures isn’t that good and they can punish me too much for tapping 4 mana on my main phase. Also, they have Path to Exile and I’m taking out my Scions, so Sower is just too risky. Jace is one of my best cards against them, because they don’t have enough pressure to kill it quickly if you use the +2 ability and they can’t do much with the extra card, since they’re likely to be using all of their mana for most of the game anyway. Furthermore, it helps you assemble the chain of Mistbind Cliques that they can’t beat. Just going through this has made me wonder if I should have 1 less Glen Elendra Archmage in the sideboard and one more Vendilion Clique, just so that I could bring it in instead of the 1 Agony Warp that’s staying in my deck as it would still be good against Five-Color Control and it would significantly help against Banefire in this matchup. This is why it’s important to think through sideboard plans in advance. It lets you see when you can afford to dedicate a slot only to 1 deck (like Glen Elendra Archmage) and when another deck needs one more card to come in and you have to try to find a card that can do double duty.
Against Kithkin, Scion of Oona is less likely to get killed, so you can keep that and bring in Sower of Temptation. In general, they’re faster, so it’s more important to have removal than countermagic. As a result the sideboard plan is more like -2 Thoughtseize, -3 Broken Ambitions, +2 Sower of Temptation, +3 Infest. Note that I’m not cutting Jace here. I really like Jace and try to cut it as rarely as possible. When I’m bringing in more removal I just want to trade to protect it and use the card advantage to win the late game.
Against B/W Tokens, Thoughtseize is better because they won’t be killing you with Banefire, their deck relies more heavily on specific interactions, and they have Bitterblossom which you would really like to deal with. Terror is terrible as it doesn’t kill most of their creatures, and spot removal is still relatively ineffective. LSV said he fears Sower of Temptation from the B/W side of this, but I just can’t see bringing it in. Spot removal isn’t good enough and they can punish you too much; also, they could have up to 8 spot removal spells for your Sower. The plan here is -2 Terror, -1 Vendilion Clique, +3 Infest. On the draw I would also cut 2 Broken Ambitions for 2 Thoughtseize.
Against midrange decks like G/B Elves, Doran, and Bant, the sideboard is something like -2 Jace, -2 Thoughtseize , +1 Remove Soul, +1 Flashfreeze, +2 Sower of Temptation. This can change a lot depending on their deck, as Peppersmoke, Infest, and Sower of Temptation change value a lot depending on their exact builds, but that’s the basic idea. These decks hit hard enough that Jace can’t protect himself by using his +2 ability, so this is the one place where I don’t particularly like him.
On all of this, it should be noted that none of these plans are set in stone. Manuel recently said that he doesn’t like to think about what he’s doing in advance because he wants to be able to change his plan to combat a specific opponent’s deck and play style. There is a lot of truth to that, and it’s important to know when to change up, but where most players are at, they would do much better to think more about sideboarding before they enter an event.
Until next week…
Sam