MTG Sideboarding Guide: How to Win Games 2 and 3

Most players build a 75-card deck and only ever think about 60 of them. They sleeve up a maindeck, register a sideboard they barely tested, and then in game 2 they pull out whatever "feels" right against the opponent. That's how you lose a match you should have won. The sideboard is where best-of-three Magic is actually decided, and sideboarding well is a skill you can learn — most people just never bother.

Here's the thing about a competitive match: you only play game 1 blind. After that, both players know what they're up against, and the player who adjusts better usually takes the set. Your maindeck is a guess about the whole field; your sideboarded deck is a precise answer to the exact opponent in front of you. This guide covers how sideboards work, how to build a 15, the actual decision of what comes in and what goes out, the staple sideboard cards by archetype, and the play/draw and mulligan wrinkles that change how you board.

How sideboards actually work

The rules are simple and worth getting exactly right, because people lose games to deck-construction errors that a judge will catch.

• A sideboard is up to 15 cards, separate from your minimum-60-card maindeck. You don't have to play all 15, but you can't go over.
• You only board between games, never mid-game. Game 1 is always your registered maindeck (your "main 60").
• You can swap any number of cards, one-for-one, between your maindeck and sideboard between games. Take out three, put in three. The deck must return to its registered 60-card minimum before the next game.
• You must "de-sideboard" back to your maindeck between matches (or as the rules require between rounds), so your registered decklist is the baseline you always return to.

That one-for-one constraint is the whole puzzle. You don't get to add cards — every card you bring in forces a card out. So sideboarding isn't "what good cards do I have against this deck," it's "which of my maindeck cards are weakest in this matchup, and is my sideboard card actually better than the thing I'm cutting?" If the answer is no, you don't board it in. Lots of players board in five hate cards and quietly gut their own game plan in the process.

Building a 15-card sideboard

A sideboard isn't a junk drawer of cards that didn't make the maindeck. It's a set of targeted answers, and every slot should have a job. Build it against the metagame you expect, not against every theoretical deck.

Think in terms of what your maindeck is bad against. Your 60 is tuned to beat the field on average; your 15 patches the matchups where the average plan falls short:

• Dedicated hate for the matchups your maindeck genuinely loses to. If aggro runs you over, you want cheap removal and lifegain. If combo is faster than you, you want disruption or counterspells.
• Flex threats and answers that improve grindy mirrors — extra card advantage, a haymaker the opponent can't easily answer, a planeswalker.
• Catch-all answers that hit multiple decks at once. A single card that's good against three archetypes is worth more than a card that's great against one fringe deck.

A common, healthy split for a tuned constructed sideboard is roughly: a few cards for the aggro matchups, a few for control/grind, a few for combo or graveyard decks, and a couple of flex slots for whatever's spiking your local meta. Avoid the trap of dedicating a card to a deck that's 2% of the field — those slots are nearly dead weight. Every card in your 15 should improve a matchup you'll actually play.

The test for a sideboard card: "How many games does this realistically win me, and against how much of the field?" A four-of hate card against a deck nobody plays is worse than a flexible two-of that's solid against half the room.

What to bring in and what to cut

This is the part people get wrong. Bringing the right cards in is the easy half. Knowing what to take out is where matches are won and lost.

Start by identifying your dead cards in the matchup — the maindeck cards that do nothing or close to nothing against this specific opponent. Those are your cuts:

• Removal with no targets. Against a controlling, near-creatureless deck, your creature-removal spells are blanks. Cut them for cards that pressure or disrupt.
• Slow, durdly value cards against aggro. Your three-mana "draw a card" engine doesn't matter if you're dead on turn five. Cut it for cheap interaction.
• Narrow cards that the matchup doesn't reward. A card that only shines in long games is dead against the fast matchup, and vice versa.

Then bring in the answers that match. The key discipline: don't over-sideboard. It is almost always wrong to swap out 7+ cards. Your maindeck is a coherent, tested machine; gut too much of it and you create a clunky 60 with a broken curve and inconsistent draws that's worse than what you started with. Most good sideboard plans move 3 to 6 cards. If you find yourself wanting to change ten, your deck is probably wrong for the meta, not your sideboarding.

Two rules that keep you honest:

1. Keep your curve intact. If you cut three two-drops and add three four-drops, your deck now stumbles on early turns. Side in and out at similar mana values when you can.
2. Don't cut your win condition. Boarding into pure interaction with no way to close the game is how control players lose to decking or to a topdeck war they had no plan to win. Keep enough threats to actually end the game.

Common sideboard cards by archetype

These are the workhorses you'll see in real sideboards across Modern, Pioneer, and Legacy. Names and exact legality shift with bans, so the categories matter more than any single card, but here are reliable, currently-legal examples by job.

A few of these are MVPs worth calling out. Rest in Peace and Leyline of the Void are the gold standard for graveyard hate — Leyline can even start in your opening hand for free, locking out a reanimator deck before it does anything. Veil of Summer is one of the most efficient protective cards ever printed (and so good it's banned in Pioneer while legal in Modern and Legacy). Mystical Dispute is a near-free counter in the control mirror. Abrade is a flexible answer that kills a creature or an artifact, so it covers two jobs in one slot.

The lesson here isn't to run all of these. It's to identify which jobs your sideboard needs based on your local meta, then pick the most flexible card that does each job. Two cards that each cover two matchups beat four cards that each cover one.

Writing a sideboard plan

Top players don't improvise their sideboarding — they write it down before the event. A sideboard guide is just a short note for each expected matchup: what comes in, what goes out, and any play/draw tweak. Building one forces you to actually think through the cuts instead of fumbling them at the table with a clock running.

A simple format for each matchup:

• Vs [Archetype]:
• IN: the specific cards and counts
• OUT: the specific cards and counts (must match the IN count)
• Notes: on the play vs on the draw differences, what to play around, your win plan

Write these for the five or six decks you expect to face most. The act of writing them is half the value — it surfaces problems like "I have nothing to bring in against this deck" or "my cuts don't match my adds" before they cost you a match. Test the plans in practice games, not for the first time on camera. A plan you've never tried is a guess with extra steps.

Play, draw, and the mulligan interplay

Sideboarding doesn't happen in a vacuum — it interacts with whether you're on the play or the draw, and with how you mulligan.

• On the play, you're the aggressor. Tempo matters more, you have less time to be reactive, and you generally want your proactive, cheap cards. Some plans cut a card on the play to lower the curve.
• On the draw, you have the extra card and usually a bit more time. You can afford a slightly more reactive configuration — an extra removal spell or counter — because you're a card up and trying to grind.
• Your mulligan decisions shift after boarding. Once you bring in dedicated hate (say, graveyard hate against reanimator), a hand that has that hate is more keepable, and a slow hand that doesn't is more shippable. Post-board games are often won by mulliganing aggressively toward your key answer. If you want the deeper framework on keeps and the London mulligan, see our MTG mulligan guide.

The throughline: your sideboard plan should bake in a play/draw note. "On the draw, +1 removal, -1 threat" is a one-line adjustment that wins close games. And remember the choice itself — the winner of game 1 decides play or draw for game 2 (almost always choosing the play), and the game-3 choice goes to the game-2 loser. Knowing you'll likely be on the draw in game 2 if you won game 1 lets you plan your boarding around it.

The mistakes that lose games 2 and 3

The same handful of errors cost players matches over and over:

• Over-sideboarding. Swapping 8+ cards and turning a tuned deck into a pile. Move 3 to 6 cards in most matchups.
• Boarding to a fantasy opponent. Bringing in hate based on what you assume they're playing instead of what you saw in game 1. Side to the deck in front of you, not the archetype's stereotype.
• Cutting the win condition. Loading up on answers until you can't actually close the game. Keep enough threats to win before you deck out.
• Wrecking your own curve. Cutting all your cheap plays for expensive hate and then stumbling on the early turns. Match mana values when you cut and add.
• Forgetting to de-sideboard. Starting the next match (or game 1 of the next round) with last round's sideboarded deck. Always reset to your registered 60.
• Never testing the plan. Sideboarding a card combination for the first time live. Test post-board configurations in practice — the post-board games are the majority of games you'll play in a best-of-three.

Sideboarding is the highest-leverage skill most players never practice. Game 1 you play blind; games 2 and 3 you get to be precise — and over a tournament, you play far more post-board games than game 1s. Treat the 15 like part of your deck, write your plans, test them, and you'll start winning the matches you used to split.

Quick Action Checklist

Tightening your sideboard game? Run through this:

• Build your 15 against the meta you expect — every slot needs a job, no dead-weight cards for fringe decks
• Favor flexible cards that improve multiple matchups over narrow single-matchup hate
• Identify your dead maindeck cards in each matchup first — those are your cuts
• Move 3 to 6 cards in most matchups; if you want to change 10, your deck is wrong for the meta
• Keep your curve intact — swap at similar mana values, don't gut your early game
• Never cut your win condition into pure interaction with no way to close
• Write a one-line sideboard plan (IN / OUT / play-draw note) for your top 5-6 matchups and test it
• Side to the deck you actually saw in game 1, not to the archetype's stereotype
• Reset to your registered 60 between matches — don't forget to de-sideboard