Brainstorm is arguably the best card in Legacy. It’s probably the most played card in the format, and almost every blue deck plays the full 4. In fact, several decks that are blue probably wouldn’t be if Brainstorm didn’t exist.

Brainstorm is also one of the most complicated cards to play. It costs only 1 mana and it’s an instant, which means that you have a lot of choices for when to play it. Then, when you do play it, you have to select two cards from however many you have, and you have to select an order to put them back. Finally, you can often choose to draw the cards you put back or shuffle with a fetchland.

Simply put, there are so many things going on with Brainstorm that it’s rare for two players to even resolve it in the same way. Give a group of 10 people a Plains, a Wasteland, and a Mother of Runes, and you have two variations of how that plays out. Some people Wasteland first, and some play Mother of Runes first. Give the same group an Island, a fetchland, and a Brainstorm, and you will probably have ten different unique plays.

In this article, I’m going to talk about what Brainstorm can do, and how to play it.

The most important thing to keep in mind when considering a Brainstorm is that it is actually three cards in one, each completely different from the other. Every time you cast Brainstorm, you should be looking to do at least one of those things, and you should know which ones before you cast it. If you cannot identify what your purpose is when casting Brainstorm, you’re much more likely to mess it up. Here are the three versions of Brainstorm that I use to help guide my plays:

1) The first version of Brainstorm is a filtering spell, like a Thirst for Knowledge. It allows you to get rid of the cards that are bad in a matchup or situation, and replace them with new cards. This is the preferred mode in reactive decks.

2) The second version of the card is a digging spell, like Ponder or Preordain—you draw three cards deep to find what you need. This is usually how Brainstorm is used in combo, and often how it’s used in proactive decks in general (though those use a combination).

3) The third version is a “put cards specifically on top of your library” spell. This is used when you have cards you want to have access to soon, but they can’t be in your hand right now, either because you need to draw them (Terminus) or because you’re dodging a discard spell.

The Filtering Spell

This is the “classic Brainstorm”—drawing three new cards and replacing two bad ones for the situation, then shuffling your deck.

As a general rule, the more reactive your deck/hand is, the more you want to make sure that you can filter your cards away, which means that waiting until you have a fetchland (which, in turn, means not Brainstorming on turn 1). A deck like Miracles has tons of reactive cards (Swords to Plowshares and Terminus to name a few), and it might not want them in certain matchups. In those matchups, drawing those cards again is horrible, so you really want to wait until you can shuffle both away. The same can be true for any deck that has cards like Fatal Push, Abrupt Decay, or Spell Pierce.

Imagine that you’re playing Miracles against an unknown opponent. You’re on the play and your hand is:

Imagine that you keep this hand, and you lead with Tundra. Your opponent goes Underground Sea into Ponder, and shuffles. At this point, you should not cast your Brainstorm, because you don’t have a reason to. You aren’t looking to get rid of any cards yet because you simply don’t know if they are good—your opponent could be playing Storm, or they could be playing Delver, or anything in between, like B/U/G or 4-color Stoneblade. If it turns out that they are playing Storm, then you want to shuffle those Swords away, but you can’t get rid of both if you cast it turn 1. You also don’t know what you need—do you want a Force of Will? How about a Terminus?

In this spot, given that there’s no reason for you to Brainstorm, you probably shouldn’t. I would just untap and draw my card.

The filtering mode is also the preferred one when you have a land-heavy hand in a deck that operates with few lands, like Delver. For those decks, each land past a certain number is effectively a dead draw, and you want to shuffle them away.

Imagine that you’re playing Delver against Reanimator, you’re on the play, and your hand is:

I like keeping this hand, but I would do so assuming that I’m not casting a Brainstorm on turn 1. If my opponent’s turn 1 is “land, go,” I’m going to untap and draw. I am already flooded, and I am likely to want to get rid of multiple lands, which I can only do if I can follow the Brainstorm with a fetchland activation immediately. Because of that, I’ll wait until the main phase of turn 2, even if that costs me a mana in the process.

The Digging Spell

Usually, combo and more aggressive decks want to use Brainstorm proactively. For those decks, it’s not as relevant that you shuffle two cards back as it is that you use your mana efficiently. They often have fewer dead cards anyway, and if the game is over, they don’t care about how many dead cards they have in hand. If you have mana to play things on your turn (i.e., it’s not turn 1), then you usually just want to cast the digging Brainstorm main phase. Most people wait until the end of their opponent’s turn, but that doesn’t make much sense when you can draw into something to play sorcery speed.

Imagine that you’re playing Sneak and Show, and this is your hand after a mulligan and a scry to the bottom:

You lead with Island, and your opponent plays Underground Sea and Ponder-shuffle. In this spot, I think you should Brainstorm at end of the turn. You’re looking for a big creature, and, if you find one, you can get it in play turn 2. If you wait for that Scalding Tarn to come online, you could be giving your opponent the chance to either kill you or disrupt you. In this spot, I think it’s better to just cast Brainstorm.

Now, imagine that your hand is the same, but with Island instead of Ancient Tomb:

In this spot, I’d rather wait a turn to dig one card deeper, since even if I find a creature immediately I cannot cast the Show and Tell unless I also find a double land or a Petal, which is going to be unlikely. I would untap, draw, and then main phase I would cast Brainstorm (so that I can also cast a Ponder if I find one).

Now, imagine you’re playing True-Name Bug. You lead with Tropical Island and Noble Hierarch. Your opponent leads with Polluted Delta into Underground Sea and Ponder. You untap, draw, and your hand is:

Here, you don’t have a fetchland to shuffle things away, and you have no clue if that Force of Will and Fatal Push are going to be good, since you don’t know what your opponent is playing. Should you cast Brainstorm? I would, because I think you’re in digging mode here. You don’t know what your opponent is playing, but your goal with this Brainstorm is not to get rid of that Fatal Push—it’s to advance your own game plan. You need to find a land to play that Jace next turn, and by Brainstorming main phase you can hit Noble Hierarch or Deathrite Shaman in addition to it. You can also potentially hit Ponder, Thoughtseize, or Wasteland, all of which will be great for you. In this spot, advancing your own game plan is more important than waiting until you have perfect information or can shuffle the cards away.

The “Leaving Cards on Top” Spell

Discard is prevalent in Legacy, and Brainstorm is one of the best ways to get around it by hiding your more important cards on the top of your library where they can’t be reached. As a result, it’s often correct to pass with a mana up, and if your opponent casts Duress, Thoughtseize or Cabal Therapy, you Brainstorm in response.

Brainstorming in response to a discard spell is easy, but sometimes you have to cast the Brainstorm before you see the discard spell and you have to plan accordingly. Imagine you have this Sneak and Show hand against Storm:

You lead with Island, go. Your opponent has the standard Underground Sea into Ponder. Now what?

I believe you should definitely cast Brainstorm here. You want to find an Ancient Tomb, City of Traitors, or Lotus Petal. If you find any of those, you can win turn 2. But now imagine that you do not find them. Instead, you draw 3 lands. Your hand is then:

You have to put 2 cards back. Which ones?

It might be tempting to put back two lands, draw one, and then shuffle. But at this point, what are you really looking for? You already have everything you need as long as your opponent doesn’t disrupt you. You’d love to have a counterspell for next turn, but that’s going to come too late regardless.

So what I would do is put back Show and Tell first, and then a land. I’d draw the land and leave Show and Tell on top. Then, I’d play Scalding Tarn and pass, and I would not use it. This way, your most important card—the Show and Tell—is protected from discard. If your opponent plays Duress or Cabal Therapy, they’ll flat out whiff. If they see your hand and then play Cabal Therapy, they’ll take Griselbrand, but you still have Emrakul as a backup, which might not be enough, but will be your best chance.

In a spot like this, you have to recognize that the key to winning the game is protecting that Show and Tell, and that’s more important than getting rid of one extra useless card.

Sometimes, when you cast Brainstorm, you’re forced to keep a card on top anyway. In those spots, if you’re playing against a black deck, it’s better to just keep the better card on top. Imagine that you’re playing a Czech Pile mirror, and after you’ve exchanged a lot of resources, you draw a Brainstorm. Your board is:

After thinking for a while, you decide to cast Brainstorm now, rather than waiting until you have another card in hand. You find Scalding Tarn, Jace, the Mind Sculptor, and Underground Sea. Now, you want that Jace, which means that you’re not going to shuffle with Scalding Tarn before you draw it. Since you want both Jace and Scalding Tarn, you should keep the Scalding Tarn in your hand and leave the Jace on top, so that your opponent cannot discard it with Thoughtseize or Hymn to Tourach. You’re not casting it this turn anyway, so it doesn’t matter if it’s in your hand or on top of your library, and on top it’s better protected.

Another use of the “putting spells on the back” is to enable cards like Terminus, Counterbalance, and Predict. If you have those cards in your deck, you should time your Brainstorm so that they work.

For example, imagine that you have this hand against an unknown opponent:

With a hand like this, you should plan for a turn-2 Counterbalance, which means casting a Brainstorm turn 1. You also want to make sure that you have control over the card on top, which means that you don’t want to sacrifice a fetchland after you’ve Brainstormed. You should, therefore, lead with Flooded Strand, and plan on casting a Brainstorm and putting the appropriate mana cost back second from the top (in this case, a 1 or a 2, depending on what the opponent leads with).

That’s what I have for today. I hope you enjoyed it, and see you soon!