Choices in Magic

A while ago, Gavin Verhey posed the following question on Twitter:


Take a moment to think about it. I know this is hardly a decision that you can make in a vacuum (and making it in a vacuum is never going to happen in an actual game), as it’s incredibly context dependent, but pause and decide toward which decision you’d be inclined.

When Gavin posed the question, 79% of the respondents chose “attack,” and only 21% chose “not attack.” Gavin eventually wrote an article about the question, elaborating on what factors could lead to attacking or not attacking.

My answer, however, is different than most. I chose “not attack.”

I want to be the person making the choice instead of presenting the choice to my opponent. In today’s article, I’ll elaborate on why. If you look at the Tweet’s history, you’ll see that most people’s arguments for attacking boiled down to: “if I’m playing a bear it means I’m aggressive, so I should attack and trade damage.” Even if you assume that playing Glory Seeker implies aggression (it does not), I believe this argument is flawed because attacking does not mean trading damage—it means giving the opponent the choice of trading damage.

Let’s analyze the possible outcomes:

A: You trade damage.
B: You trade creatures.
C: Nothing happens (you don’t attack).

If you choose to attack, you’re not opting for “A” and you’re not opting for “B”—you’re giving the opponent the choice between “A” or “B.” Your attack means “we’re either trading damage or we’re trading creatures, but you’re going to choose which one.” If you select door number 3, however, you lock both players into that option. It can only be correct to attack if you believe that the worst option between A and B is still better than option C.

If I’m assigning values, let’s say that option A is worth 5 points to you (you want to trade damage), and option B is worth 1 point (you’d rather not trade creatures because you can get them through later in the game with a card like Magma Chasm or Overrun). Option C, of doing nothing, is worth 2 points. In this case, even though the best-case scenario is the damage trade, even if the average of A + B is higher than the average of C, it’s still likely correct to pass without attacking because you’re not getting the average, you’re getting their choice. It’s a decision between getting a “2” or “your opponent’s choice of 1 or 5” and, spoiler alert, they’re going to choose 1.

You Always Want to Be the One Making the Final Choice

In Magic, being able to make a choice is a huge benefit. Making the final choice is the difference between Fact or Fiction, one of the most powerful cards to ever have been in Standard, and Steam Augury, a card that hardly saw any play.

I remember back when people played Zoo and had the option between playing Kird Ape or Loam Lion in their decks. Most people chose Kird Ape (or something based on their mana base), but some writers would tell you that Loam Lion was the right choice, because people would spend their Deathmarks on it instead of saving them for your Tarmogoyfs and Knights of the Reliquary, which were better targets most of the time. But you aren’t forcing your opponent to kill your Loam Lion, you’re simply giving them the option to do it. Do you want them to kill Loam Lion? Perhaps you do in certain game states, but you certainly don’t want it in all game states. In this spot, you are deciding between giving them a choice between killing it or not, or making the choice for them before the game starts. Making the choice for them is generally going to be better.

When you make the final choice, you get the best of any option—when your opponent makes the final choice, you get the worst. Cards like Charms or Commands might not have any single mode powerful enough to justify playing, but when you get the best possible for each situation, they’re suddenly much better. With punisher cards, it’s the opposite.

If someone asked me “what’s the most consistently overrated card,” I’d have two answers: first, Cranial Extraction, and second, Browbeat. People look at Browbeat and think “5 damage for 2R is good and 3 cards for 2R is good, so this card is a good rate no matter what.” While that technically is true, the fact that your opponent gets the ultimate choice means the card is not even playable.

In the example above, if you attack with your Glory Seeker, they’re making the ultimate choice, not you. Much like Browbeat, you’re going to get the worst of the outcomes.

If You Can’t Make a Choice, Play in a Way that Your Opponent Can’t Either

Sometimes it’s not up to you to decide something, but you can still make sure your opponent can’t decide anything either. If you have no choices, then it’s better to force both of you to make a play that doesn’t benefit you than to let your opponent choose from among alternatives that include that play.

For example, imagine a scenario in which you’re both at 1 life. You have a 10/10 in play and your opponent has a 1/1 deathtouch, and you’re both empty-handed during your main phase. In this scenario, you don’t really have a good choice—if you attack you’re just going to trade, and you certainly don’t want to trade a 10/10 for a 1/1 deathtouch. In this spot, however, you must attack, because your opponent can force the trade if they want to. By attacking, you say “we’re trading.” By not attacking, you say “it’s your choice between trading and not trading,” and they might not want to trade if they draw a removal spell, a creature with convoke, and so on. In this spot (and in similar spots where they have the ability to force you to block), attacking is mandatory because you must take away their ability to choose.

Make Sure You Have the Most Information Possible Before Making a Choice

If you have a choice to make, then make sure you give yourself the most amount of information you possibly can before making it. This means that if you’re casting Ponder and Tarmogoyf in the same turn, you should almost always cast Ponder first, because casting Ponder first might change the rest of your turn, whereas casting Tarmogoyf first never will—if you cast Tarmogoyf first, you have to follow it up with Ponder, but if you cast Ponder first, you don’t have to follow it up with Tarmogoyf .

The biggest offender here by far is the “end of the turn Wasteland.” You see it so much, yet it’s so incredibly wrong. Using a Wasteland at the end of the turn is almost always strictly worse than using it after you draw your card. You’re not going to have the Wasteland mana anyway, so you might as well draw and make sure that this is what you want to do. This is true when you have so much mana that you can still play whatever you draw. Imagine your opponent plays Tireless Tracker and passes the turn, and you have Ultimate Price in your hand and 10 mana in play. If your deck has nothing that costs more than 8 and no draw spells, there’s absolutely no reason to cast that Ultimate Price at the end of the turn. It’s better to wait and draw for your turn because, who knows, maybe you’ll draw a Dead Weight you’d rather cast instead.

Give as Little Important Information Away as Possible Before Your Opponent Makes Their Choice

If you tap out to play a threat and then attack, you’re giving them 2 important pieces of information: one, that you have said threat, and two, that you don’t have anything else. For this reason, you generally attack before casting spells.

Back when Fact or Fiction was Standard legal, people would often just do whatever they were going to do and then cast Fact or Fiction at the end of the turn, which was not always correct—some of the time, you’d want to cast it in response to your opponent playing a spell. Now your opponent wouldn’t know whether you had a Counterspell in hand or not, so they’d have to make their choice without the right context. Imagine a spot in which you have 2 cards and they cast a lethal Fireball, and you respond with Fact or Fiction flipping Counterspell and 4 good cards. It’s possible that your opponent splits them 4-1 (Counterspell vs. the rest), and then you take the 4 cards and cast the Counterspell you already had in your hand. If you simply counter it and cast Fact or Fiction at the end of the turn, the split is almost certainly going to be worse for you because your opponent will be able to make a more informed choice.

There are spots, however, in which you should purposely present your opponent with some information with the goal of expanding the universe of things you can have to give them an even harder choice. If I cast Drownyard Explorers before attacking my 3/3 into your 5/5, then I’m now including Confront the Unknown into the mix of cards I can have here. Before, your opponent had to make a decision including XYZ cards—now they have to include this one as well, so, paradoxically, you give them less information by giving them more.

It’s the same with basic lands. Imagine I have 3 Plains in play, and I play a precombat Island in a format where blue has no tricks and attack. At this point, I’ve given you free information—my second color is blue and I have access to it right now, as well as a fourth land. It’s very likely that I’m just giving you free information to make a decision, so I should not play the Island precombat. If the land I play is a Forest, however, then things change. Yes, it’s true that I’ve given you information about my second color and about the status of my fourth land drop, but I’ve also added a bunch of unknown factors that you have to take into account because green is the color of pump spells.

When you’re forced to let your opponent make a choice, you should do your best to make sure the most important factor for the decision is the one your opponent knows the least about. One of the best ways to play against cards like Fact or Fiction is to bank on information you have and they don’t, much like when you play with Fact or Fiction yourself. I remember a story from when Marcio Carvalho played Shuhei Nakamura in a tournament, and Shuhei cast Brilliant Ultimatum. Shuhei was at 5 life and knew Marcio had Soulfire in his deck, but was otherwise in a comfortable position, and Marcio had just tapped out to play a 5/5. The Ultimatum came up to 4 good cards and a Tidehollow Sculler, and Marcio split them 4-1. Shuhei thought for a long time and took the Sculler, and then Marcio revealed a land in his hand. In this spot, Marcio got the better end of it because he had information Shuhei didn’t (his hand) and he made the choice contingent on that information.

There are, however, some moments in which you don’t mind—or actually prefer—if your opponent makes the choice. They are:

If the Worst Choice Benefits You More than No Choice

Going back to my Glory Seeker example, there are definitely spots in which the worst of A and B is still better than C, and in those spots you should attack—if there’s a 5/2 that you want to play this turn, for example, or if you’re going to cast a Gravedigger in an aggressive deck. In those cases, you can say that trading damage is a 4, trading creatures is a 5, and doing nothing is a 2. Forcing your opponent to choose is going to result in at least a 4, so clearly that has to be better than a 2.

If You Believe You Have Information Your Opponent Does Not

This part is very tricky. Most of the time you have information the opponent doesn’t have, but you ignore that they also have information we don’t have. In my experience, people vastly overestimate the disparity of information between themselves and their opponents—people tend to believe that their plans are top secret and their opponent’s plans are an open book when in reality each player has roughly the same amount of knowledge. You should never forget that you also don’t know your opponent’s hand and deck, and it’s possible that their knowledge trumps your knowledge for that particular situation. In the dark, you should not assume you know more than they do.

There are, however, spots in which you can be pretty sure you know more than your opponent, or that your knowledge is more relevant than the opponent’s. You could, for example, have a 5/2 in your hand that the opponent doesn’t know about. In this spot, they might choose to trade not knowing that this is actually the option you’re looking for. Or you could have a 1/3 that will blank all future attacks from their 2/1, but they might not trade because they don’t know that, so you’ll get in 2 free damage. Or, perhaps, you aren’t actually an aggressive deck but the way your deck behaved in game 1 led your opponent to believe that, so you think they’re going to trade, which is what you want. Just know that all of these situations can be reversed. Maybe they are the ones who have a 5/2 in hand, for example.

Another example is the practice game between Luis and Paul that I’ve already mentioned in other articles, where Paul was tapped out, Luis passed the turn, and then on Paul’s upkeep he attempted to cast Pull from Eternity on Paul’s Ancestral Vision that was about to come off suspend. Paul thought it was weird that Luis was giving him the opportunity to counter it when he could have just dealt with it while Paul was tapped out, so he let it resolve, and sure enough Luis had Imp’s Mischief in his hand. In a spot like this, Luis had the option of giving Paul no choice (casting it main phase) and giving Paul a choice (casting it on his upkeep), and Luis decided to give him a choice because he had information that Paul didn’t (it didn’t work, however, because Paul realized Luis was giving him a free choice for no reason). As a general rule, if your opponent is competent and gives you a choice that they didn’t have to, you should play as if they want something other than the forced play, and therefore you should just do what they could have done but didn’t.

If You Believe They Might Make the Wrong Choice

In general, people tend to underrate their opponents. They think everyone else is “not as good as they are” so they don’t mind giving those players more choices. Most of the time that is wrong, but sometimes it’s correct. If I think my opponent is not going to attack with their 1/1 deathtouch into my 10/10, for example, then suddenly not attacking becomes more appealing for me. If I think they are going to Deathmark my Loam Lion every time, then I probably want Loam Lion over Kird Ape. If I think they’ll take the damage every time, then I can treat an attack as “trading damage” as opposed to “the worst option between the two.”


  • Try to be the person making the final choice. Cards that let you do that (such as Fact or Fiction or Charms) are usually good—cards that let your opponent do that (Steam Augury, Browbeat) aren’t.
  • If you attack, your opponent has the choice of blocking or taking the damage—unless it’s a scenario in which both are better than staying put, you should not attack.
  • If you’re making a choice, be as informed as you possibly can before making it.
  • If they are making a choice, try to give them as little information as possible.
  • It’s usually better to make a play that’s bad for you than to give your opponent options where one of the options is that same play.
  • Exceptions can be made when you believe you have more information than they do or when you think they’re going to mess up, but in general people tend to overestimate how much more information they have in relation to their opponent.



Scroll to Top