A lot of the time, these differences of opinion get lost in the way we present ourselves to the public. So Mark and I decided a month or so ago to occasionally do a "Point/Counterpoint" type of column where we air out some of the arguments that haven't yet been conclusively resolved. This week, we're tackling (as I understand it) card drawing on value spells like Divination or Thirst for Knowledge.
The gimmick works as follows: He and I are writing our articles entirely independent of one another. Normally I'd get to enjoy a bit of an unfair advantage, as I could revise my article after reading his Monday column such that I appear all smart and clever by rebutting his points. So I'm pinky-promising to have turned this in before I get a chance to respond directly to whatever he's saying. Then you guys can judge in the forums and via email which one of us makes more sense.
Would you like fries with that? How about a used drive belt off a Yanmar 2GM20 marine diesel engine?
It's really tempting to act, by default, like "options" are a good thing. "Variety" carries positive connotations. "Selection" is a word advertisers use to suggest that your preferences are more likely to be met at a particular supermarket chain. But even leaving aside the vast wealth of happiness research that indicates too much choice leads to decreases rather than increases in satisfaction—think of whether you really need, for example, 64 different kinds of detergent in the laundry aisle, or 350 different channels of cable, including four separate shows in which people chase alligators through the swamp—it's easy to think about how, in our daily lives, we don't go out of our way to give ourselves options we don't need.
I'm sitting at a coffee shop right now, for example, as I write this column. I'm sitting in a very comfortable chair. I bet it's possible to design this chair such that it could store, in a compartment below its seat, up to four Betamax® tapes while preserving all of its existing functionality. Moreover, I just ordered a delicious 16-ounce dirty chai latte. Weirdly, though, every item on the menu is some kind of coffee variant. I'm looking with my own eyes at a spray-bottle of whipped cream, a bottle of honey, and no fewer than eighteen different varieties of syrup—as well as (I can't figure out why) a jar of kosher dill pickles. But none of these things are on the menu. None of these things are presented as options available for consumer purchase.
I wonder why that is?
Imagine the following card:
I have had conversations with a number of Magic players—most of them very competitive—who wish most cards worked this way. They want a bunch of 2/3s for five mana with abilities like ": Target red Penguin becomes green and black until end of turn," which in turn interacts with some card like ", Instant: Target black Penguin gets –1/–2 until end of turn." More opportunities, after all, to increment value! More opportunities to grind out advantage over opponents!
My contention, however, is that such a design wouldn't actually accomplish what they want. In order to have opponents, you have to have players. And nobody would ever play this game.
Design is about elegance. Design is about simplicity. Design is about cards that make sense, that people can understand, that do what they say they do and do it in an appealing way. Platypus should, as an aggressively costed 4/4 for four, be an appealing card. But it's not, because it's got this gunky do-nothing ability that costs a million mana and confuses people. And this is not conjecture. We have done actual market research where we basically present people with two cards:
Overwhelmingly, on basically every axis, people prefer the "worse" 3/2 with no abilities. It's not because they're dumb or that they don't understand the difference between a 3/2 and a 3/3. It's because the ": Lose 10 life" ability is a joke and a lie and an insult; a ploy by a designer to revel in a player's tension, confusion, and discomfort. It's a trap. It's not that people are going to accidentally activate this and lose because of it. It's that we, as human beings, trust designers to create things that are purposive. When they aren't, it feels bad for us—like the hand that promised to guide us is just as lost as we are.
Now, occasionally there will be a deck that can take advantage of the ability to lose 10 life. I realize that. In fact, I just designed a mechanic, fateful hour, that would actually really love to have access to this kind of card. But in Dark Ascension—which Maro himself lead-designed—there was no card like this whatsoever anywhere in the file. That's not accidental. It's not that we never thought of it. It's that printing such a card is just so obviously wrong.
It's interesting to me that he realizes this, but doesn't realize that "Target player draws two cards" is the exact same thing.
Mark talks all the time about his admiration for Apple. Apple has become one of the most successful companies in the world because its entire product development strategy involves not making products like Platypus. It does not present its consumers with superfluous "options" that, 95% of the time, are traps toward behavior that the user doesn't want. Instead, its products do a few things well, and they do those things reliably.
"Draw two cards" is Apple. "Target player draws two cards" is Linux.
I realize, of course, that statement is controversial. I also realize that a lot of people like Linux, and a lot of people think Apple should present its users with more options. I'm not here to say that one or the other is better. But as a designer, I feel like I have a responsibility to the people who use our products, who spend their hard-earned money because they believe we can help them create memories through Magic. I feel like I have made them a promise. And when it comes to delivering on promises, the Apple way of doing things has a pretty good track record.
Walk the Walk
Okay. I realize I've equated ": Lose 10 life" with "Target player draws 2 cards." Why does that parallel hold up?
Glad you asked. Let's break it down.
It's tough to quantify, but most evidence suggests that well over 90% of Magic games are played as duels. Now, I realize there are a lot of multiplayer aficionados out there, so I'm going to go ahead and point out I'm not trivializing multiplayer Magic. A ton of multiplayer gets played, too, but that's because a ton of Magic gets played, period. What that means is that while we make sets like Commander for a certain segment of our audience, we have got to design most of our cards for most of our players. And most of our players experience Magic in the form of one-on-one duels.
What that means is that printing Divination as "Target player draws two cards" gives me a false option. It allows me the privilege of casting it on my opponent, but that option is overwhelmingly just a massive downside for me. It's worse than losing 10 life. It's actually allowing my opponent to cast Ancestral Recall—after all, I'm down a card and my opponent is up two (a net gain of three cards), only this time I had to spend three mana and my opponent didn't spend any.
This produces in players the exact same kind of discomfort that having the option to halve their life total does. It's not so much that they are going to run around doing it all the time, although that will certainly happen (I am thinking, for example, how frequently I've had Sign in Blood cast against me at Prereleases on the second turn). It's that Magic is demonstrably more successful when it minimizes the extent to which its cards are counterintuitive.
I realize there are times, even in duels, where you want to force your opponents to draw cards in order to deck them. I'll deal with that in the next section. But given that you really want to do that only when it'll kill your opponents—otherwise, you're giving them the juice to kill you, which is very likely to happen when your opponents get to cast zero-mana Power Nine cards in Booster Draft—it's still exceedingly rare that such a situation comes up. So I'll frame it another way. Given the following two cards:
...I know which one I'm printing.
Previously, we've been operating under the assumption that (a) you're not about to deck your opponent and (b) you're playing a two-player game. Under those conditions, my contention has been that you don't want to put ponderous and unattractive options on a card just to make yourself feel better. However, even granting Mark that I am about to deck an opponent, or I am in a multiplayer game, I'm not convinced his solution is better for Magic.
At the 2002 World Championships, my boss, Hall of Famer Dave Humpherys, won one of the most memorable Sunday games in history by Deep Analysis-ing and Cephalid Coliseum-ing his opponent to death after having all of his Psychatogs exiled by Lobotomy. The game was also so grueling and intense that he received a slow-play warning in an un-timed match. Situations like this are cool precisely because they are rare. The more we enable the ability to deck your opponent with card-drawing spells, the more game states we create that are miserable and grindy and counterintuitive and (under normal circumstances) lead to unintentional draws.
One of the most painful formats I can remember was Lorwyn-Shadowmoor Constructed, where a prominent strategy for the Five-Color Control mirror match involved decking your opponent one card at a time with Oona's Grace. I realize there are people out there—myself included—who really like this sort of thing. Most people, however, would literally rather do anything else, and my job as a Magic developer is to please my consumers, not to gratify myself. By taking away players' abilities to kill one another with incremental card-drawing spells, we make it so control decks have to actually play finishers that end games in reasonable amounts of time.
This illustrates, by the way, one of the reasons why I'm limiting the conversation to incremental card-draw. I was on the development team for Mirrodin Besieged, where we printed Blue Sun's Zenith—a targeted card-drawing spell. That spell ends games and, as a rare, is less likely to confuse less-experienced players. So—since Constructed testing never demonstrated that it was so powerful that we should have removed the "target player" victory-condition-restriction as a balancing measure—we worded it as "target player." It ends games.
As for multiplayer?
When I cast Divination right now, I draw two cards. I have no options. It powers me through my deck to my lands and spells. Add the ability to target, though, and all of a sudden my value card becomes political. My decision to target myself becomes a decision not to target my teammates, or for that matter anybody else at the table. Maybe someone in a better board position would require me to target them with my card-draw spell, "or else." Maybe I think I need the cards more than my teammate, and my teammate thinks he or she needs the cards more than me, and now we have to have a dumb argument over something that doesn't matter. Again, if I'm casting a bigger spell, I think these conversations are appropriate to have because it's going to substantially affect the narrative axis of the game. But brick-and-mortar cards like Divination aren't designed to contort themselves around these kinds of decision points. They're design to help your deck work just a little bit better, to do their thing with minimal intrusion.
As we make it as easy as possible for the cards that do work to do work, we get to spend our complexity and decision points on cards that really matter.
Now you’ve read both Mark Rosewater’s Pro and Zac Hill’s Con, it’s time to vote. Which of these two options do you support?