Tuesday, February 20, 2018

Lessons from the Exam

Lessons from the Exam
By Jim Harbor

Welcome Artisans, I hope I find you well. The GDS3 has stoked a fire under the custom magic community and I am glad to see it making all of us better designers, regardless of how well we placed. In the spirit of that, I am going to take a look at some of the trickier questions from the GDS3 exam. Instead of looking at just the right answer, I’m going to break down exactly what WOTC was testing; what they see as the signs of a good designer. Hopefully the takeaway from these questions will help us all tune up our design skills.

2. What is the MOST appropriate rarity for this card?
Scrappy Survivor
[CMC: 6]
Creature —
Menace, prowess
Mythic Rare

Question 2 tests the ability to look at the purposes of the different rarities, looking beyond word count. At first glance this card may look like it could be common. It only has two words of text! Nothing too complex there right? This question is making sure you know commons are about more than just being simple.

In Limited, commons show up the most. Any given common will on average show up about twice in an 8-person pod draft. It's possible to see decks with multiples, and it’s close to guaranteed to show up in the games you play. That means you have to think about its play pattern. Big evasive creatures like Serra Angel and Air Elemental are put in uncommon. Their role is to act as a finisher to give the game momentum. If you slam one of these, it’s not unbeatable but they will carry the game away unanswered. They are put at uncommon so they show up in just enough games and at just the right time where they can push the game to its conclusion, but not in a volume high enough games would wrap too quickly.

The interaction between these two keywords is also something you need to check out. Double-scoop French Vanillas do show up at common sometimes, but menace and prowess combine to form a play pattern that is degenerate at common. If you get attacked with this, you are pressured to multi-block because of the menace. If you are lucky you may trade, but any instant speed spell is going to pump the Survivor and risk you getting blown out. A card that easily forms scenarios where the best-case scenario [for your opponent] is getting two-for-one'd, has no business at common.


5. What is the MOST appropriate rarity for this card?
Come Work for Me Temporarily
[CMC: 4]
Gain control of target creature until end of turn. Untap it. It gets +1/+1 and gains haste, hexproof, and trample until end of turn.
Mythic rare

Question 5 also tests to see if designers can think about rarity at a more nuanced level. Looking at current designs, threaten effects at uncommon will maybe add a keyword and/or a stat boost. But two extra keywords on top of a stat boost is taking something uncommon and adding more on top of it; Wouldn't that make it rare?

The answer is no, because hexproof doesn't add much to this card. Your opponent is going to get their card back at the end of the turn anyway, so they are very unlikely to use a removal spell; The extra keyword doesn’t work well with the base effect. This question tests to see if you realize just dumping keywords or effects on lower rarity cards isn't enough to pump them up to high rarities. The context matters.

The Typo version of this question has a different answer, but you still got points if you picked it and it teaches a similar lesson:
Come Work for Me Temporarily
[CMC: 4]
Gain control of target creature until end of turn. Untap it. It gets +1/+1 and gains haste, deathtouch, and trample until end of turn.
Deathtouch and trample interact in a confusing way, WotC outright refuses to mix them on a single card that has them naturally. (The closest we got were the two Garruks whose pluses and ultimates could combine to make a Deathtouch Trampler, but that was pretty small corner case.)

A creature with deathtouch and trample can assign 1 damage to a creature—enough to kill it—and then trample over all the rest of its power to the opponent. So a 5/5 death touch trampler that hits a 4/5 will still send 4 damage to the enemy.

This card wouldn’t be done, but if it were, where would it go? The second lesson of the typo question is that complexity doesn’t scale up with rarity. Rares are more complex than uncommons, which are more complicated than commons, but Mythic Rares are actually simpler than rares.

Mythics are meant to be splashy hyper sexy effects that can blow a player away. If it’s hard to understand what they do, that visceral impact is taken away. This is the reason why Warp World was put at rare in Magic 2010, and similar “crazy wall of text red enchantments” like Sunbird’s Invocation have kept that slot.

This is why Mark Rosewater’s Drive to Work says “If we ever did deathtouch and trample together, it’d be at rare.”


25. Why do green's common creatures tend to be a bit more efficient than white creatures?

White has stronger removal.
White is the enemy of black and red.
Green players love powerful creatures.
Green is the enemy of blue and black.
This is the way it has always been.

Most of the color pie questions on this test are easily answered by just looking them up. But this one asks you to go deeper and explain why the color pie is set up like it is, instead of just regurgitating what Maro’s told us.

Saying “Green players love powerful creatures” is a trap here. Green has the most powerful creatures, which is why Green players like them. It’s mixing the cause and the effect.

White is the color in MtG with the widest breadth of removal. It can remove artifacts and enchantments, and can take down creatures as long as a condition is met. These conditions are often not that big a deal, sometimes you have to “remove the removal” to “escape” or otherwise it will only hit a certain subcategory of a creature (i.e. tapped, attacking, blocking, or a certain type or color). The power to kill any creature but compensate your enemy for it is the most powerful version of this, giving us mistakes like Path to Exile and Swords to Plowshares. These “downsides” often make the removal cost less, making them more efficient. This, coupled with the ability to dip a toe into counterspells make White the color in MtG with the single best suite of answers.

Because of this, it has to be weakened in other areas to compensate. White is the worst at card draw effects because having a deck full of answers combined with the ability to draw into those answers would make the color very overpowered. (Old school Blue players are familiar with this.)

It’s natural to assume this logic would extend to creatures. While White and Green are both the creature colors, White doesn't have the most efficient ones because a single color shouldn't have both the best threats and the best answers.

This question rests on your ability to understand the reasoning behind the color pie to figure out a fact that makes sense, even though it hasn’t been told to us directly.


(Today's Guest Article comes courtesy of  Jim Harbor. If you would like to contribute a Community Spotlight article, read Write for Goblin Artisans, then send a brief pitch of your idea to zefferal on gmail.)


  1. Thanks Jim! This was helpful.

    Maro's Twitter reports that Top 8 emails are out and also that any who didn't make it are free to share their designs. I was one of those who didn't make it, so I'm probably going to share my designs and self-critique on my own blog shortly.

    If we get enough people to check in like this, maybe we can figure out which Artisan(s) made the top 8 by process of elimination? Exciting times!

    1. That sounds a lot like cheating. The top 8 arent allowed to say so so I frel trying to suss it out breals the spirit.

      Anyway my fatal flaw was not checking my Planeswalkers for wordiness. I checked ALL THE OTHER CARDS but forgot to check my walkers, leading to an 11 line walker.

      Also I screwed up a targeting/templating thing ams accidentally made a card that doesnt work.

    2. How many words is that 11-liner? That sounds like a whole lot. I'm curious to see the design.

    3. It was a walker based around the idea of storing counters and redistrubting them

      Grassa the Grafter (mythic rare)
      Legendary Planeswalker -Grassa
      +1: Remove any number of counters from among other permanents you control, then put that many
      loyalty counters on CARDNAME.
      -X: Distribute X +1/+1 counters among any number of target creatures you control or distribute X loyalty
      counters among any number of other target planeswalkers you control.
      -9: You get an emblem with “If one or more counters would be put on a creature or planeswalker you
      control, that many plus two of each of those kinds of counters are put on that permanent instead.

      If I had checked it for wordiness I could have just has the minus only work for creatures and/or the ult say "permanents" or the like.

  2. I really appreciate the point of understanding why the color pie works the way it does, which is a lot harder than memorizing which ability is primary/secondary/tertiary where. I remember Ken Nagle making the point in GDS2 that if you have every color huge common creatures, it was no longer clear why you'd bother with green and its ramp spells instead of black and its removal.

  3. Yeah, I often find the rarity questions hardest, and think they're one that shows off good judgement best. And making colour pie judgements about things which are philosophically clear but don't have much precedent is likewise a good test.

    That said, I liked some of these questions a lot more than others. The first one seemed pretty clear to me. Just from size alone, it's probably not common, but not everyone will know that immediately. And a 5/5 that grows and evades is a lot less common. And I wasn't *sure* it wasn't rare, but it didn't seem like it had to be so it probably wasn't.

    The second one, I was really unsure, because Act of Treason plus two other things is usually going to be uncommon. Act of treason itself is on the verge. But the specific abilities are ones that... don't add much to the effect. How often is the opponent going to kill their own creature here? But not be able to kill it in response to the spell? How often are they going to chump block when it's only attacking for one turn? This is something like "Act of Treason + three damage to opponent". I think I should have said, "act of treason is already on the verge, anything else makes it definitely uncommon". Or "random irrelevant verbiage pushes it up in rarity." But I kept looking at the card as if it didn't have those, because it seemed like they weren't adding much to the card, they might as well be removed. (I think I put common in the end but I'm not sure.)

    Whereas, I should have said, what are they testing here? The fact they changed the question means they thought it was important it had SOME abilities, but not which they were, I would have viewed the question more as you put it.

    (Indeed, at first I thought they were asking, "do you know that trample+deathtouch is definitely rare", but apparently that was just an oops)

    Likewise for the last question, I could see what they were asking, like, "starting from where we are now, why couldn't they move big creatures into W instead of G". But actually, the reason IS that they were there all along. If in alpha, white had big creatures and removal, would they have definitely moved creatures into green? Maybe they'd have moved removal into green instead. Who knows? The colour pie has shaped itself *around* green being the creature colour. You could ask the same question in reverse, "why couldn't green have removal like white? because it has the best creatures". I think it's a good question, but as actually written it wasn't clear enough.

    PS. Congratulations times a million and good luck to anyone who got through to final eight, if anyone did. And many many commiserations to everyone else.