Wednesday, May 16, 2018

Between the Lines of GDS3 - The Design Test

I'm going to examine some of the comments the Wizards team reviewing the GDS made. Today, the Design Test. (I'm obviously biased, discussing more of my own. Feel free to discuss anything in the comments.)

from Ari's test:

1—Steamfist Enforcer
Erik: This card is narrow and would only exist in an artifact-heavy set. However, the request was to design to an average Standard-legal set. So this is not a good fit for the given criteria.
Mark: We asked you to design for an unspecified Standard-legal set and you made numerous cards that could only work in very narrow circumstances.
This comes up a few times. The instructions were to design for a Standard-legal set, not an average Standard-legal set. There have been at least 12 artifact-heavy sets printed for Standard. Also, the idea of an "average Standard-legal set" is nonsense; Even the core sets have a unique mechanical identity to them. (I will say that a set being both heavy in artifacts and hybrid seems unlikely, as both ease color requirements significantly, though hybrid artifacts would work.)

2—The Raven Lord
Eli: Be careful with "as long as [condition], this has [bonus]" effects; if the condition can turn off in the same turn it turns on, losing the bonus can cause creatures to die. This means that either the condition must last for the rest of the turn (as you've done here), the bonus must not protect creatures from death, or we need to rejigger the ability.
It's odd to specify this here, as the submission was already mindful of this.

7—Taiga Eruption

Both Erik and Mark call this out for not being a creature, but fail to recognize that you couldn't print this as a creature card or a land card because it both has land types and a mana cost.

Mark points out "we tend to avoid putting haste on creatures with flash … as the two abilities are non-synergistic" but they did on Samut, Voice of Dissent and it makes sense there because she has an activated ability, which Taiga Eruption also has.

from Chris's test:

2—Keeper, Underworld Warden
Ethan: The rate on this card seems very strong to me; it can exile two permanents or four cards from hand for only four mana.
I can only conclude that Ethan missed the death trigger that returns all the exiled cards to their owners.

Melissa: This can be a mono-green card, and is not much different than Prey Upon. The added flexibility here is cool, but still not black.
The other three judges disagreed with Melissa here. The judges disagreeing is something we'll see a lot of. It's a good thing that not everyone in R&D agrees about everything, and it's common for a designer to get conflicting feedback on their work, but it's awkward in a contest because the designer is caught in the cross-fire.

from Jay's test:

1—Recall with Purpose

Eli offers the modal solution which is clearly best.

2—Echidna's Two-Headed Pup

Melissa points out this is one mana too cheap and she's right. I didn't expect to lose points based on our ability to cost cards with minimal playtesting. (I don't actually know if I lost 'points,' but it's a fair conclusion.)

I was relieved Ethan and Mark felt this was distinct from Two-Headed Dragon because I don't remember having looked that up before submitting.

3—Third Degree

It was a relief to me that the team liked Third Degree because many of the community members hated it. I was happy to find an elegant combination of red's "deal 2 to a creature and its controller" and blacks" "deal 2 damage and gain 2 life." {1}{B}{R} is absolutely more appropriate for this common.

4—Cosmic Revelation
Eli: I can't tell if you want to lock in the target's size right away or if you want it to update throughout the turn. Words exist for either way, but how should it be tested? One trick some designers use is to put in plain-English reminder text answering the question until the editors change the words or I change the rules.
I'd probably lock in the numbers for simplicity. I was grateful for Eli's invitation to clarify intent in the future, but would come to be disappointed.

None of the judges noticed that drawing a card (and doing it first) eliminates the math and confusion a player would suffer from when they plan this spell out in their head. Your hand size not changing between thinking about casting this and it resolving is critical.

5—Lure of Treason

The judges called out how likely this is to end the game and how that's not fun and they're 100% right. I originally had it at four mana, if you can believe that. What cost/rarity do you think this would work at? I don't think it's novel enough to be a six-mana mythic. Seven-mana rare? Should it just not exist? I can imagine a variant where you have to pay {2}{R} to get the Act of Treason effect and {1}{G}{G} to get the Alluring Scent effect, perhaps on a fuse-split. I'm also okay with just burning this idea.

6—Illusive Vagabond
Melissa: I understand what you're trying to do, but I dislike this card. I would avoid forced sacrifice. Even though I'm gaining value, sometimes I don't want to sacrifice my creature. This body is sized in a way where he will not connect very often. I do like the idea here.
Yup. Adding a 'may' is the one change I'd make to this card. Good call.

7—Bucket List

Though I was aware of the Standard requirement when I began, I honestly forgot while submitting and was later nervous this one card would get me eliminated for failing to follow the rules. It did occur to me that one could argue this as a black-bordered mechanic, even though it's closest kin have been silver- so far.

I was delighted to find it well-liked, of course, but surprised I didn't even get the Standard-legal-set talking to.

9—Fiona, Peaceseeker

Erik feels the "0: Gain life" ability is too strong and Melissa feels it's irrelevant. The weird part is, Magic is so many games, they might both be right. Planeswalkers are hard. That's why I put mine last.
Ethan: I like this one a lot less than I liked Jori. There are so many game states where I won't even want to cast this. If my opponent has more creatures than I do, or if my hand it empty. I generally prefer planeswalker designs where I can usually cast my planeswalker with confidence and then figure out which ability is best to activate.
I think some planeswalkers want to be great in any deck that can play them, but I'm a huge fan of the more common design for the last few years where a walker is only at their best in a deck built around it. If we can make build-around uncommons, I feel like we can make build-around planeswalkers.

10—Bought the Farm
Mark: We've done Oblivion Ring–style effects on Auras, but never on an enchant land. Lands are traditionally difficult to destroy, so I'd probably want to make sure the set had an answer to it, but I like the general design. I'd change the flavor, but it's a cute design name.
This was the design I was least confident in. I still love the flavor despite Mark's opinion, but I'd definitely cost it at {2}{G}{W} next time. (5cc is closer to reasonable but just too late to care about one-mana acceleration.) Lands being difficult to destroy is irrelevant because this is strictly easier to answer than Oblivion Ring.

from Jeremy's test:

Melissa: I think this is a cool Control Magic effect, but it's costed very aggressively. The drawback here does not justify the mana reduction here. I think this looks sweet with some number adjustments.
I also really dig this design. Anything that Melissa warns against, I'll take extra time to look at and test, but my sense is that this is very close to printable: Most of the creatures who pack their value in effects rather than size won't survive the transfer.

4—Pam, Shapeshifting Planeswalker

The graphic designers at Wizards would do a better job splitting the text here, but it's still too much (it's 10 lines on a normal frame). Always use MSE to visualize your big cards.

5—Sign In Someone Else's Blood
Melissa: There isn't really anything new here, this is just Altar's Reap. There is an added bonus that sometimes your creature sticks around, and an added drawback that your opponent can remove the target, but I don't think there is anything very innovative here.
Melissa dings iterative designs like this a few times. She's not wrong, but I could absolutely see a design like this in a set like Ixalan. I would tweak it so you can target yourself. I might also make it a sorcery so it replicates the Altar's Reap play pattern less.

6—Whimsical Djinn
Ethan: I don't really buy this card as a mythic rare; it's too goofy and random. These types of cards are very unappealing to Spikes, so I wouldn't want to push this for Constructed. That narrows its potential roles to Limited bomb and/or casual card. It's probably appealing enough on those axes to make it into a file for a playtest.
That's a surprising claim right after Erik and Melissa just finished talking about how strong it is. Spike is fine with randomness when all the results are efficient and relevant.

It amuses me that no one took notice of the fact that the Djinn returns itself when it's done all three effects, presumably because you'll have cinched the game by then.

9—Remember Your Charlemagne

No one calls out the fact that this card skirts the attached-auras issue but runs head-long into the attached-equipment issue. (And we run into the Standard-set BS again.)

from Ryans' test:

4—See You Tomorrow
Eli: …This card is easily fixed by tweaking the phrase to "next postcombat main phase this turn," but not all designs can be saved so easily.
This is a helpful tweak, but why add that last phrase?

6—Kaya, Kingslayer

The only difference between this and Jeremy's Sorin is the cost and the ultimate. It shows restraint that the reviewers didn't compare them, but I do find the difference in their comments curious. Obviously two mana is radically different from five mana, but cost is also the easiest thing to change on a design at this stage.

7—Cranky Hydra

I'm surprised Mark didn't make the comment that one mode should feel red, one green, and one red-green (one feels red and two feel green).

Overall Commentary
Erik: Sometimes the key is to treat a card design as a draft, then come back and revise the card later, perhaps the next day.
We see a number of comments like these throughout the judging. The intent isn't derogatory, but the effect is a bit condescending. It's like Erik can't imagine the contestants as potential peers. That's not a surprising attitude for an expert in a field very few have worked in professionally, but it's not the most productive either.

Ethan and Mark go on to suggest that Ryan didn't playtest. [There was a communication error prohibiting playtesting with others.]

from Linus' test:

2—Lorelei, Centaur Leader
Melissa: I don't see anything innovative on this planeswalker…
To this day, I don't have a firm grasp of what each of the judges believe is innovative or not. Calling out cards that do new things in new ways as indistinct because their gameplay pattern isn't as new feels like an awfully high bar to me.

3—Feint Strike

This card is Diabolic Edict plus half of a Dual Shot or Geistflame. There's hidden value in that you can ping the creature you don't want sacrificed (even if it doesn't kill it) in order to force something better to be sacrificed—and that's very cool—but I'm not sure it needs to be uncommon and cost {2}{B}{R}. If the intent was that you get to target which creature they sacrifice, then it's slightly under-costed.

5—Ashiok, Dream Collector
Eli: Casting things off the battlefield is theoretically possible, but it causes no small amount of weirdness. What if the other player want to activate its ability in response? Priority fights aren't fun. What if it's face-down? I can move it to the stack, look at it, and then determine whether or not I can or want to cast it. Perhaps you can get to the same goals by having the emblem let you pay a permanent's mana cost to exile it and cast it for free, or even having it simply gain control of a one by paying its mana cost.
When I saw this comment, I scrapped an entire folder of designs. If only I'd managed to scrub the idea fully from my mind

6—Red Rover

Eli and Ethan point out rules issues and digital issues with this card, both of which are real considerations for a Magic designer and not something you could reasonably expect anyone to know who isn't receiving a steady paycheck from Hasbro.


At least two judges missed the fact that this spell can only target cards you don't control, and so building around it is not the primary use of this card. It's somewhat heartening to know that even the professionals misread cards at times, but in a competition where designs are judged without the possibility of clarification, it's pretty frustrating.

from Scott's test:

1—Easily Excitea-Bull
Mark: I'm a fan of this card. To deal with some of Eli's issues, I'd contemplate making the card a little more expensive, moving it up to mythic rare, and granting a +1/+1 counter per life gained as that's clearer how many counters you get.
This is the kind of response I'll shortly claim we could stand to see a lot more of. No one cares if our submission doesn't make it through the entire process untouched. As long as the nugget that makes it fun and unique is appreciated, that's great. Seeing a fellow designer riff on the design is actually a great compliment because it means you found something that excites them.

6—Come at Me Bros
Melissa: I really dislike this because it causes huge feel-bad moments for the opponent when they plan out their turn without attacking, tap out, then pass the turn. "Go, you say? Actually hold on, before your combat step, I cast this." According to the rules you must go back to combat, and the opponent must attack you. This card is a huge blowout. A solution to this is to make it a sorcery and say that the opponent must attack during their next turn. That will give the opponent time to plan. Overall I would avoid cards like this.
This is a brilliant insight. Not something most people would find without playtesting.

7—Elixa DuMort
Eli: I triumphantly use Elixa's last ability and you give your creature hexproof in response. I get no awesome Zombie Dragon. "Fizzling" is always a risk, but the greater the investment you put in, the more likely we want a rules or templating trick to avoid it. Contrast the overall multiple-turn investment for Elixa versus your blue/red design's cost of just paying some mana. We've got just the trick to get around that: "Target player creates . . ." Now if the creature's an illegal target, you still are a legal target, so the ability resolves and you get your token.
Another sweet save. I love seeing all these little templating tweaks that improve play.

In addition to the -2 and -6 effects feeling too similar, I feel like there's negative theme in turning any corpse into a zombie dragon. Zombies work that way, but dragons do not.

The judges seem to judge the cards as if they're the final design and no iteration is possible; Like we can't even tweak the templating or the costs. I'd like to have seen more, "this is the best part of this design, and here's what I'd do with that," which is a more productive form of collaboration common to teams in the rest of the industry.

The judges ask the designers to be more innovative in their designs, even while acknowledging that it was the polish of their safe designs that got them into the Top 8. What do you think is the judges' priority from their comments and choices?

Lest I fall into the trap of leaving the solid stuff unpraised simply because it's not unusual, I must say that all the judges brought a metric ton of wisdom and insight to the review, I rarely took issue with Ethan's comments, and I appreciated Mark's enthusiasm and support. Melissa and Eli seem to have a sixth-sense for impending problems. Erik frequently answers with one-liners; It might come off like he has better things to do, but I appreciate not saying more than needs to be said—that's a skill most people lack.

It was nice not having a Gleemax reviewer. Dealing with this much criticism without the chance to explain yourself or adapt is stressful enough without someone insulting you. I wish that the public would treat the R&D team with as much care.

I'd be very curious to see what numbers you think judges might have assigned to each contestant's submission every round, based on what they've written. See if you can predict who advances.


  1. I thought Melissa's comment on Mama Bear was a really cool way of making the card less of a beating in Limited while retaining the feel of it.

    1. Agreed.

      That comment for reference:
      "At common, this card will be the backbone of any green or white Limited deck, and a three mana 4/3 is very aggressively costed. I think the flavor is very cute, but remember to keep in mind how frequently players will see these at common. I would have taken a different direction with this; adding a counter to the smaller creature instead of Mana Bear. The flavor remains, and you won't be beating down with an overstated creature on turn four in most of your Limited games."

  2. I playtested, but not with others. I specifically messaged Mark because he claimed that he’d told us we could playtest, but the legalese we got back made it sound like it was against the rules. He said he’d been edited in the FAQs and would make it more clear moving forward.

    1. Thanks for the correction.
      I'd forgotten how we only learned after the fact that playtesting was allowed.

    2. I think solo playtesting, though it has some limitations, is the biggest design tool available to me, and I'm always amazed how reluctant everyone else seems to do it.

      If your game isn't interesting enough to enjoy when you get to make 100% of the decisions, how will it be interesting enough when you only get to make half/a third/a quarter of them?

  3. Trying to make a hard-core elimination-style contest out of a type of work that leans so very heavily on collaborative iteration really ends up going to some strange places in judging and evaluation.

    In "real life," all these concerns and critiques would lead to discussion and debate and potentially changes in one director or another. In the end, it might even turn out that you (as the initial card designer) were right all along. And in "real life" you probably don't consider that your job is "at stake" during this whole process, unless maybe all you ever get is negative feedback.

    But that common collaborative work dynamic gets weird when you're in a position where disagreement could potentially just end things. It makes for interesting tension for the "audience," but ultimately it's kind of a bad way to go about doing this all.

    1. And up until Erik's critique, I still did not understand what Ashiok's ultimate on Linus's card even did. Now I get it.

    2. Yes. This experience was so different from what I'm accustomed to in the games industry.

  4. Oh! Are we going to have a thread/challenge/what have you on the first top 8 tribal challenge?

    1. I honestly have "Things To Say" about guideline 3 in that challenge. I think a lot of the problems I have with how the Ixalan set shook out can be explained by that guidance. I was experimenting with designs yesterday afternoon and kept coming back to that rule and shaking my head at it.

    2. I am excited to see what tribes people chose.

    3. I am really excited for this too. I wonder if mechanically caring about the creature type has to mean referencing that creature type, something I tend to dislike at low rarities. For instance, is a common card that cares about very high toughness, with proper flavor, sufficient for a Treefolk tribal card? Or need it be narrowed only to Treefolk explicitly. My gut feeling is to prefer broader cards, but maybe I am thinking more of sets with samplings of tribal rather than full-on tribal sets.

    4. Yeah I wanted to go with apes and have them care about forests. I really like Avatars but that would be tough to flavor. Also I wonder if unusual tribes get bonus points

  5. "Seeing a fellow designer riff on the design is actually a great compliment because it means you found something that excites them."

    Oh, good! I always worry that I'm being terribly rude

    1. I think iterating on an idea is a lot of fun. It forces you to find the good/interesting part even in a design that is subpar.

  6. Regarding Bought the Farm, I was a little surprised that none of the judges connected it to Chained to the Rocks.

    I did think repeatedly as I read through the judges' comments on all the tests that the rules hadn't been stated clearly enough. It must have been a frustrating thing to be on the receiving end of contradictory comments.

    I think I liked Ethan's comments the best; it's a pity he's not a judge on the rest of the challenges.

  7. Your analyses, as always, are informative to read. I would caution you, though, against critiquing your potential employer too harshly in a public forum. I would hate to see that held against you, but maybe I'm just overly paranoid.

    1. The hope is that they appreciate feedback as much as I do.

    2. Definitely, and if I were your coworker in any field I would value the critiques you provide. The picture we have of R&D from Maro and from old multiverse comments is also one where critical feedback is common and encouraged, so it's probably not a problem.