Designer Fun – Set Doctor

During the last week I became aware of a project that GDS2 finalist Jay Treat is working on with some… other people he met on the internet? I actually don’t know how they got together on this, but they are all aspiring Magic card designers. They decided to make their own version of M13 for fun. It’s not a prediction of what WotC will do, but instead they’re collaborating to see what kind of core set they can make together. Design is best learned through experience, and this is a great way for them to get better at it.

In today’s article I’m going to do a little Set Doctoring – it’s like Deck Doctor, but instead of a constructed deck, I look at a set design and viciously tear it apart. I want to say up front, before I get into it, that I respect what these guys are trying to do. Designing Magic cards is very difficult. Also, I’m looking in less than halfway through the process, so of course there are problems. They haven’t finished design, let alone tried their hands at the development process. I hope my advice to them will help them improve their set. I will certainly be criticizing some of their cards and decisions, but I want the reader to know that this team is doing a solid job so far, and has a great shot at making a realistic (and fun) set if they keep working at it.

Comments & Communication

A set’s quality depends a lot on the quality of people working on it. Even if one of you does something wrong, if you have great people, several of them will point it out. The reverse case is also important. If one of you designs something brilliant, the others will call it out and you can be sure to keep it. Be sure to comment if you like a card just as often as you comment because you see a problem with a card. The [card]Goblin Artisans[/card] (I really should have asked what they wanted to be called) are doing well here. They have lots of comments in their googledocs excel file, and I’ve seen several cases of one person catching a flaw with a card or commenting that they like what a card is doing.

Basic Elements of a Core Set

Core sets are different from the big fall set in a lot of ways. They don’t have an overall story or flavorful theme. Instead they are guided more by what’s come before them and what will come after. The first step for core sets like M10 and M11 (and I assume M12) was to decide what cards from previous core sets should be repeated. As a general rule, current core sets repeat about 50% of the cards seen in the core set immediately prior. The Goblin Artisans team is doing a good job on this front. They’ve observed the 50% carry-over and are attempting to maintain that. They’re even tracking the patterns they see in each card slot (as much as can be done from M10 to M12) and trying to match what’s already going on in the new style core sets. They’re keeping count of how many new cards they do at each rarity, for example.

However, I think they’re taking the concept of a set skeleton too far. They’re tracking all the card slots too literally. For example, at WotC I never saw card slots for “common white first-striking flier.” Just “common white creature.” When a set moves into the card design phase (which is almost immediately for a core set). The design lead says “everyone bring white common creatures to the next meeting” and each designer goes off on their own and types out 10 to 20 common white creature designs. I never thought much about specific roles for the commons while designing them, I just designed 20 creatures that should be white and could be common. I tried not to make any of them too similar to one another, and just by doing that I would end up with a variety suitable for filling out the commons in general. I invariably ended up with a couple of uncommon and rare designs too, which I would tuck away for a future meeting.

At the meeting, we read all the designs and picked our favorites (as a group, usually with a vote of some kind). We put the favorites into a list with the cards we chose to carry over from the previous core set, sorted by CMC. Then we looked for holes in that list. Only at that point would we narrow in on roles, like “we’re short a flying creature for white, and based on the CMC of the other cards… it should cost 3.” That meant only a very few cards were designed to such tight specifications.

The precise categorization of [card]Roc Egg[/card] (in their excel file) as “guy who wants to chump” and then putting in an effort into designing alternates… is cute, but I feel like doing it repeatedly over the entire set design is misdirected effort. You should be trying harder to generate top-down designs for things that you feel Magic has never done, or has never done “right.”


This team decided they wanted to try a mechanic they are calling terrain. It’s basically the [card]Slavering Nulls[/card] ability – creatures that get a bonus if you control a particular basic land type. When I first learned of their set Jay asked me what I thought of this mechanic and my reaction was that it seemed to be fine. I’ve had more time to think about it since then and there are some things I like about it, and some I don’t.

I like that it cares about basic lands – which is something every limited game and almost every constructed game of magic (especially in standard) has. I like that it’s easy to understand – if you have the land it asks for, the creature gets the ability. I don’t like that it asks you to know about your opponent’s lands in order for you to know what their creature does. Even the top players don’t usually know what lands their opponent has in play at all times, while the medium and new players almost never know.

Let’s compare:
Island Lair – unblockable (~ is unblockable as long as you control an Island)
Islandwalk (~ is unblockable so long as defending player controls an Island)

These two are similar abilities, but they have a very critical difference. When my opponent plays each of these creatures, and I read it, my reactions are different:

Case 1: I ask “Do you have any islands?” you say “no” and I forget about the guy’s ability. He’s vanilla in my mind. On your next turn you play an island, attack with the guy, and I try to block it. You’ll have to remind me about its ability.

Case2: I read the guy and I immediately know if I have islands. I know because I’m playing my deck, that I built, and I sure as heck know if I put Islands in it. I also know if I have one in play without even looking.

It seems like such a small difference, but it very much matters. Terrain-like abilities were tried in Zendikar block – [card]Slavering Nulls[/card] is the example I used earlier – so this issue is clearly not a big enough problem to prevent a card from being made. The question remains: is this good for a major mechanic in a core set?

Now compare terrain to a variation they’ve already got in their set, on cards like this:

Sunblessed Tactician
Creature – Human Soldier
Whenever you play a Plains, you may tap a target creature.

This card essentially has Plainsfall (landfall requiring plains). Landfall has proven to be a great mechanic, and the specific version of it is also good. I believe landfall is a better mechanic than terrain, because it’s much easier to interact with. When your opponent plays a land, they tell you about the trigger, do the thing, and move on. Later, when you’re attacking and blocking, you don’t need to know if they control a plains or not. You can ignore their land pile. I think my greatest fear for terrain is that it looks bad next to landfall. This problem is exasperated because terrain effects have to remain small, because having a land always turns them on. How would you make an [card]Admonition Angel[/card] with terrain? [card]Admonition Angel[/card] can be so powerful because after you have 6 lands to cast it, you need a 7th to get the first effect. Then, you can keep getting the effect over and over – that would never happen with terrain. A terrain card with a big bonus (+5/+5 & trample, for example) would still have to cost enough to be fair (5+mana), but it would also always have the bonus in any reasonable deck, because you wouldn’t get up to casting it without hitting your second color anyway.

The Goblin Artisans team has already started discussing how much color splashing you want. The terrain mechanic very much needs to know what that plan is, and so do many of their individual card designs. The set is clearly still in a state of flux as they work out these issues, so some of the internal conflicts I’m observing in the set are a result of that. In case they don’t realize it, though, I’m making sure to point it out: Many of these cards don’t work together in the same set. Perhaps specific examples are in order…

Time to take a look at some cards I liked and then some that I didn’t like. I didn’t have time to carefully comb over everything, but I tried to read it all at least once and grabbed some designs that caught my attention.

First, Some Cards I really liked:

For God and Country / In Elspeth’s Honor
Creatures you control gain lifelink until end of turn.

Simple and effective. Development may seek to make this cost 2W, as there’s no particular need for it to be so white-focused. It may even be 1W, if there are a lot of other 3-mana white spells at common.

Plated Griffin
1/4 Flying Griffin

Another simple set of numbers that hasn’t been done (there are no 3-mana 1/4 fliers in all of Magic). They even have an excuse for all that toughness in the name.

Supreme Anthemist
Creature – Angel
P/T = # of Plains you control.
Other creatures you control get +1/+1 for each Plains you control.
Canopied Seaside Terrain- ~ has hexproof as long as you control a Forest or an Island.

You’ll note I used some strikethrough on this one. I like it, except for the clunky additional terrain ability. I know hexproof is the new hotness, but you guys have added it to far too many cards, especially in White. Also, I hope you know you can’t use this “Canopied Seaside Terrain” wording, even if you use the names of your land-typed duals. In other news, I’d submit this to development with a more aggressive 4WW or 5WW mana cost. The third W is unnecessary on a card that already cares about plains, and it’s reads so awesome with the “everybody [card]Empyrial Armor[/card]” ability that I want it to actually be pretty good. Nerf it up after you test it, if you need to.

Deity’s Chosen
First Strike
Whenever ~ deals combat damage to a player, put a divinity counter on target permanent. That permanent is indestructible as long as it has a divinity counter on it.
Terrain – You may have ~ assign its combat damage as though it weren’t blocked as long as you control a Forest or an Island.

Another awesome mythic card on which I’ve cut out some absurd terrain text. WWW but you want a Forest or an Island? Really? You should use designs to teach new players how to build decks, not to encourage them to build manabases that won’t let the cast anything they draw. This team has a lot of design knowledge, as shown in some of their excellent posts on Goblin Artisans and some of the great comments in the excel file, but this design contains a significant flaw. However, it was easy for me to rescue it by deleting the terrain text. It’s easy to crowd extra text onto a card that you designed because you want to be sure others will like it. I learned this first hand when I submitted a card for Fifth Dawn. I gave it two abilities, but the first one was so cool that the second just looked silly next to it. The card actually got moved into unhinged. Which card? [card _____]_____[/card]. (The card that can change its name to any name.)

Back to Deity’s Chosen! The base design is very exciting – making more and more of your stuff indestructible with each attack sounds awesome. Almost everyone will realize you should make itself indestructible with the first hit, and that makes the player feel smart for figuring it out. How to get it through… well that’s the part of the puzzle that makes deck construction fun. Maybe you just put it in a regular white weenie deck. Maybe you load up on equipment or auras, maybe you have removal to clear blockers. You’re best plan as a designer is to leave that up to the player to figure out.

Please also note that WWW is a very restrictive mana cost, especially on a small creature. There are 8 cards in all of Magic with this cost, and the only recent one is [card]Devout Lightcaster[/card] – a color hoser for one of White’s enemies. It might be okay here, but due to the desire to combine this guy with evasion, removal, etc, I’d go for 1WW instead. I personally remain mystified as to why amateur designers gravitate toward adding extra color symbols to their designs. Does it look cool? Like flames on your hood?

Rageturn Sphinx
P/T = # of Islands.
Whenever you gain life you may pay that much mana, if you do, draw that many cards.
Terrain-Lifelink as long as you control Plains or Swamp.

Maybe this is the kind design that can make terrain viable on rares. It has a build-around ability focused on a life = cards engine. The terrain enables it to self-solve, but you’d still want to gain more life to draw more cards, so there is still build-around potential even once you realize you should play black or white with your Blue sphinx.
As I will emphasize and explain below (on a card I hate), the P/T scaling with islands is a bad idea here. As is the UUU in the cost. I would make this a 4UU 5/5.

Sanguine Hypnotist
Creature-Vampire Wizard
Lifelink (Reminder text.)
Pay 10 life: Target player discards three cards. Activate this ability only during your turn.

This might turn out to be unfun, or really bad for Magic, but I like what you’re trying to do here. I’d at least want to test it. Oh, and this design is not a common. It’s just not the sort of thing you need two or three of in a limited deck. Also “pay 10 life” is too far out of the comfort zone of most players for it to appear at common.

Knowledge Sap
Whenever an opponent discards a card, draw a card.

An excellent build-around card that provides a different reward for making the opponent discard. There’s a good chance this is absolutely miserable to play against and therefore shouldn’t ever be printed, but it certainly should be designed and sent to development so they can find out.

Fan the Flames
Whenever a red sorcery or instant you control deals damage to an opponent, you may draw that many cards. If you do, put that many cards from your hand onto the bottom of your library in any order.

A powerful card for Red burn decks to take advantage of. This wording is probably too strong and too “in control” for Red. R&D would most likely have you discard that many cards at random. You could try “whenever an instant or sorcery you control deals damage to an opponent, put your hand on the bottom of your deck and draw cards equal to the amount of damage dealt.” Though it loses the smoothness of “deals damage -> draw that many cards.” Try a few variations to see if you can come up with a slightly more Red-feeling wording.

The Many Eyed Serpent
X/X Hydra
~ can block up to X creatures and can’t be blocked by fewer than X creatures.

I like the use of scaling up the blocking and can’t be blocked on this card, it’s got great flavor – each of the heads being able to act independently.

Heavenly Perch
Whenever a Plains enters the battlefield under your control, gain 1 life.

This is a great alternative cycle to the [card]Angel’s Feather[/card] cycle, though I don’t know why they don’t count the opponent’s lands as well. They’d be great in a set that encouraged mono-color decks… but in this M13 they fight against the terrain cards and splashing of other colors. (Note that in their own comments the group seems to still be discussing this point.)

And now for some cards I really disliked:

Players can only untap one of their Swamps and one of their Mountains during their untap steps.

What is this, 1995? Magic stopped making savage basic land hosers a long time ago, and for good reasons. They are far too punishing, especially for casual play. Most playgroups end up banning them after one game. This card would be bad for Magic.

Mirage Mage
Human Wizard
Whenever you draw a card, you may have target land become a basic land type of your choice until end of turn.

This fixing is far too powerful for Blue. The common Blue land altering cards turn stuff into islands.

Contest of Gills
Counter target spell unless its controller pays 1 for each Island you control.

This [card]Mana Leak[/card] variant sounds cute at first but it lacks [card]Mana Leak[/card]’s drawback (which is: later in the game player can afford to pay 3 for their spells and get around it). It’s nice that it encourages mono-Blue decks, but it’s so good that it goes beyond encouragement into forcing control decks to be mono Blue in order to play the best counterspell available in the format. That’s bad.

Reclaiming Waters
Creature – Elemental
~’s power and toughness each equals the number of Islands you control.
Terrain — Whenever Reclaiming Waters deals 3 or more damage to an opponent, if you control a Plains or a Swamp, exile target permanent that player controls.

This card has the same problem as Deity’s Chosen, except with this design there’s nothing to save. It needs lots of Islands to make it big, but it wants a plains or a swamp in order to do a super powerful effect… so how do I build my deck? 4 [card]Terramorphic Expanse[/card], 4 [card]Evolving Wilds[/card], 1 Swamp and 17 Islands? That’s only if both of those basic fetchers are available, which they probably are not. The Goblin Artisans have talked about partly solving this problem with dual lands that have both basic types (like Ravnica duals without the life payment for untapped clause), and maybe it works out, but the player who opens this card won’t always know those lands exist. You can’t say “but look over here, I’ve designed all these other cards in order to make this card work” – this card still has to stand up for itself. Cards that are more exciting after you learn about other cards in the set that make them better are good. Cards that don’t work or look bad unless you know about the other cards in the set that make them work out, are not good.

I found a black card with a similar problem. (There’s a red and a green one too). Counting lands doesn’t work with your terrain mechanic. You will have to choose one or the other.

I noticed [card]Doubling Season[/card] was suggested by many of you, though luckily one of you noted that it’s too good with planeswalkers. R&D tried [card]Doubling Season[/card] in M10, briefly, and it was completely busted with Planeswalkers. That card will never be printed again.

Keep at it!

Finally, I want to offer them some encouragement. This M13 project has a lot going for it, mostly in the talent it has recruited to work on it. It’s at a critical stage where it needs to figure out (or get rid of) the terrain mechanic. It’s important to test what needs to be tested. If you’re unsure of your mechanic, test it out more, compare it to other options, and spend more time thinking about all the different ways it will affect players. If you’re confident the idea is solid at the core, spend your time exploring a greater variety of implementations and uses for the mechanic. In the case of terrain, it may be that they’re thinking of these creatures the wrong way around. What if they Red creature that asks for Plains is not a reason to splash White in your Red deck, but instead a reason to splash Red in your White deck. Would that lead to different designs? I hope they can figure it out.

If you want to check out what they’re doing, or just give them a shout of encouragement, you can check it out here

See you next week.
Oh, and feel free to follow me on twitter: @GregoryMarques

Share this


Scroll to Top