Tuesday, April 24, 2012

Developing Our Fan M13 Set

As many of this blog's viewers may know, we've been designing a fan M13 set for several months now.

While individual design members have been playtesting the cards with their local playgroups, we have not been playtesting as a group during design. Now, we've established a way to play online and a development team is slowly coming together.

The deadline for our fan M13 project is midnight, May 29th, and we only have slightly over a month to make this a polished set that is fun to play and is reasonable as a real set.

I'd like to discuss the things that need to be done to make this happen.

Before I continue I would like to mention that this project started out on the basis of open participation and it still is an open participation project. Although there will be a development team to make decisions, anyone is welcome to make suggestions and design cards (that they don't mind us using and presenting to WotC), or help out in playtesting.

First, the things we need to decide for Limited:

Set the Pace
One of the first things we need to decide is how fast or slow the environment is.

One goal which I believe in is to try to support both aggro and defensive styles within the environment. People have different preferences, and I think sets are more intriguing when both are possible, rather than just the slow games of M10 or the super-fast games of M12.

There is an argument to be made for fast, tempo-oriented environments, as they tend to revolve around combat more than card advantage; beginners intuitively understand the meaning of actions that reduce life totals, while card advantage is a higher concept sometimes invisible to them.

However, since this is coming after the fast-paced M12, I do think it makes sense to slow down in M13. I also believe we also can't go back to the slowness of M10 that often caused games to be determined by bombs. I think a good place to aim for is "slightly faster than M11." I would like to pursue how much of a contrast we can make between aggro and slower strategies within the same set, giving tools to both.

Identify the Archetypes
Once the pacing is set, we need to determine the archetypes that we want to support for this set.

Zac Hill has talked about how Development discusses the archetypes it wants to support for each color pair early in the development stage. I feel that this is exactly what we need to do at this stage.

Core sets need to serve the needs of both beginners and advanced players, which is a tricky goal. I believe archetypes are one key to making that possible. For the sake of beginners making their first forays into Magic, core set cards are much simpler than expansion set cards, but even when the individual cards are very simple, the archetypes you can draft with those cards can be very deep and fun, as sets such as M12 show.

For example, a Zephyr Sprite might be boring as a card to some players who've seen it before. But how about drafting a deck where you enchant Zephyr Sprites with Unstable Mutation and Oakenform protecting them with counterspells? There's no removal, you just disable opposing ground threats temporarily with cards like Frost Breath and Unsummon just long enough to outrace them. That's not boring. It would feel exciting - like flying by the seat of your pants.

A Craw Wurm might be boring as a card to some players. But how about drafting a deck that plays Overgrowths to pump out Craw Wurms and uses Pyroclasm to clear the board of small evasive threats? That's not boring to play. It would feel satisfying - your patience pays off, and you get to crush smaller opposition with overwhelmingly big guys.

I'm not sure whether supporting these particular example archetypes are feasible, or whether we want those particular archetypes in this M13 set, but the point I want to illustrate is that a set can give you many varied experiences and many varied play styles, each producing a different feeling, even when the cards are simple.

Zac Hill also talks about how this kind of depth and variety in game play doesn't happen automatically. As far as I understand, in real sets it happens because Development ensures that a sufficiently large number of strategies each get enough support. They also ensure that a large portion of the cards care about what deck they go into rather than being universally good, so that the viable strategies don't all converge into one "collection of best cards in the color" strategy. There may be many other factors as well, for example taking pains not to stifle particular strategies.

While design produces ideas and elements of game play, development is where these ideas get woven together to actually become a game. I believe Development is very much a kind of Design, except you're designing game play rather than any single card. Because of that, I don't think we can calibrate anything in the set until we know how we want the set to play and what strategies we want to make available. Without that, we have no goal or target to calibrate towards.

So, we need to generate ideas on how each color pair can play. I hope to be holding discussions on that in a Google Docs document for posting ideas on the identities of each color pair.

Playtesting to Spot Budding Archetypes
One thing to remember is that we shouldn't just choose the archetypes in a vaccum; we should let them grow out of the cards in the design file as much as possible. For one thing, if we always determine the archetypes in a set based on the natural traits of its colors, we may come up with very similar identities for archetypes every time, such as RW Boros or UR Spells. For each set to feel unique, we should try to build on the quirks and traits of the cards submitted.

To build on the card designs that we have as much as possible, we need to play a lot of limited with this set and find what budding archetypes and synergies exist. Once we identify those potential archetypes we can support the ones that seen fun to try to make them into full-blown archetypes. I think the best way to realistically testplay in the beginning is to testplay a lot of sealed online, since drafts require many people to be online together at once.

We should set up a style of playing where we're not trying to win but to try out the most strategies. With each sealed pool, players try to build 3-5 different decks that use different colors, or focus on different synergies, so that different styles get examined.

Doing it in Parallel
I wrote above that we shouldn't design archetypes in a vacuum, but on the other hand, I don't think we should wait until the initial testing is done to start discussing archetypes. First of all, these online discussions take time, as members log on to comment then log off at different times, responding to others over different days. We should try to do as many things in parallel as possible. There are some aspects of the identities of colors that need to be discussed that are basic to most core sets, and we can get started on that right away.

Secondly, some card designs don't occur to you until you've started to think, "what factors can cause a fatty RG deck to want to draw upon different cards than a RG aggro deck does?" So it's good to start thinking about it right away rather than just wait for it to occur naturally.

Design Archetypes Around the Bond Creatures
During the design stage, I proposed that each Bond creature should spearhead its own archetype based in its two associated colors. Each color pair has two Bond creatures in it, so preferably one of them would be the aggro version while the other would be the control version. This could be one way to build archetypes into the set based on the existing designs.

For example, Kird Ape could be a centerpiece of the RG hyper-aggro deck.

Hillside Charger could be representative of the GR ramp-fatty deck.

Vaportrail Imp could represent the UB evasion-and-tempo deck.

Dark Cove Pirates could represent a slow BU controlling deck that grinds out the opponent and kills slowly and steadily with a few evasive win-conditions.

One of the reasons to take this approach is because Bond creatures should be powerful, and I think it's good when the powerful cards in an environment each lead to different directions and each shine in a different deck.

The reason I think Bond creatures should be powerful is because they are like multicolor cards; the main point of playing them would be that the extra land requirement allows you to get a better creature than what you usually get for that mana range. (We still need to see whether a "Grab land fixers and every Bond creature you see" strategy can be fun or not, and should try to make it work if possible, without turning the Core set into Shards of Alara 2.)

If Bond creatures are powerful, they create many interesting implications. They lead you down different color combinations in draft, and make building a Sealed pool more than a trivial activity.

So we could try to use the bond creatures as motifs and envision a deck where each of those bond creatures would make the most sense in. However, some of the Bond creatures like Windborne Pteron, Minotaur Cultist, and Mastodon Calf are strong in a general way so that any deck of their colors would want them. I think that's fine.

These might not directly contribute to deck diversity, but I still think it's good that they're appealing the way they are. (By the way, the Pteron and Mastodon may need change because they have some other issues, but that is another topic). We can still try to ensure a variety of archetypes exist in their colors using the other cards that surround them.

For example, Minotaur Cultist's BR counterpart is:

This guy probably wants to be in an aggressive BR deck that uses its removal for getting early blockers out of the way rather than for long-term board control. I've read that RB naturally gravitates towards a slower board-controlling style; it doesn't particularly need help to exist. If we can support the aggressive BR deck somehow, we may be able to have both the aggro style and the control style.

We could try putting in other high-power creatures at common such as the 5/1 vanilla Rotting Fensnake in the black commons or the power-pumping Feral Ridgewolf alongside the 4/2 Scorched Shambler so that there are enough of these guys to form the basis of a strategy. Then we could support those high-power cards with cards like:

(These last 3 support cards aren't currently in the file, they're just possibilities.)

This might allow an aggro RB deck and a slower RB deck to coexist. Of course, this is just an idea. It's all theory and the actual "high-power" deck might not work out when we playtest them - it might feel too artificial, gimmicky, and/or single-minded. It might feel like a "pre-designed" deck that is forced onto the player. Or, it might be extremely swingy based on the order you draw spells and based on whether the opponent drew removal. But it's just an example to show that something like this might be possible.

By the way, I'm not sure about what should be done with the last two Bond creatures; they don't seem particularly important for any strategy.

Maybe the Royal Navy is subtly good, but the Druid seems weak since GW is a color combination that usually needs a little help. Perhaps they could be good in an Aura-themed deck because of the vigilance? Not all the archetype contrast need to be about differing speeds. However, we should also consider changing these cards, including having a different Plains Bond effect.

Two More Things to Do in Parallel:
Here are two more things that need to be discussed in parallel:

We need to consider what trends we'd like to create in Constructed.
We need to make sure there are enough cards that impact Standard in various ways, whether it's by being the center of a new deck type, answering a problematic card, synergizing with Return to Ravnica cards, etc.
There should be Constructed impact at all rarities if possible (such as Lightning Bolt and Mana Leak being in various core sets).

Find Product Selling Points
Most sets have a cycle or two of very iconic splashy cards that people identify with the set. We have couple of cards that might possibly be worked into such a cycle (such as sorceries with drastic effects, or fatty creatures with powerful abilities) but they need rework. Or, if they're not working, we could install a new cycle, such as a cycle of Commands etc.

Getting all of this done in time would be quite a challenge, but hopefully through the awesomity of our enthusiastic designers we can do it. I will be looking forward to sharing any lessons we learn through these blog posts.


  1. Minor Flavor issue: Dark Cove Pirates looks like Mutant belongs somewhere on its type line.

    1. Right. Human Pirate Crab Mutant it is. Now, if we can fit in a Turtle Ninja Mutant somewhere...

    2. Presumably he's a Pirate of Dark Water.

  2. I'm really excited to have Chah leading the Development of M13 because he's put a lot of thought and passion into this and I'm confident our team of Developers will ensure that the set is as fun to play as it was to make.

    I agree with and support most everything Chah has laid out here. I do want to clarify that the Limited archetypes aren't complete unknowns going into this phase as the Design team (including Chah) was very cognizant of them from the beginning. That said, only play testing can show which of our plans bore fruit, which are legitimately fun and which need a lot more work.

    While ensuring that there is an aggressive as well as a controlling build for every allied color pair is an excellent place to start, it's important to remember that symmetry has no value at this abstract level and it's much more crucial to simply find as many awesome deck archetypes as we can.

    Chah covered what we're hoping to see from the allied color pairs, but we'd also like to see some enemy color pairs if the set can support it. Going farther afield, I've been hoping that Cde, that is one color splashing its two allies is a possibility as well.

    I would argue that there's no particular value in forcing a splashy set-selling card from each color into a cycle. Five unrelated awesome cards will sell the just as well. In fact, since a cycle homogenizes a concept—even one as exciting as the Titans, for instance—you'll often garner more excitement with a handful of unique cards.

    We got our first round of online playtesting in this last Sunday and it was very cool. If you like Prereleases, consider joining us for some M13 Sealed games because it has all the fun of discovery with some added excitement knowing that there are cards whose brokenness hasn't been found or expurgated yet.

    1. I'm commenting to add emphasis to the third paragraph of this comment.

  3. Maybe we should explicitly test certain color pairs? We could do it in Sealed or we could sit down and simulate a WB draft deck by selecting up two of each white or black common and up to 6 total uncommons.

  4. I'm a little bit worries that Dark Cove Pirates and Vaportrail Imp both fit into both decks too well. You do want to hold the ground while flying over, and it's not like Pirates isn't an evasive clock. Alternatively, Imp is a fast finisher and can trade for a good number of 4 and 5 drops on defense. Then again, that should be fine if we can otherwise differentiate the archetypes.

    On other fronts, Bloodsplatter Devil + Matron Willow is sweet, and we could possibly make either token production or sacrifice for value easier to do in red green to make one or the other a build-around card.

    Another potential archetype I noticed was a white deck using Pyroclasm as a tempo play. Regal Unicorn, Mastodon Calf, and Elephant Rider all survive it and beat down effectively at common.

    Cursed Tomb has the potential to make a reanimator archetype with an easy sacrifice outlet and looting, but isn't really there right now.

    It seems like there could be a cool WB deck with Islands to make Winborn Pteron and Vaprotrail Imp fly, but needing a third color doesn't work well with white-black's highly color intensive nature.

    That's all I've got for now, more to come with more testing.

    1. RE: Vaportrail Imp + Dark Cove Pirates - A bind that you guys seem to be finding yourselves in is that you've focused a good deal on making sure the creatures are reasonably playable regardless of whether you've activated the bond-quality. As a result, there's something of a flatness in terms of quality — everything is either OK or really good. I think this pair really highlights that, as non-bond Vaportrail Imp goes from negligibly worse than Barony Vampires to better than Daggerclaw Imp. Likewise, Dark Cove Pirates is your typical blue-fortress durdle that can also turn into a relatively efficient beater.

      Somewhere you're going to have to imbue greater variance in these commons, which I believe is something Jay plans on addressing with his forthcoming article on Mastodon Calf and Windborn Pteron.