General The Problem is the Solution is the Problem [Archetypal Design]

[Kyultu]

I've been trying to tighten up my cube lately, reducing its size and focusing a little more on archetypal centric design. It's working, the decks I'm supporting and getting drafting and playing well and incoherent piles are a little less common. But It's gotten me worried that the more defined I make the supported archetypes in my Cube, the less varied the experience will be from draft to draft. I'm not even close to getting there yet, but I fear reaching a point where my cube becomes a "solved format" much like most released sets are before the next release with everyone on relative autopilot. Worst case and exaggerated scenario is something like a cube night becoming the same 8 people each playing 1 of the 8 supported decks that I might as well have pre-built them and handed off as they walked in.

Clearly, the addition of newly released cards and tinkering on my behalf will change the experience over time, but my anxious inner Johnny needs soothed. So, 2 questions:

1. Am I just chicken little worried about the sky falling? Is the fear that overly defined archetypes in a cube draft environment railroading the experience (eventually) at all legit? Is there any value in being able to just draft a unplayable pile or synergy-less world beater?

2. Regardless of the answer to the first question, is there a way to create a better experience (i.e. have more fun) by systematically introducing variety? Is there opportunity for innovation? As an example, I'm thinking something along the lines of having multiple ~120 card "packages" or "expansions that are randomly added to a 240-360 card base cube set. Each package would make certain archetypes viable or introduce a theme. There are probably plenty of other ways to do it, got any ideas?

 

[Jason Waddell]

I'll check in on this another day (busy today), but for the moment here is an article tangentially related to your question: http://www.channelfireball.com/articles/cube-design-the-poison-principle/

 

[Aoret]

Not sure that I really have the answer to your question, but I'll give my thoughts.

No matter what, I think there is some risk that a static format eventually becomes "solved". If we take Innistrad or Rise limited for example, the available archetypes were basically known. On the one hand, that doesn't necessarily make drafting them un-enjoyable. On the other, if you're looking for more variety, you have to find a way to inject that.

In my opinion, the problems aer very likely to solve themselves. If your format isn't balanced, it will become apparent, and you'll fix it. That guarantees us that eventually (with enough time and effort) you'll arrive at something balanced. Further, since you're reading and posting on Riptidelab, we know that you're going to be exposed to good design principles that will make your cube fun (layering archetypes, etc). Once you have something that is both balanced and fun, the question of whether you're stuck in the same tired old archetypes arises. I submit that you won't be able to resist modifying your cube anyway, so you will naturally solve this problem by introducing volatility from week to week.

Your idea on expansion packages is interesting, but is too work-intensive for my tastes. I think if I were that concerned, I'd rather just go above 360, break singleton more often to maintain my "core", and call it a day.

 

[Eric Chan]

I'm no expert on this topic, either, but I think that Jason and Skrap offer great advice. To build on ideas they both touched on, I'd look to emphasize cards that fit into multiple archetypes. Rise of the Eldrazi used Eldrazi Spawn producers like Kozilek's Predator to great effect, as the versatile 0/1 tokens could be used both in ramp decks to power out a scary endgame creature like Artisan of Kozilek, or by tokens decks, with Raid Bombardment as the engine. Meanwhile, as good as Modern Masters was, there was only one deck that ever wanted Frogmite or Myr Enforcer, so that made for a less interesting drafting dynamic when nobody was fighting for them. (Though in fairness, that set did layer its other archetypes about as good as could be hoped for.)

The difficult part about using expansion packs for your cube is what Jason likes to call the "testing bottleneck". That is to say, you might have a very finely-tuned list of 360 cards that is well balanced and plays out in an interesting manner. But your expansion packs wouldn't necessarily have the same level of playtesting, and using them could throw the balance of your cube off, or just reduce the number of playable archetypes. I assume this is why people start building more than one cube when they want a different experience, rather than patching a new set of cards onto an existing list. I think it's just easier to design, develop, and test for one environment at a time, with a fixed set of knobs and levers to adjust.

 

[Changling Bob]

Meanwhile, as good as Modern Masters was, there was only one deck that ever wanted Frogmite or Myr Enforcer, so that made for a less interesting drafting dynamic when nobody was fighting for them.

Wouldn't this by somewhat beneficial if these wasn't the core affinity creatures; like, there are cards that only really fit into one(ish) deck, so you can somewhat rely on picking them up late if you go into and/or (more usefully) find yourself accidentally that deck?


Disclaimer: I didn't draft MM.

 

[Eric Chan]

Yeah, every archetype does need some "reward" cards, cards that you can count on to wheel because they're much more valuable to you than to the other drafters. This might be something like Burning Vengeance in Innistrad. I pick on the Affinity archetype in Modern Masters, though, because between Court Homunculus, Arcbound Stinger, Arcbound Worker, and Paradise Mantle, there are a lot of cards that fit into that archetype and that archetype alone. This ties back to the original poster's concern of having eight pre-built decks in the metagame that drafters simply choose from and then pick up the cards for.

But you're right. In cube, while you can layer and interweave your archetypes, you still want some cards that will reward your drafters for following a specific path, cards that in all likelihood you can pass on with your first pick and reliably hope to table. This might be something as innocuous as Steppe Lynx for white aggro, a beefy green creature for ramp, or Blood Artist for the black zombie deck.

 

[Jason Waddell]

An analogy I've had in my head for archetypes is a multidimensional Venn diagram. You'll note that there are some regions that only support one "archetype", but these ideally should be pretty minimal and pull a lot of weight. They are cards to incentivize you to go into the "dedicated ____ deck". Think traditional limited build-arounds like Burning Vengeance.

What you don't want is this (sorry for the crappy internet image):

Obviously there's a continuum, but in the first picture we have dozens and dozens of decks with different textures to them, and in the second we have four rigid archetypes with non-overlapping support. The former will give you lots of variability from draft to draft, and the latter will see the same four cookie cutter strategies each draft.

The Poison Principle article gives some of the concepts for connecting archetypes (fixing goes a long way), and other examples can be found in Wizards sets.

Something like RTR or GTC was actually fairly rigid because most of the decks were just guild decks, with a peppering of allied guild-combination decks (e.g. Azorious + Izzet). There's no simple solution, but if you have specific examples from your cube design to discuss I would be happy to discuss.

 

[Kyultu]

Thank you for the replies, they're all very thoughtful.

I'm quite honestly not at the point with my own cube where it is stable enough to suffer from the stagnation I'm describing, where week to week the same decks are drafted and played out, maybe with some interchangeable parts (Rakdos Cackler instead of Goblin Guide), difference in strength (damn, didn't see Jace TMS like last time) or drafter/pilot. The best example from my own experience is that every week, without fail, there is a Green Ramp deck. Unless the drafter is completely new, there isn't really a question of whether or not the deck will come together, the archetype is well defined and heavily supported by design. The deck isn't oppressive and can be very fun to play. I'm not worried about the existence of the deck, but the inevitable existence of it every week.

Once my design is tightened up I fully expect other archetypes to become just as defined and persistent. Usual suspects like Red Aggro, Esper Control, Bant Midrange or Flicker, White Aggro, Black GY, and Reanimator/Big Cheats. I'm just not sure were that situation crosses over from being a format with a meta and degenerates into feeling static.

I think the "Well, if a problem comes up, just design an answer" proposal is inherently true and that my concern is probably just a theoretical one without a concrete answer. That said, HOW do we design an answer that goes beyond just neutering or swapping certain builds whenever we get bored? Between reading the Poison Principle and some replies, and doing some more thinking, here's what I can come up with as possible solutions:

1. No solution is necessary. The rate of people getting bored with your cube is significantly lower than the overall rate of change. Expectations are good and your players have better experiences when they know what to expect. Design well and it'll all work out.

2. Vary the pool. The easiest way to change what people consistently draft is to change what they can draft consistently. The most common change is increase the size of the cube. You could also make changes to the contents or order of packs. Maybe seeding certain cards so they only appear in packs 1 or 3. I believe the utility land draft is an example of this idea? Messing with your size and density clearly trades off reliability so that sometimes decks just can't come together no matter how well they are drafted. But, restriction breeds creativity, so it could be a positive thing that your Reanimator deck never finds a fatty and you audible into a Rock Value Recursion deck for the first time. On the Multi-dimensional Venn-diagram above, you aimed for one circle and ended up in an adjacent one on accident. I think there's value in that experience. You just have to watch out for too much homogeneity and repetition between effects/interactions or else the diagram will start to shrink and collapse into a big mushy midrange disaster.

3. High Risk Cards/Archetypes. You know about these. The very high risk/high reward "build-around-me's" as opportunities for players to try and craft a sweet pile. Birthing Pod, Pestermite and Wildfire come to mind. They're really only there for the people who are looking for something different and are willing to try. Personally, I know what the worst card in my Cube is, because it is in there for the day some Johnny opens it up and accepts the challenge. One week it'll be sweet and break up the norm, until then, Kalia of the Vast is just going to go 15th. In this case It's probably just a very poor design idea, but the concept of "hidden" archetypes has strong appeal when I'm worried about "solved" formats.

 

[CML]

Hmm, now this is interesting because over here almost everyone in green wants the mana dorks REAL BAD, so the ramp is usually split up (and when it isn't, it's a pleasant surprise.)

Poison principle, archetype anchors, the necessity of some mono-dimensional cards like Goblin Guide etc. (which are better for design than cards that just go ridiculously well in everything) are well-worn topics here.

Don't make your Cube too big.

Cut the Kaalia-like cards and actively ask your drafters which ones they want out. Three-color cards are just kinda bleh in general.

Ehhh solved formats don't exist in limited as often as they do in constructed. Limited metagames seem to be more durable in this regard, whereas if we just kept playing RTR Standard over and over again results would start to converge. Metagames are nice though, it's what allows MTG to approximate sentience to the point of self-perpetuation (like poker)

 

[Jason Waddell]

I'm quite honestly not at the point with my own cube where it is stable enough to suffer from the stagnation I'm describing, where week to week the same decks are drafted and played out, maybe with some interchangeable parts (Rakdos Cackler instead of Goblin Guide), difference in strength (damn, didn't see Jace TMS like last time) or drafter/pilot. The best example from my own experience is that every week, without fail, there is a Green Ramp deck. Unless the drafter is completely new, there isn't really a question of whether or not the deck will come together, the archetype is well defined and heavily supported by design. The deck isn't oppressive and can be very fun to play. I'm not worried about the existence of the deck, but the inevitable existence of it every week.

I think it's more important to vary your design space than your card space. Playing with > 360 may superficially change the cards that you see from draft to draft without actually giving you greater archetypal variety. You only see 360 cards per draft, and you want to make sure those cards are doing work for you.

To put this into practice, if ramp is something you focus on (as it is in my Eldrazi cube, for example), I would put an emphasis on making sure there are many possible variants to the deck. From my main cube, I've seen BG Rock Recursion ramp decks that combine Primeval Titan + Volrath's Stronghold, GW ramp with tokens and an Elesh Norn top end, an "aggro-Valakut" style GR ramp deck that beasts down while topping off with an Avenger of Zendikar, etc.

Sometimes people, as CML said, pick off the ramp spells without really building a ramp deck. I 3 - 0'd with a deck once that used Growth Spasm to set up Sarkhan the Mad sacrifices. This is a good example of overlapping support. Even if there was always a ramp deck in my main cube (there isn't), it would look different from week to week. Now, the mechanic behind that is the fact that many of the cards that ramp wants are also in demand for other decks. Everybody wants Deathrite Shaman. Pod decks love 1-drop elves. GW beatdown takes Noble Hierarchs. Control takes Farseeks as random acceleration. The chips fall differently every draft, and sometimes and archetype isn't really there because other people have eaten into your space.

With overlapping support though, you can take whatever cards you have and transition somewhere else. It's not like drafting Infect and having to abandon ship entirely. The skill comes in reading the packs and identifying what's open.


As another example, we see Pod decks about 75% of the time, but they're always different. Bant Pod, aggro Pod, Zombie Pod, RUG Pod, etc. Last draft a player built a 4-color Pod list that combined with Volrath's Stronghold, to sacrifice creatures with Pod and put them back into the library with Stronghold. It was genius, I had never thought of that before.

The cube doesn't get boring because the archetype overlaps are structured in such a way that there are dozens of successful decks people can make without feeling like they are building a cookie-cutter archetype.

 

[Jon Metzger]

This general idea is the most interesting thing about designing a cube to me. From reading around here and thinking about my new cube pretty obsessively for the last week or two, I'd like to propose a way to think about cards with 3 roles for cube:
1) Good cards - things like Ponder, Lightning Bolt, Hero of Bladehold - cards that basically any deck in those colors will want. Obviously some decks will want them more or less (aggro wants Bolt the most, but control is happy to play it as removal, Hero is amazing in tokens, but a fine single card finisher for any deck), but these are just good cards that will be picked highly and make any deck playing them better. These are important because they're fun, sweet cards to play with, but also so that drafting doesn't turn into drafting the same rigid decks over and over.
2) Build-around cards - Birthing Pod, Runechanter's Pike, Splinterfright, Goblin Guide - these are the incentive cards that you see and you try to draft a build-around deck with. This is many people's most fun thing to do in conventional draft, and something that's easier to support in cube with a less randomized environment. Drafting one of these cards is like picking out the title for the short story you're going to write.
3) Support cards - these are the cards that are lower power in general but can be all-stars in the right deck. Intangible Virtue isn't a card that you pick normally, but if you already have several good token cards, you might pick it over a substantially better card. These are designed so that they can wheel to the right deck; they may be the 23rd card or a cut in the wrong deck (you don't want them to be completely unplayable) but they'll be stars in the right deck.
4) Bad cards - these are in normal draft sets to help people learn card evaluation and make drafting easier (having a totally flat power level is bad), but I think that cubes generally will eliminate this category in favor of Support cars and Good cards.

One thing I noticed about Modern Masters is that a lot of the Support Cards are off-color. For instance Mothdust Changeling is generally a Faerie, but you can also see it in the more rare UW or UB Rebels deck if that's the way the cards split. The Affinity deck was UW in the design articles that talked about MM, but it often ended up being UR, WR, or UWr in practice. I think the "art" and replayability of a Cube is helped by finding Support cards that can support multiple archetypes and that suggest taking archetypes in different color directions.*

A few examples from my cube to illustrate my thought process here:
- Swords to Plowshares - this is just a Good card. Better in control than aggro, but it's mainly in there to make sure that quality removal is available. Should be an early pick for any deck in white.
- Crib Swap - Support card. The player who picks up Haakon should be able to count on this wheeling to her.
- Fling - Off-color Support card. A red deck will possibly play this as a late pick, but it might also get the player who's drafting the GB Lhurgoyfs to splash R for the "Fling 20 at your face" instead of the U they would usually splash for Tracker's Instincts/Armored Skaab.

*Having good fixing is important for this - you need to be able to be confident that you can switch your second color or support a splash.