Thursday, March 12, 2015

Back-Forward Design

I wrote this post last year, before Wizards announced the big shift to a two-set block structure. I'm not sure whether I forgot about it after seeing the announcement, or because I wanted to think of a better example for the conclusion. It's no longer terribly relevant to Magic, but I'll share it now for the perspective it shares on Magic and game design in general. And I won't delay it further by looking for a better example.

July 17, 2014…

Wizards has gotten better and better at block-planning. They look years ahead to know what kinds of blocks are coming. They begin development of one block before finishing the last one, and use that information to interweave synergies that make those blocks more than just circumstantial neighbors. Within a block, advanced design gives them a better idea of what will work across the block and what won't, so they can hit the ground running and not run out of gas. They've also been playing with block structure a lot, finding new ways to keep each expansion relevant throughout the year.

I've got one more idea.

Wizards grants priority to the set closest to release over sets that will come out later, often borrowing cards or mechanics from upcoming sets and even other blocks to ensure it's the best it can be. The first set of each block also gets more time and love because it's such a big deal; It's a large set and if it turns players off, most of them won't be back for a while. The second set is an evolution of the first set, usually directly, sometimes indirectly like Gatecrash. The third set is a further evolution (again, often supplementary like Journey into Nyx, or complementary like Avacyn's Rise). In any case, that last set is hard-pressed to deliver on the block's existing expectations, while still innovating, without venturing into inelegant design space.

This is not an unreasonable model. It's a bit like filling a jar with stones, then filling the stones' gaps with sand, and filling the invisible gaps of the sand with water. But I don't think it's the only way. With advanced design, R&D has started thinking ahead about the needs of the last set, and setting mechanics aside for it. What if we took that one step further?

Consider the novel. There are many ways to write a novel, but if it doesn't end in a satisfying way, the whole thing tends to be regarded as a failure. It's not uncommon for an author to figure out the story's premise and then determine how it will end, before then going back and writing the middle of the book to beg that conclusion and then writing the beginning of the book to feed that core. Working backwards is a surefire way to end with a bang. Imagine a Magic block where the team finds some epic way for the set to end, something really exciting, unique, and—this is important—that you couldn't do in a first or second set of a block. Then use the first two sets to set the stage for that ending, to make it possible.

The first set will never be a footnote to the last set, so it's not just setting up that brilliant climax. It still has to set the tone for the block, drive excitement for a new idea, be fun to draft triple boosters for 14 weeks, and feel like Magic. That it's acting as a building block doesn't negate any of that. Nor does it reduce the set's quality. Think of all the good books you read. The first third establishes the setting and conflict, it always paves the way for the rest of the book. But if it's not enthralling on its own merits, you won't make it to the end of the book… and it wouldn't be a good book.

This is all pretty abstract. I'll take a light stab at an example, just to get a little more perspective on the idea. Suppose we brainstorm a really cool mechanic that we just can't print without a whole lot of support, both in terms of other cards to make it relevant to the format, and in terms of layered concepts. Say we've explored a generic-mana-cost-matters theme that has a lot of merit (however unlikely that seems), but we're finding that it requires dozens of other cards in the set to really impact Limited, and that casual/new players are having a lot of trouble differentiating between colorless mana (that you produce/spend) and generic mana (in costs that you pay) or between the terms "converted mana cost" and "generic requirement in its mana cost. We can build the first two sets to include enough spells with very specific generic components to their mana costs and to include a few cards or a mechanic that get players thinking about the different parts of a mana cost, like devotion.

(Editor's note: This example was written long before Battle for Zendikar was announced, but isn't referring to the colorless mana symbol {C}.)

Those considerations won't likely define the first or second sets, but they serve as an additional parameter to build them around. Maybe the first set brings back devotion as a major or minor element of its thematic basis. Maybe there are a lot more or a lot fewer CC and CCC spells, the idea of which lingers barely noticed until the third set reveals how relevant that is, and everyone goes "Oh. They planned this." Not an inspiring example, I know, but hopefully enough to convey the basic idea without distracting from it.


  1. Some blocks have been ever so slightly like this-- Alara and Scars block come to mind. I don't think all-multicolor or Phyrexian-world could have been done as a first set.

    1. Also, I think this is an awesome idea, assuming that there exists a great concept for the third set.