Every year, I write a column examining the previous year's worth of designs. I call the column "State of Design," as it's modeled after the annual speech given by the US president about their take on the state of the nation. This is my twelfth "State of Design" column. You can see the first eleven below:

This past year was the first year of our new Two-Block Model, where we started doing two blocks each year, and as such, I will for the first time be changing up how I structure this column. In order to be able to focus on each block, I'm dividing my column into three sections: one for overall Magic design to look at the year as a whole, one for Battle for Zendikar block, and one for Shadows over Innistrad block. For each section, I am going to go over what I felt the highlights of the block were and then talk about the lessons I learned from them.

A few quick caveats. One, this is a column primarily about design, so I'm going to be focusing on how I feel design went. There are a lot of other wonderful departments doing very strong work. My commentary is not focused on any of them but rather on the part of the process that my team and I are responsible for. Second, there is more design than just the four main sets, but due to space restraints I'm not going to be able to talk about them today.

Before I get into the blocks, I always start by asking: How was the last year of Magic design? It was a transitional year. We were changing over to an entirely new model, and that transition proved more difficult than I think we realized at the time. There were plenty of highlights, as I'll get into in a second, but I think this last year to me was more defined as a year of important lessons. We're on the verge of another big step up in Magic design technology, and this last year has been a valuable learning year.

That said, let's dive into the last year of Magic design.

Overall Magic Design

HIGHLIGHTS

  • Two blocks a year is a success

I wrote an article two years ago about how Magic was going to have a metamorphosis. Now that we're on the other side, we get to start looking back on the many changes that occurred. Let's start with the biggest one. Magic shifted away from its "one three-set block plus core set" yearly model to one with two two-set blocks and no core set. While there have been lots of issues on our end making the change, the public response has been almost exclusively positive. Two blocks a year is just more fun than one.

The concentration of three sets into two has allowed us to focus our efforts and make stronger designs. I also believe we're getting out before players are growing tired of the block, something that often didn't happen with three-set blocks. There was also a concern that players wouldn't treat the spring set like a new block, as that wasn't the pattern they were used to, but you all adapted easily and Shadows over Innistrad definitely felt like a new block and not just another spring set.

  • We're getting better with story integration

Another big change was how we were handling story. The goal was twofold. First, we wanted more continuity. We wanted a sense that the story was more ongoing and connected. To do that, we created a stable cast of characters (the Gatewatch) and have started interlinking the story of the blocks more closely together. Second, we wanted the story more integrated into the cards and the design. We needed the story to be part of the gameplay, to allow players to experience elements of the story as they played Magic. This was accomplished in numerous ways. From a design standpoint, we worked hard to make sure that what you, the player, were doing when you played with the set dovetailed into what the characters were doing in the story.

Zendikar Resurgent
Zendikar Resurgent | Art by Chris Rallis

This integration was also very successful. I have heard from many players that for the first time in years of playing they are actually aware of what the story is, and they're enjoying how playing with the cards is allowing them to understand, loosely, what's going on. There's still lots of room for growth in this category, but I'm happy with how much we've accomplished for our first year in.

LESSONS

  • Complexity is too high

Battle for Zendikar block and Shadows over Innistrad block were both above average in complexity, especially at common. Now, some sets get to be above average, but as this is the third block in a row to have been quite a bit above average, it's an issue we need to concern ourselves with. I know a lot of people worry whenever I talk about lowering complexity, as they think it means lessening strategic depth, but the relationship between those two things is actually much more detailed.

I often talk about the importance of lenticular design—creating things that appear simple on the surface but become deeper as you're able to understand more facets of the game. In the past, I've focused lenticular design more on individual card design, but in the last year I've spent a lot more time thinking about it on a larger holistic level, looking at how whole blocks can be more lenticular in nature.

The takeaway here is that there are many answers out there for us to explore, but the current execution of creating mechanical layering through the piling on of mechanics is fundamentally a flawed one. Battle for Zendikar block, in particular, suffered greatly from this.

  • We're making too many named keyword/ability word mechanics

This lesson is an offshoot of the previous one. Part of what is raising the complexity of the sets is the sheer number of named mechanics we've been including in them. I'll get into more detail below as I talk about the individual blocks, but this was a global lesson that applied to both blocks this year. We could have easily removed at least one mechanic per set and I don't think anyone would have noticed it missing.

  • There's too much change between the big and small sets

I believe that there wants to be some change between the large set of the block and the small one that follows it. I also believe both Oath of the Gatewatch and Eldritch Moon changed more than was necessary from the large block that preceded them. That's not to say that both small sets don't have fun designs (they very much do), but as the head designer looking back, it's clear to me that we could have made the small sets just as exciting without having to change so much. This is more an internal issue than anything else, because what I'm fundamentally saying is we could have made the small sets just as much fun for all of you without using up valuable, finite resources on our end.

So in a nutshell, we're adapting well to our new model of making Magic sets, but we need to be more careful about what and how much we put into them.

Battle for Zendikar Block

HIGHLIGHTS

  • We had a couple hit mechanics

Landfall was much beloved when it premiered in original Zendikar, and our research shows it was beloved again on its return. (The one negative note I did hear was more enfranchised players disliked that we ratcheted down the power level of landfall. This is more a developmental note than a design one, but we were addressing a lot of the criticisms of how landfall played in original Zendikar.) Awaken from Battle for Zendikar and surge from Oath of the Gatewatch were new hits. I expect to see them both return one day. Colorless mana, while not quite a keyword, also stirred up a lot of excitement.

  • You still love full-art lands

The full-art lands started in the Un- sets but made their way into expansion sets with original Zendikar, where they were a big hit. Nothing has changed. They were just as beloved today as they were in the first Zendikar. (I have loudly heard the note, by the way, that you're eager to see full-art lands of worlds other than Zendikar.)

  • Zendikar Expeditions excited a lot of players

We tried something new in Battle for Zendikar block with having older cards in new form show up at a rarity higher than mythic rare. The vast majority of you were quite excited by them.

Godless Shrine
Godless Shrine | Art by Noah Bradley
  • We have begun embracing card-frame components and symbols as an important tool in design

Both devoid and the new colorless mana symbol were key mechanical aspects of the design that leaned heavily on visual components. As we seek out new design veins to explore, having visuals as a tool to help us has proven a necessity. A special thanks to Liz Leo for helping find the visual solutions needed to make these mechanics work.

  • We pushed through taboo areas to find new design space

Magic design has to constantly be on the lookout for new design space. As I mentioned above, part of that is finding new tools. Another part is being willing to revisit areas you have previously made off limits. Battle for Zendikar block (with ingest and Processors) found new space by messing with an area, the exile zone, that I had been very reluctant to explore—but I was happy that we were able to find new design area without undermining the reasons I had wanted to keep out of exile in the first place. This willingness to re-examine old decisions is an important one as Magic design moves to the next level.

LESSONS

  • We focused on the wrong part of Zendikar

I talked about this lesson in a column earlier this year. Players loved the plane of Zendikar, so we returned to let them revisit it, but in doing so we focused on the wrong aspects. If you look at all the market research, the thing that players loved most about Zendikar was its adventure world feel. It had a sense of old-school Dungeons & Dragons and Indiana Jones and Flash Gordon and a whole genre of pop culture.

Then to justify shifting the final set mechanically, we introduced a brand-new villain, the Eldrazi. While Rise of the Eldrazi is flavorfully remembered for its deep and interesting Draft environment, the set itself got only a lukewarm reception by players. So in choosing to return to Zendikar, we picked up the cliffhanger of the original block and then focused on the aspect of the plane that the players liked less. Battle for Zendikar was too much Rise of the Eldrazi and not nearly enough Zendikar. We plan to one day return to Zendikar for a third visit, and I plan to not repeat that mistake next time.

  • There were too many Eldrazi

Some things are good in quantity and some things are not. I like the Eldrazi as a villain, but I think they work best in small doses. Giant, unfathomable world-eating aliens are cool characters, but they're very difficult to make compelling cards out of when you have to make more than a handful. A lot of the problems we had in design centered around trying to solve issues (how to make a set filled with Eldrazi) that we shouldn't have ever had in the first place. I should note that the idea to focus the block on a war between the Eldrazi and the Zendikari was my idea, so I'm trying to own up to the fact that, at the core of this design, I made a fundamental mistake.

  • Players don't like "marker" mechanics

One of the challenges with bringing back the Eldrazi was finding a way to mechanically reference them. The creatures all shared the Eldrazi creature type, but there were also a lot of spells we wanted flavored as Eldrazi. Last time around we used the tribal card type, but as we've chosen not to use that any more, we needed to find a new answer.

That answer was colorlessness. The problem was we didn't want to usurp the color pie for developmental reasons, so we came up with devoid. Originally, devoid wasn't supposed to even be a mechanic. The color indicator, frame, and mana cost were supposed to convey how the cards worked. Playtesting showed it didn't, and we ended up creating the devoid mechanic to make it clear.

Here's the interesting thing. Players enjoyed the "colorless matters" theme, but they didn't like devoid, the thing that made the theme possible. I spent a lot of time digging into it, and the problem boiled down to one issue—devoid didn't do anything. It was what we call a "marker" mechanic, in that its role was mostly to mark a subset of cards so the set could mechanically care about them. The problem was it didn't match the expectation of what a mechanic is.

Interestingly, I believe the correct answer would have been to remove the devoid keyword and just write out what was the reminder text (tweaked slightly for rules purposes) on each card. This would have kept the public from thinking of it as a mechanic and thus removing the "it doesn't do anything" issue. This is a very important lesson moving forward because as Magic slowly uses up every element of the card already naturally marked, we have to figure out how to create "marker" technology that doesn't run into the "it doesn't do anything" problem.

  • The colorless mana symbol should have started in Battle for Zendikar

We made a conscious decision to hold back the introduction of the colorless mana symbol until the set that made it mechanically relevant came out. The result of this decision is that we had cards that looked different but functioned similarly in the same Limited environment. I believe that decision was a mistake, and was one of the contributing factors that made the new colorless mana symbol harder to learn than it needed to be. In the future, we should start new ongoing things for Magic (that is, things we plan to introduce and then keep doing in every set) at the beginning of a block and not partway through.

  • Our hit rate for mechanics was too low

Processing/exile-matters, ingest, rally, and devoid from Battle for Zendikar all did okay in market research, but our goals are to do better than "okay." Converge from Battle for Zendikar and support and cohort from Oath of the Gatewatch did poorly. We definitely had some successes, as I listed above, but overall, our hit rate for the mechanics in this block was lower than I like to see. As design is primarily responsible for mechanics, I take full blame for this. I've spent a lot of time looking back at each of the mechanics and all the feedback we've gotten on them to understand where there was room for improvement. Design is committed to keeping the average quality of our mechanics high.

Shadows over Innistrad block

HIGHLIGHTS

  • We did Eldrazi correctly

It is liberating, when you do something wrong, to then watch yourself do it right. Shadows over Innistrad block found the proper way to use Eldrazi effectively. The first set was a mystery where the Eldrazi were just hinted at, and then Eldritch Moon found a way to integrate the Eldrazi in low numbers that reinforced their feel and was something we could execute cleanly and efficiently on cards.

  • We focused on the right part of original Innistrad block

Shadows over Innistrad got right another thing Battle for Zendikar got wrong. Original Innistrad block also took a sharp turn in its third set to allow a mechanical reset. But instead of making this block return to Avacyn Restored, we correctly made it a return to Innistrad. Players wanted to see monsters running amok, not righteous Angels cleaning up the world.

Conduit of Emrakul
Conduit of Emrakul | Art by Raymond Swanland
  • The mechanics on the whole were very well received

Double-faced cards were the breakout hit of original Innistrad block, and their return was equally beloved. I feel we both used transformation in cool new ways and came up with an innovation for double-faced technology—the meld mechanic. The numbers are still coming in, but it's clear that meld is a home-run hit. (The biggest complaint about meld? Not enough of it.) Other hits for mechanics were investigate (more on this in a second), emerge, and the return of the madness mechanic. All in all, our hit rate for mechanics in Shadows over Innistrad block was where I like to see it.

  • We nailed the top-down design

Innistrad is currently battling with Ravnica for first place as most-beloved plane. I think a lot of Innistrad's success has come from how evocative the cards and mechanics are of the Gothic horror world we've created. Cards like Triskaidekaphobia or Harmless Offering or even Emrakul herself have done a wonderful job of marrying cool flavor with evocative gameplay. One only needed to watch the finals of Pro Tour Eldritch Moon to see the climax played out on the battlefield. (And yes, Liliana and Gideon did come together to help defeat Emrakul.) Top-down design is tricky, but I feel part of what made Shadows over Innistrad block such a success was design (and development) being on our A game.

LESSONS

  • We shouldn't have removed investigate

This lesson is really the culmination of a number of lessons from above. I think we're making too many mechanics and also making too big a change between the big set and the small set that follows it. Players really liked the investigate mechanic. It was flavorful and played well. We removed it from Eldritch Moon because we were adding three new mechanics—meld, emerge, and escalate—and we simply didn't have room for it. We rationalized removing it because the story had moved past the mystery part, which was what investigate was flavorfully conveying.

Looking back, I wish we had kept investigate and not introduced escalate. Meld and emerge were important, as they were conveying the influence of Emrakul, the major theme of the set, but escalate was more there to introduce another new thing. I would rather have held back on escalate for a set where it had more meaning and synergy and simply continued to use investigate. I think there was plenty of simple design space left with room for a few tweaks. I think this would have made players happier and created less complexity.

In general, moving forward, I am questioning removing mechanics in the large set from the small set. The majority of the times we've done it we've created more dissatisfaction than happiness. This dovetails with my desire to lower the number of new mechanics we're introducing overall per block.

  • We brought back Eldrazi too quickly

The rationale behind bringing back Emrakul was a solid one. We wanted to convey that we were changing-up how story was being told, and we really wanted Shadows over Innistrad to feel like a continuation of the story and not just a completely separate tale unrelated to the block that preceded it. That said, I feel some of the ill will created by Battle for Zendikar block got brought to Shadows over Innistrad block. Rather than just excitement, I feel the reveal of Emrakul being on Innistrad created some disappointment. Now, I believe we did a great job of execution and won over most of the cynics, but it's hard not to look back with a critical eye and question if there was supposed to be a block or two between Battle for Zendikar and Shadows over Innistrad.

Wrapping Things Up

To sum up this year, I believe we made a change that was bigger than we realized. The act of transitioning thus proved more difficult and the first block out of the gate, Battle for Zendikar block, was not as jelled a design as I normally like to hand over. Development, led by Erik Lauer, took on a Herculean task to get it playable, but a lot of the flaws of the block were already baked in. Shadows over Innistrad block then managed to fix a number of the mistakes of Battle for Zendikar block, but still fell into the trap of overstuffing the block and creating more mechanical flux between sets than was needed.

As I look forward, I believe the upgrade in design execution from Shadows over Innistrad block to Kaladesh block is as big as, if not bigger than, the change from Battle for Zendikar block to Shadows over Innistrad block. I'm quite excited to be here a year from now reviewing how you all received Kaladesh.

As always with my State of Design columns, I am now eager to hear what you all thought of the last year of design (and of my take on it). Please send me an email telling me your thoughts or drop me a note in any of my social media accounts (Twitter, Tumblr, Google+, and Instagram).

Join me next week when I finally (finally!) get to start talking about Kaladesh.

Until then, may you have enjoyed this last year of design as much as my design team and I enjoyed making it for you.

"Drive to Work #360—Tweaking"

This podcast talks about how we tweak old cards in design to make new cards that fill a similar function but are just a little different.

"Drive to Work #361—Twenty Lessons: Emotional Response"

This is part six of my 20-part series "Twenty Years, Twenty Lessons" where I cover the lessons from my GDC talk about the many things I've learned from designing the same game for 20 years. This podcast I talk about how to not confuse "interesting" with "fun."