Last week, I started a new series called "Lessons Learned" based on a series I do on my Drive to Work podcast where I talk through lessons I've learned from sets for which I led or co-led design or vision design. For each set, I explain the main lesson I learned from designing the product. I got through Unhinged last week, so that means today I'm beginning with the original Ravnica.

Ravnica: City of Guilds (October 2005)

Lesson: "Structure should follow theme."

Ravnica was the first set I led after becoming head designer. As I mentioned last week, I was focused on the idea of block design, that is, planning out each set of the Magic year before finalizing details for the first large set. The design of the first large set of the year depends on the role that each set plays in the bigger picture. In other words, each block had to have a preplanned structure. Now, saying that is one thing, acting on it was a bit more difficult.

Ravnica started with the goal of being a multicolor block, but one different from Invasion, our previous multicolor block. Invasion was about playing as many colors as you could, so I decided Ravnica should be about playing as few as you could, meaning two, the smallest multicolor combination possible. I also decided to treat ally and enemy-color pairs the same structurally. Up to that point, we tended to favor ally-color pairs because we were trying to capture the flavor that the ally colors like working together more. (This, by the way, is why there are more cards with ally colors than enemy colors.) I just felt strongly Magic was more fun if you had additional options, and that the flavor hit wasn't too big of a deal.

Anyway, our first Ravnica playtest had all ten two-color pairs, plus hybrid mana, this new thing I'd created that I wanted to find a home for. After the playtest, R&D member Henry Stern came up to me. He said, "I made the top four in back-to-back World Championships, and this was too hard for me to do. I lost track of how many different piles I had to make." I then talked to other people in the playtest and discovered that everyone felt similarly. Meanwhile, Brady Dommermuth, the creative director at the time, was tackling how to flavorfully capture the ten two-color pairs. While exercising one day, he came up with the idea of a city plane filled with guilds.

Now we get to the big lesson of Ravnica. I loved Brady's guild concept and recognized that I had to address the complexity issue that Henry had raised. That's when the idea hit me. What if we doled out the guilds over the course of the block? With a large-small-small block structure, I pitched four, three, and three. My plan was to embrace the set's theme of guilds and use a structure that reinforced that theme. At the time, it was a radical idea. People kept suggesting that we should have a little bit of each color combination in every set, but I held firm. If we were going to make the block work, we had to have a clear and focused message for it. This simplified each individual set and gave the block a clearer structure. Ravnica went on to become one of the most successful sets and planes we'd ever created and probably the most influential set on future design.

Future Sight

Lesson: "Understand your audience."

The Ravnica block was my first block design as head designer. I used what I called the "pie method" where I took ten themes and chopped them up between the three sets. Time Spiral was my second block, so I wanted to try a different method. The block had a time theme (we had new mechanics suspend and split second and introduced flash as an evergreen keyword), so I was looking for a way to structure the block that matched the theme. I was interested in a structure that could look at the same theme but in different ways. How do you divide time into three parts? The answer seemed obvious—past, present, and future.

The past part was easy. The set could just do a lot of things that were reminiscent of Magic's past. The idea of the timeshifted sheet (i.e., a bonus sheet) that had old cards with the old frame arose from us exploring the past as a theme. The present and future were a bit trickier. For the present, I came up with the idea of an alternate-reality present where Magic had made some different choices in the past, both mechanically and creatively, and we showed what Magic would look like had those choices been made. The future, though, seemed trickiest of all, so I chose to lead the design for it. (Brian Tinsman led Time Spiral design, and Bill Rose led Planar Chaos design.)

In the end, we decided to have Future Sight give players a peek at possible futures. We made an entire timeshifted sheet filled with mechanical and creative elements that Magic hadn't done yet. Part of the design included doing mix-and-match cards, where we took mechanics that we'd done and combined them in ways we hadn't before. All of this resulted in the set having almost as many keywords as had existed in all of Magic before the set came out.

The Time Spiral block, but Future Sight even more so, did something we'd never seen before. Organized play, including all the sanctioned tournaments, was up, but sales were down. These two signifiers had always gone hand in hand before. If a set was played a lot in sanctioned play, it sold well. For the first time ever, these two forces went in opposite directions.

At first, we were baffled, but something had to be causing it, so we figured it out. The players who play the most are those we call "enfranchised players." These are the people most invested in the game. Many participate in sanctioned play and make up a lot of the online community. They are the most dedicated players and the ones with the highest visibility. That group adored Time Spiral. It was a love letter to the game, and they soaked it up. There were a lot of old mechanics being used, but they already knew what they were, so it was fun for them.

The other group we dubbed "the invisibles." These are people who play Magic but lack visibility. They don't show up much in sanctioned play, they don't take our site surveys, and they don't frequent parts of the internet that we can observe. Our only chance to learn about this group is doing what we call a deep-dive, where we ask questions to the public at large. Deep-dives are hard to do and take a lot of resources, so we do them less frequently than most other market research.

What we realized was that "the invisibles" were a big group. Yes, they buy far less on average than the enfranchised players, but they represent a much bigger pool of the audience. The invisibles were turned off by the Time Spiral block. It was too confusing. There were way too many mechanics, and the themes required a lot of knowledge of existing cards. "This card is a combination of two cards you don't know. This card is an alternate-reality version of a card you've never played. This card hints at something we've never done, but you don't know we've never done it." The block was just too much for them, so they checked out.

So, the lesson here was about understanding all audiences. It's very easy to focus on the audience you understand best, but that doesn't result in a product that makes the most people happy. The offshoot of this lesson is that there's a time and place to focus on a subset of the audience—supplemental products. That is why we made Modern Horizons. There was an audience that adored the Time Spiral block, so we catered a product specifically to them, but our run-of-the-mill premier set must keep the larger audience in mind.


Lesson: "You can overdo a theme."

I came up with hybrid mana during Ravnica design. I was enamored with it. Where traditional multicolor cards are "and," hybrid cards are "or." It just played in unique design space and created different synergies. There was a lot going on in the Ravnica block, though, so we chose to just sprinkle in hybrid mana. There was one vertical cycle (a common, an uncommon, and a rare) for each of the guilds. It was brand new, so using it for splash value made a lot of sense. My plan was to bring it back one day where it could play a larger role in a set.

Flash-forward a year. There was a desire to have a fourth premier, non-core set for the year, so I came up with the idea of two mini-blocks, each with a large set and a small set. They would represent two aspects of a plane that went through a change. Lorwyn, the first mini-block, had a creature-type theme. Shadowmoor, the second mini-block, needed a theme that was synergistic with Lorwyn, but its own theme. What was another trait like creature type that could matter in one mini-block and then appear in the other? How about color? Most of the Lorwyn cards would be of one or more colors. We could make a set about colors mattering, and I had just the mechanic to use.

When I started the design, my goal was to have as much hybrid as I could in the set. We ended up choosing half. I felt that was a bold statement. In fact, when we premiered the set, we first showed the audience cards from a sample booster where none of the cards had any rules text. You could just see the names, mana values, art, and frames. That boldly demonstrated what the set was about.

The lesson from this set: you can have too much of a good thing. I think hybrid did a great job of being the center of a "colors matter" set, but we erred on the side of using too much. It forced us to do more color bends than we should have, and I think the set suffered for some of the constraints it put on us. I do think my goal of "as much hybrid as we can" was fine. Half was just too much, and we should have figured that out before releasing the set.


Lesson: "Understand the audience's perspective."

The lesson of Eventide is an offshoot of the lesson of Shadowmoor. Because we chose to make half the set hybrid, we ended up using all the design space for ally-color hybrid. The solution to that problem was to make Eventide enemy-color hybrid. The Invasion block had divided the sets into ally color (Invasion and Planeshift) and enemy color (Apocalypse), and that had gone over well, so this seemed like the natural solution.

The problem was that the two had to be drafted together. The Invasion block was focused on playing as many colors as possible, so being able to get ally and enemy-color combinations into the same deck wasn't difficult. The block instructed players to pick up as many sources of different colors of mana as they could. At the time, I wasn't worried about this. In my mind, the Shadowmoor block was about monocolor play. In fact, in all of Magic, it's the set that made it the easiest to choose to play just a single color (defined as using only one basic land type for your mana). You pick your color in your first Shadowmoor booster and just keep drafting that color. If you pick black, for instance, you're drafting blue-black and black-red hybrid cards in the Shadowmoor boosters, and then drafting white-black and black-green hybrid cards in the Eventide boosters. Simple.

Here's the problem. I was thinking like a designer. Yeah, at its core, the Shadowmoor block draft theme was about monocolor, but that was something players had to learn as they drafted. The messaging of Shadowmoor was that of "it's an ally-color set," so players drafted ally-color decks. The messaging of Eventide was that of "it's an enemy-color set," so players drafted enemy-color decks. Those two things don't naturally work together, especially without mana support, which wasn't there.

Yes, some players figured it out, but a lot didn't. They tried to follow along with what the messaging of the sets told them and ended up with unplayable soup. That's on us. As I explained last time, it's the job of the game designer to properly message to the players what the set is asking of them. In retrospect, I think Shadowmoor and Eventide would have been helped by the hybrid volume being lower and the messaging of the monocolor theme being louder.


Lesson: "Reward the players for doing what they want to do."

Zendikar came about because I was obsessed with the idea of having a set based around land mechanics. It was a theme I realized we'd never done before, and I had faith we could do something cool with it. Few others in R&D shared this belief, though. Randy Buehler, who was my boss at the time, put it at the very end of a five-year plan I'd laid out, which ended up becoming a seven-year plan, with "Landsapalooza," as I'd jokingly named it, in the seventh year. Then, when it finally came time to make the set, Bill Rose gave me and my team three months to prove the theme. If we couldn't do it to his satisfaction, the set theme had to change. Bill did not normally do this, so it was clear he was nervous about it.

We spent most of those first three months (this was back when it was design and development and design had a year) working on land mechanics. I'm not exaggerating when I say I think we came up with 40-plus different mechanics. One of the ones we spent a lot of time on was a mechanic where you could spend your land drop as a resource on other things. For example, there was a creature that you could use your land drop to put a +1/+1 counter on. We were very intrigued by the concept that your land drop could have other functionality.

The problem came when we playtested the mechanic. Strategically, the correct play most of the time was to use your land drop for land when you had it and use it for other things when you didn't. This way, it was a resource you could make use of every turn. That's not how our playtesters played it. They were lured into doing other things with their land drop and would do those over playing land. They would then get mana hosed and lose. This obviously wasn't much fun.

We then came up with what was an out-of-the-box idea at the time. What if we turned the mechanic on its ear? What if instead of getting a reward for not playing a land, you got one for playing a land? This was how landfall came to be. The worry at the time was that we weren't really using a resource. The player was going to play the land anyway, so why were we rewarding players for it? There was a lot of skepticism, but we figured it was worth a playtest.

Unlike the last playtest, this one went fantastically well. Getting rewarded for something you already wanted to do felt amazing. It was like you were getting something for free, and all the playtesters adored it. That's where this lesson comes from. Our goal as game designers is to make players enjoy the game. Not everything has to be a mental strain where players must weigh options. Yes, there's a time and place for tension and hard decisions, but we can also have mechanics that just reward players for doing what they already want to do. This was a big revelation for me personally that really made me rethink how I approached Magic design. I can't just think about how something makes the players think, I also need to examine how it makes them feel. The key to a good design is that players enjoy playing it. An important part of meeting that metric is having a better understanding of what you're asking of your players and how it impacts their emotional approach when playing.

Live and Learn

I've hit my 3,000 words for today. Part three will be coming, but not right away. I'll continue this series when I find opportunities to do so. As always, I'm curious to hear your thoughts on today's column and any of the lessons. You can email me or contact me through my social media accounts (Twitter, Tumblr, Instagram, and TikTok).

Join me next week for the start of March of the Machine previews.

Until then, may every act have an element of education.