Lessons Learned, Part 9
I write this weekly column and answer questions on my blog, Blogatog. In addition, I publish a weekly podcast called Drive to Work, which I often record while driving to work. Each episode is about 30 minutes long (that's how long it takes to get to work) and I post two episodes a week. One of my recurring series on Drive to Work is called "Lessons Learned" where I take a set I led or co-led and walk through all the things I learned while leading its design. In 2023, I started writing article versions of those episodes. Today, it's time for part nine. Here are the previous eight "Lessons Learned" articles:
Last time, I talked about Phyrexia: All Will Be One and March of the Machine. This time, I'm starting with The Lost Caverns of Ixalan.
The Lost Caverns of Ixalan
Lesson: "The mechanical theme has to serve the needs of the set."
At the time of writing this article, I've led or co-led the design of almost 50 sets. I believe The Lost Caverns of Ixalan was the set where I most lost my way. Every set evolves from what's handed off by Vision Design. But of all the sets I've led, no set has changed as much as this one. There were a number of external reasons for this. The biggest reason was a decision made by the Creative team after the set was handed off. Originally, the set was going to take place on a new plane, but late in the process, far later than normal, the decision was made to change the setting to Ixalan, a known plane.
The whole point of doing these "Lessons Learned" articles and episodes is to take an honest look back at what I did with the design. The hard truth is that this change happened primarily due to the set's vision design. It was trying to accomplish something that, in the end, didn't service the needs of the set. Fundamentally, the entire point of vision design is to determine the core identity of a set. While there are definitely pieces of Vision Design's work in the final set, The Lost Caverns of Ixalan didn't establish its vision properly. And that's on me, the lead vision designer of the set.
We'd talked about doing an underground setting for years. There's ample trope space to work with, and we thought we could create a lot of cool visuals. When we officially greenlit an underground setting, I was excited to lead the vision design. As I explained in my preview articles (Part 1 and Part 2), there were a bunch of different paths we could take when doing an underground world. The first big fork was the genre. An underground setting could be done as an adventure or horror setting. Magic had done a lot of horror in sets, so I was excited to explore the adventure path. The next fork was generational. There have been two big executions of underground settings in gaming. One of those was based on the dungeon crawls from role-playing games. The other was a resource-acquisition video game. At the time, the latter had been less explored and skewed younger, and Magic had been looking for ways to attract a younger audience, so we chose to take the latter path.
I don't think any of those decisions were incorrect, but once we committed to going underground, I became convinced that this setting would be an ideal place for a theme I'd been trying to get back into the game for years: "color matters." The color pie is an organic part of the game. When Richard Garfield created Magic, color was a very important mechanical element. There are early mechanics that specifically reference colors and land types (protection, fear, landwalk, and landhome). There are whole cycles of cards that negatively affect enemy colors. There are cycles that reward you for playing certain colors. Color is embedded in the game, and that is even more true the further back you look. Over the years, a lot of that has been weeded out; we found caring about color to be too swingy. Either it didn't matter at all or it dominated games.
The Shadowmoor block leaned heavily on hybrid mana. As a side effect of that, many cards had to be multicolor. That enabled a mini-block where color mattered mechanically. (The Shadowmoor block famously allowed you to consistently draft monocolor decks.) I had great fun with it and wanted to revisit the theme again, but as the years rolled by, I couldn't find a set that wanted it. That is, until we began vision design for The Lost Caverns of Ixalan.
In the second Great Designer Search, we had each of the finalists build their own world, and two different finalists built underground worlds. One of those underground worlds, created by Jonathon Loucks, had a theme of light and dark. Intrigued by that idea, I began the vision design process with an interest in exploring color as an execution of light.
As we started getting into the design, I found a better use for color. The set became about finding resources and using those resources to craft powerful magic items. The key was to find a way to create "recipes" or allow players to create new items by combining existing ones in a specific pattern. I was enamored with those items being actual gems (as they often are in video games). That led us to remember that Magic already had five iconic gems: the Moxen.
You would collect tokens of five different colors—opals for white, sapphires for blue, jet for black, rubies for red, and emeralds for green. Because we wanted them to have a secondary function, we chose the low-hanging fruit. They were just like Treasure tokens except they only produced one color of mana. This meant you could use them for your craft recipes or to cast spells.
The Vision Design team also leaned into a "color matters" theme. There was a mechanic named "illuminate" that cared about how many different colors you had among your permanents. (This would see print as vivid in Lorwyn Eclipsed.) We used twobrid costs (spending a mana of a certain color or two generic mana) to allow more decks to have access to other colors. We included cards that created off-color tokens. We made more cards that could change their color. We hit the theme pretty strongly.
During the last month of vision design, we have a meeting known as the Vision Summit. This is where the Vision Design team shows off their work to the rest of R&D to get feedback. The different invitees represent different concerns downstream of Vision Design. The meeting allows the Vision Design team to hear a lot of practical feedback about the set that they can work on before handing off the file. The Lost Caverns of Ixalan (then codenamed "Offroading") had a rough Vision Summit.
Play Design hated having basically five different Treasures. They felt it was hard to track and would lead to a hyper-fast play environment. Also, the tokens' mana abilities were so much better than their role as a resource, so there was a concern that no one would use craft. The Vision Design team pivoted from artifact tokens to counters as a result of the feedback. Counters had no secondary function and were created by exiling a card of the appropriate color from your graveyard. This was done with a new mechanic called "dig." Cards still had recipes on them for craft.
The finished set kept craft but shifted from caring about colors to card types. In fact, the entire "color matters" theme was jettisoned from the set. It also spent a lot of the structure on capturing elements of Ixalan once we decided the set would take place there (things like bringing back explore and Dinosaur typal).
Looking back, my biggest flaw in this set's design was that I was more focused on finding a way to use a particular theme than on finding the best execution of what the environment needed. I was so enamored with trying to make "color matters" fit that I didn't ask myself, "Is 'colors matter' the best fit?"
I'm a big believer that design has to make bold choices. You can't execute on a vision without exploring what happens when you go down a certain path. I'm not unhappy that I explored "color matters" as a theme. I think I was right to try it. Looking back, my frustrations lie with the fact that I wasn't able to see when that choice was restricting the design more than enabling it. Here's a quote that I like by psychologist Abraham Maslow from his book The Psychology of Science, "If the only tool you have is a hammer, you tend to see every problem as a nail."
There is a power in having a strong focus, but you have to recognize when that focus is leading you astray. I held tight to the idea of an underground world being the way to bring back "color matters" from the day I signed up to lead the set. I championed that theme all the way through exploratory design and vision design until the day I handed the set off. It was wrong, and I should have recognized it sooner.
My biggest lesson from the design was that I was my largest obstacle.
Murders at Karlov Manor
Lesson: "Stories need plots; Magic expansions need settings."
Murders at Karlov Manor originally began on a brand-new plane and ended up taking place on an existing world, although the change for this expansion was made during the middle of vision design, allowing us to adapt to it. The idea that got the set greenlit was simple: a top-down murder mystery set with an elaborate series of puzzles baked into the set. I co-led this set with Mark Gottlieb. Mark is a brilliant puzzle creator. From the beginning, the plan was that he would oversee the puzzles. I oversaw exploratory design and early vision design and handed off the file to Mark. He led the second half of vision design and the beginning of set design.
The design was treated as a top-down design. We started by asking ourselves what the audience would expect from a set that captured the genre of murder mystery. We began with the low-hanging fruit. We brought back investigate, first seen in Shadows over Innistrad. What do detectives do? They investigate murders. Well, we literally already had a mechanic for investigating.
Next, we tweaked morph. A core part of murder mysteries is the mystery. There are unknown facts that must be discovered by the investigation. The idea of a face-down mechanic felt like a good fit for the unknown. Your opponent has to figure out what the creature is. We had talked about updating morph for years (creatures had improved so much over the years that paying three mana for a 2/2 wasn't as strong), but we finally did it with this set. At first, we tried making 3/2 creatures, but we ended up with 2/2 creatures with ward
We created the Detective creature type (it technically first appeared in a set of Commander decks, but it was borrowed from Murders at Karlov Manor) and added a few cards that care about it. I became very focused on finding ways to get specific words into rules text for top-down designs. Making Detective a creature type allowed us to have cards that give off murder mystery vibes. We made the suspect mechanic, mostly to allow us to create card text that made creature suspects. We added Cases, a new enchantment subtype, to the set. This was another mechanic we'd fiddled with for years, starting back in Zendikar. Originally, these were similar to the Quests from Zendikar, where if you complete a certain task, you get a reward. We finally cracked the right way to do it with Cases. Finally, we made the collect evidence mechanic, which uses the graveyard as a resource.
The shift to Ravnica came about because we realized that the story carried so much more weight if people already knew the characters involved. We had considered setting it in New Capenna, but the audience was unfamiliar with most of the characters, and the lack of a robust legal system on the plane made it an odd fit. We added some elements to the set to highlight Ravnica. We added a higher as-fan of multicolor cards, split cards, and a lot of individual cards that reference previous Ravnica sets.
In the end, the set had a number of problems. First and foremost was that "murder mystery" is a genre of story but not particularly one of environment. Gothic horror and Greek mythology, in contrast, have tropes associated with them and come with whole environments, something that gives the setting a tone and a mood. Murder mysteries mostly have shared character archetypes and plot elements. That let us make a bunch of cards but not a world. It's why we had such a hard time creating an original world for the set.
In hindsight, I think we hit the murder mystery theme too hard and didn't spend enough time leaning into Ravnica. War of the Spark had gotten away with it because the two sets preceding it were traditional factioned guild sets, so the audience had already gotten their fill. The Lost Caverns of Ixalan, while also a "backdrop set" (a term we made to describe sets based on a world but not focused on that world's mechanical identity) did a much better job of making mechanical nods to the world.
If I had to do it over again, here's what I would have done differently: I would have kept investigate. I would have kept disguise and removed cloak. Cloak added unnecessary complexity, and Duskmourn: House of Horror would do a tweaked version of manifest later in the year. I would have removed suspect. I would have kept the Detective creature type but removed all the Detective typal cards. I would have kept Cases. I would have kept collect evidence, but I would have changed the name to make it a little more universal. That would make it easier for us to bring back. (I really like the mechanic.) I would have dialed down the as-fan of murder mystery-themed cards.
With that space freed up, I would have added more Ravnica-centric designs. I would have committed to more guild-based cards. I would have contemplated a ten-card cycle with one card per guild, maybe even with watermarks. I would have done more to make the set feel like a Ravnica set.
The big lesson for me from this design was that all genres are not created equal. I'm a huge fan of resonance. I love taking things players already know and creating whole sets around them. But Magic at its core is about telling environmental stories. We don't control which cards you open, so it's crucial that, when you open a booster, the feel of the world can be spread across all boosters. That means we need to make a compelling environment. Something that's visual and can be reinforced in all the details of the world. The fact that we kept putting detective hats on characters should have been a sign that the genre wasn't environmental enough to hold the weight that it needed to.
I will say that I'm glad we did all the puzzles. They didn't have the impact we hoped, as only a small percentage of the audience interacted with them, but I think it's important to keep pushing boundaries, and Murders at Karlov Manor did something that had never been done before with a puzzle event. It was engrained into the product in a way that no previous puzzles had been, and I'm proud that we accomplished that.
Looking back, I don't think there was a way to tweak the handoff for The Lost Caverns of Ixalan to get us to where that set needed to end up. In contrast, I believe Murders at Karlov Manor, with maybe a 30% change to the file, could have gotten to where it needed to be. Interestingly, the theme through both of today's sets was that the vision was off. And that rests on me, the lead vision designer of both sets.
Case Closed
And that brings me to the end of another "Lessons Learned" article. I hope you enjoyed my introspective look back at past designs. As always, I'm eager for any of your thoughts, be it on today's article or any of the sets I talked about. You can email me or contact me through social media (Bluesky, Tumblr, Instagram, TikTok, and Twitter) with feedback.
Join me next week for the start of Magic: The Gathering® | Marvel Super Heroes previews.
Until then, may you look back at your own past choices.