The entire process of designing and developing a set, plus the time it spends going through editing and the Future Future League (FFL), is about two years. In that time, the design team creates thousands of individual cards, very few of which actually make their way into the file, and even fewer get to print. It is a very rare incident indeed when a card made up in a meeting goes from design all the way to print without some pretty big changes.
On today's Latest Developments, I'm going to discuss what kind of road it takes for a card to make it all the way from inception to print, why so few cards actually make it, and what actually takes them out.
Cards get made in two major ways in R&D. The first is in meetings (usually these are cards that are there to fill a void or trying to prove a concept), and the second is homework, where people on a team are asked to brainstorm more complex ideas, top-downs, or the like, and we will go through these cards in a meeting and figure out which ones to try out. Overall, we make a lot of cards while designing sets. Mark Rosewater estimates the percentage of cards that get created that get to print well below one percent, which is probably accurate, but a bit misleading. Way more than one percent of cards that actually get tested see print, but we make a ton of cards in meetings and through homework that don't ever get tested. Many of those are just 2G 3/2s with the set's keyword or R "deal 2 damage" upgraded with the set's theme.
Most of the cards that quickly flux or die are commons, which move around a lot due to the needs of the Limited environment. Most of these are not super exciting designs and can be submitted by multiple people at the same time during meetings. Frequently, certain keywords or mechanics just lend themselves to natural designs as the right place to start. When we started testing renown, the one-mana 1/1 and the two-mana 2/2 with renown 1 were the most obvious places to start—it was just a matter of figuring out which keywords to put on them. When we are working with higher-rarity cards, most specifically either build-around uncommons or interesting rares, we try to keep them in long enough that we can try them out a few times in Limited. Even if one feels wrong after one play-through, it may take a few tries to get a deck where the card is appropriate—then we can decide if that is too hard and isn't going to work or how we can change the card to make it the appropriate power level for the set.
There is a lot of value in unique and cool designs that don't fit into a set. We want some number of cards in the Supreme Verdict/Cryptic Command vein that could realistically go anywhere. Neither has anything to do with the mechanics in their set; they are just cool and powerful designs that someone who doesn't care about guild strategies or tribal can buy into and still find cards in a set that they want. These cards are also more likely to get cut because they can go anywhere. We want a few of these in each set, but ultimately we also really need to make sure that we print what we can with set-specific themes and mechanics since the rotational period on those is much longer. That allows us to keep finding new Magic cards for a longer period of time. If we didn't have strong themes in our sets, we would have a much harder time keeping Magic interesting year after year.
Why Cards Get Cut
Now that I've talked a bit about how cards get made, let's talk about how they get cut. When a set comes over from design to development, we use the file as more of a blueprint than a set-in-stone list of what will appear in the final set. While we do allow for certain "design favorites" that we try to keep in the set, very few other cards are locked in. Almost everything else in the set is malleable and exists solely to ensure the experience is as fun as possible.
Things that can lead to cards getting cut in a set:
- Removing a mechanic from the set
- Adding a mechanic and needing space
- Rebalancing Limited
- The card does bad things for Constructed
- Creating metagame cards for Constructed
- The card is too similar to another card in a not-yet-printed set
- The card is too similar to another card development wants to add to the set for FFL
- The card breaks something in Standard
- The card doesn't work within the rules
- The card doesn't work in digital
At times, it feels pretty amazing that we manage to keep any cards at all. Development does its best to preserve what we can and keep the vision of what design wants in the set in the set. Wherever possible, we try and retain what design really enjoyed about the set and maintain the basic shapes for as many cards as possible.
Ultimately, a card getting cut from a set is not a judgment on the card's merits but an acknowledgement that no individual card is more important than the set as a whole and that we need to change a lot as a set moves through development. A huge part of the design and development process is about an exploration of what could be fun, what is fun when you first play it, and what is fun on the 30th time you play it. Plenty of cut cards have done a lot to teach us about what the final versions of cards should look like.
Why Cards Change
The ship of Theseus is a thought experiment that ponders if you replaced a ship board by board with use, would that still be the same ship after everything had been replaced? At what point does it stop being the original ship? Magic sets have similar questions about how much a card can change and still be the original card.
For example, imagine you have the following card:
Legendary Creature—Human Warrior
When CARDNAME enters the battlefield, gain 4 life.
If it moves from a 1WBG 4/4 to a WBG 3/3, is it the same card? What if it just moved to a 2GG 4/4? What if the ability changed to "gain life equal to the highest toughness among creatures you control"? What if we made up something totally new, but kept the idea of a wedge-color legend? These kinds of questions vary designer by designer and even card by card regarding what the card is trying to accomplish. A planeswalker might change all their individual abilities, but that they care about spells might be enough to consider it the same card. Reasons like this are why we don't really track who made what card or assign credits to individual cards—there is just too much movement and too many gray areas to make anyone happy with that kind of credit.
When working on Theros, design handed off the following card:
Legendary Creature—Human Warrior
CARDNAME can't attack or block unless you control twelve or more permanents.
The point of this card was to be a top-down Hercules, and the number twelve is a reference to the number of labors he was said to have completed. So, at what point does this card change too much to be considered the same card? Changing the number muddies the clear design intent, as does changing the creature type away from Human (It became a Hydra, which kind of killed the story of the card, before eventually changing to Mistcutter Hydra for FFL reasons). These are the kinds of complex questions that we get into for preserving design intent and trying to make sure everyone is happy with cards.
Because so much of the set changes, it can often feel like very little of your individual work as a designer or developer remains in its final form—but usually a lot of that work doesn't go to waste: it inspires and is the basis for the cards that end up in the final set.
How Cards Change: Case Study
One of the things I do whenever I lead a set is print out a copy of the set as it existed at the design handoff, a copy of each card near the end of development, and then the final version of each card. This is to help me remember how things changed over the process of a set, and it doubles as a good example to newer members of R&D about what kinds of changes they can expect to see. It's a good reminder to me of what I was thinking about during the whole process. As you will see, a lot of the first columns are blank—these are cards that don't have a direct descendant from the original design handoff. I will show you a few pages of my Magic Origins binder to give you all an idea of what a set looks like at different points of its lifecycle.
Gideon's Fortitude started off as a sort of "signature spell" for the Planeswalker that got better if you had spell mastery. A good enough idea, but it ended up changing for very basic reasons: the upgrade wasn't meaningful enough, and it wasn't playing enough into the go-wide renown deck that red-white was trying to encourage. The final version of the card (Kytheon's Tactics) helped to better tie in to Kytheon's story of who he was as the leader of a small gang and to call back to Desperate Stand from Journey into Nyx.
Jace's Path to Sphinx's Tutelage is a very common example of how a card can change pretty significantly but still be very much the same original idea. When Shawn Main handed off Origins, he had a cycle of "Path" cards that represented the . . . well, path each Planeswalker had to go down to become who they were. Each of these were very much build-arounds that have a quest involved. When I took over the file, I kept the idea, but ended up just making them build-arounds that represented the plane and worked with the general strategy of the Planeswalker, not actual quests.
Man of Impeccable Timing, which became Firefiend Elemental, is a card that the design team got mostly right (having a haste creature with renown was a good way of showing off the mechanic); we just spent a lot of time fidgeting with numbers. We found that 3/2 haste for five was too weak, but it was also weak as a four-mana 3/1. Both had the same problem, though: it was super punishing if you ever got hit with it, the five-mana one being more so, since the extra toughness added a lot of resilience there. In the end, we ended up learning a lot about how big the reward should be for renown and how much fun it was on either side of the table.
Also on this page, Chandra's Scorching (which became Fiery Impulse) shows how we will end up changing things to fit curves. There was nothing wrong with the original spell; we just needed a Shock to better work against the early renown creatures, and it also helped to make it an interesting possibility in Constructed.
Giant Enchantress was intended to be a reward card for the red-white deck, which was at that time an enchantments-matter deck to better tie into Theros. As we worked on the set, we decided that we needed to pare down on some of those kinds of sideways themes and ended up removing that aspect of red-white and focusing just on renown. I still liked a big creature that could be a finisher and ended up turning the Giant into Landslide Elemental (and then Seismic Elemental) to better work with the red-green deck that was based around ramp. The sacrificing of the land was mostly just for flavor and ended up getting cut to make the card cleaner.
Woodrager is an example of a cool idea for a card that both works a little wonkily in the rules and just generally can create some real problems for formats like Commander. The card looks good enough, but because any mana-sifting ability ends up letting this become arbitrarily large, it isn't actually that fun. It also led to some weird gameplay around having to track activated abilities, so we changed it to the version that got counters. In the end, we decided it might be a really cool Constructed card at lower numbers and not scaling so heavily.
Nissa's Revelation is also an interesting case; it was made after the main development of the set ended. We commissioned several pieces of art for Magic Duels, and this came back as one that showed an important part of Nissa's story that wasn't otherwise represented on the cards. I was checking with the story and art team to figure out if there were any story beats that were missing in the set and came upon this art. I ended up designing the card to the art, rather than the art being designed to the card.
That's it for this week. Join me next week when I'll be talking about developing the embalm mechanic.
Until next time,