Welcometo the second and final week of Dissension previews. Last week I showed off a certain legendary ooze mutant. This week I have something equally exciting. But rather than tease you for the length of the column, I thought I start off by just showing you the preview.

Yes, in case you don't read Magic Arcanas or just haven't visited the site in the last few weeks, split cards are back! And not just any split cards – gold split cards! So how exactly did this mechanic sneak its way into Dissension? I'm glad you asked because that is the topic du jour.

Banana Split

We start our story by flashing back to early Ravnica design, the very first design meeting in fact. I sit my team down and we have the following conversation: (as always please allow for some dramatic license)

Me: Here's the challenge ahead of us. We have to do a multicolor block. It needs to be compelling and exciting, but it can't feel like we've just redone Invasion. This is the first major block theme we're revisiting. We need to prove to the players that we can do it without simply rehashing what we've done before.
Tyler (Bielman): So what does that mean?
Me: It means that we need to avoid everything we can that reminds players of Invasion.
Mike (Elliott): Such as?
Me: Well Invasion was all about playing as many colors as you can. Control needs to focus on less colors.
Aaron (Forsythe): But we can do gold cards.
Me: It's a multicolor block. We have to do gold cards.
Richard (Garfield): Off color activations?
Me: Probably.
Mike: Color matters?
Me: No. We should avoid that.
Aaron: What else should we avoid?
Me: Kicker, gating, battlemages, divvy, domain. We have to stay far away from all that stuff.
Tyler: What about split cards?
Me: Absolutely not! Split cards were one of the mechanics that most defined Invasion. We shouldn't touch split cards with a ten-foot pole.

This begs the question: what happened?

To explain, let's jump to the beginning of Ravnica development. One of R&D's fears about Ravnica block was that it was very predictable. Once the players understood the structure, they felt, the sets would be in danger of not being exciting enough. In essence the players would know too much to be surprised. I defended my belief that knowing general structure is not the same thing as knowing execution. Sure, you all would know what guilds Dissension had, but that didn't mean you knew how exactly we'd design them. You wouldn't know the keywords or the exact strategy or the individual cards.

To be safe, they said, why don't we include something extra to send the block off with a bang? Give the audience one last surprise. Just because I didn't think the set needed the surprise didn't mean I wasn't excited to have it. Okay, I said, let's do something fun. And with those words, I committed the Dissension design team to come up with some cool twist. I had no idea what it would be. I simply knew that we had added fifteen extra cards to give us enough room to add it in. I assumed we'd figure it out.

Let's Split

Flash forward to Dissension design. Early on, the design team (Aaron Forsythe, Mark Gottlieb, Brandon Bozzi and myself) is meeting and we have the following conversation (at least this is how I remember it):

Aaron: So, in Ravnica development Rosewater agreed that Dissension would have some cool block-end twist to add extra excitement to the set. We were given fifteen extra cards to make sure we have enough space for whatever it is we come up with.
(Everyone looks at me.)
Me: I promised because I had faith in this team.
(Everyone continues to stare.)
Me: I have some ideas.
Mark (Gottlieb, not me): Such as?
Me: Gold split cards.
Aaron: We can't do them. They're from Invasion, remember? What else you got?
Me: I'm tapped out.
Mark: You said you had some “ideas”. That implies more than one.
Me: Well, I had some different ideas on how to execute gold split cards.
Brandon: I think gold split cards are cool. What's wrong with them?
Aaron: We're trying to avoid things from Invasion block to make this block seem different from the last multicolor block.
Brandon: So why did Mark, the Head Designer who laid down this rule, just suggest gold split cards?
Aaron: I don't know. Mark?
Me: Because they'd be cool. And it would be a chance for us to sneak in a couple final cards for each guild without breaking the guild model.
Aaron: Agreed. The only problem I seem to have is that you, my boss and the guy in charge of design, told me we couldn't do them.
Me: Is that all? Okay, we can do them,
Aaron: But why?
Mark: Does it matter? They would be pretty cool.
Aaron: Yes, it matters. I like them too, but I have to hear why we're just abandoning a rule Mark set out day one of Ravnica design.
Me: Why? Because we don't need the rule anymore. Ravnica block got out under Invasion's shadow. We did it. We made the two sets fundamentally different. We can do split cards because it doesn't matter anymore. Ravnica will not be seen as Invasion II. Besides, these are guild split cards…
Aaron: So you can just make up rules and then get rid of them whenever you want?
Me: It's good to be the king.
Aaron: Okay then. How does the team feel about split cards?
Me: Gold split cards.
Aaron: Gold split cards.
Brandon: Awesome.
Mark: I like them.
Aaron: Me too. Mark?
Me: Didn't Invasion do them?

Since I buck authority (even if it's myself), I had spent some time thinking about the split cards for Dissension. I hit upon the idea of gold split cards (I say gold rather than multicolor in that traditional split cards – the ones from Invasion and Apocalypse – were technically multicolored) and I felt it reinforced the guild model. And hey, we've never had gold split cards.

A quick aside, as I've mentioned multiple times in my column, the split cards are the favorite mechanic I've ever designed (and if you're interested in learning more about how it came to be, check out my column “Split Decisions”). I've been itching to bring them back since, well, about two minutes after I finished designing the Apocalypse ones. But the designer in me knew I had to wait for the right moment. My suggesting them to the Dissension design team came from my belief that this was the right time.

Once I knew I wanted them to be gold, this led me to the next realization: the two cards had to overlap in a color. The reason for this is simple. I didn't like the idea of each side being the same (it just felt more like a kicker variant than a split card) and if the two sides didn't share a color, they could only be optimized in four-color decks. But by overlapping a color, it allowed three-color decks to make use of them. From playtesting, we had learned that full block drafts were going to result in primarily three-color decks.

Once I knew I wanted one of the colors to overlap, I was led to my next realization. For Ravnica, we changed the frames for multicolored cards. As an example, here is the card Savage Twister from Mirage in the pre-Eighth Edition frames.

savage twister

Next, here is one of the few examples of a Red/Green gold card in the new border pre-Ravnica.

Ach Hans, Run!

Now let's take a look at the Guildpact version of Savage Twister.

savage twister

The major changes are the colored pin lines around the frame and the color gradation in the text box. The important point for this story is that the new frames dictated how the split cards needed to be made. Not only did the two cards have to overlap a color, that color had to join up in the middle. That is, the shared color on the left split card had to have the pinline and gradation on the right while on the right split card it had to be on the left. As we will shortly see, this little visual restriction would make the task slightly more complicated than I first thought.

4-10 Split

The Dissension design team was on board with the gold split cards. They agreed with me that they needed to overlap a color. And they agreed with me that the shared color needed to meet in the middle. With these guidelines in place, Aaron went off to make the holes for the team to fill in. It was at this point that Aaron stumbled across a very important detail.

Okay, how many of you out there know how we decide what order to put mana symbols in the cost of multicolor cards? I'm guessing not a lot, but after today you will. Here's how it works. Let's start by looking at the color wheel. The back of a Magic card will do.

For purposes of this discussion, you have to imagine a line between the green mana symbol and the white one.

Why? Because that is the natural break we make between the colors when we design and develop the cards. We call it the WUBRG order (U stands for Blue for those that don't know this convention – B was already taken for Black and L was taken for Land). This is the order we always put the colors into. It is most noticeable on the cards in two places. First, the collector number goes in WUBRG order and second, the order of mana symbols in the mana cost are ordered in this way.

To figure out what order the colors go in, pick two colors. Now going around the circle, pick the shortest path between the two colors (you may go either clockwise or counterclockwise). Draw the line you've made. To figure out the order, start with the color earliest in a counterclockwise direction. For allied colors, it breaks out as such:

white manablue mana
blue manablack mana
black manared mana
red managreen mana
green manawhite mana

All of the ally color combinations follow WUBRG order except GW. Because the closest physical connection crosses the line, it means that the mana symbols are reversed from the traditional WUBRG order.

Let's try the enemy colors:

white manablack mana
blue manared mana
black managreen mana
red manawhite mana
green manablue mana

In enemy colors, the extra separation between the colors causes two combinations to “break the line” (R/W and G/U). Why have I bothered to explain all of this? Well, first off, I think stuff like this is quite cool to know. But for those of you out there that don't get a rush out of understanding the nooks and crannies, it does explain in depth the problem Aaron ran into. Once again let's lay out the ten two-color combinations.

white manablue mana
blue manablack mana
black manared mana
red managreen mana
green manawhite mana
white manablack mana
blue manared mana
black managreen mana
red manawhite mana
green manablue mana

To make my point, let's try looking at the ten in a slightly different order:

white manablue mana
white manablack mana
blue manablack mana
blue manared mana
black manared mana
black managreen mana
red managreen mana
red manawhite mana
green manawhite mana
green manablue mana

See the issue yet?

Yes, in our mana ordering system, each color appears twice on the left and twice on the right. This meant that if we wanted the two colors to line up in the middle, our choices were much more restricted than we originally believed. But wait, the problem gets a little knottier.

The team decided that we wanted to do ten split cards such that all ten three-color combinations (the easy way to think of this is that three-color is the same as excluding two-color; that's why both two-color and three-color have the same number – ten – of combinations) had one split card that supported it. When you look at the colors based on where the colors sit on the card, you realize that there are four combinations that have a five-card cycle.

The ally cycle:

white manablue manablue manablack mana
blue manablack manablack manared mana
black manared manared managreen mana
red managreen managreen manawhite mana
green manawhite manawhite manablue mana

The enemy cycle:

white manablack manablack managreen mana
black managreen managreen manablue mana
green manablue manablue manared mana
blue manared manared manawhite mana
red manawhite manawhite manablack mana

And two ally/enemy combo cycles:

white manablue manablue manared mana
blue manablack manablack managreen mana
black manared manared manawhite mana
red managreen managreen manablue mana
green manawhite manawhite manablack mana


white manablack manablack manared mana
blue manared manared managreen mana
black managreen managreen manawhite mana
red manawhite manawhite manablue mana
green manablue manablue manablack mana

You'll notice that both the ally and enemy cycles double up, meaning that between the five cards only five two-color combinations appear, although they appear twice. This leads us to look at the two ally/enemy combo cycles. When Aaron got this far he made an important observation about those two cycles. Can you see it?

To make it easy, let me show them again with one piece of information added:

white manablue manablue manared manawhite manablue manared mana
blue manablack manablack managreen manablue manablack managreen mana
black manared manared manawhite manablack manared manawhite mana
red managreen managreen manablue manared managreen manablue mana
green manawhite manawhite manablack managreen manawhite manablack mana


white manablack manablack manared manablack manared manawhite mana
blue manared manared managreen manared managreen manablue mana
black managreen managreen manawhite managreen manawhite manablack mana
red manawhite manawhite manablue manawhite manablue manared mana
green manablue manablue manablack manablue manablack managreen mana

The problem is that the ally/enemy cycles only represent half of the ten three-color combinations. (The five trio combos listed above appear twice while five others never appear.) This meant that our only choice to have ten gold split cards that represented all ten three-color combinations was to go with the ally and enemy cycles. After a little discussion, we decided to put the ally cycle at uncommon and the enemy cycle at rare.

I realize this section threw a lot of data at you, but I wanted you to get a glimpse of all the intricate detail that comes into doing something as seemingly innocent as gold split cards.

Split Infinitives

So we finally have ten gold split cards that we know what colors each half has to be. This leads us to the next step of split card design – the names. Say what? Isn't naming the card near the end of the process, long after design is finished? Normally, yes, but split cards are a special case. You see, in Invasion, we started by making the five cards. After they were done, I came up with the _____ & _____ convention for the names (don't get me started on the _____ // _____ thing). We then searched high and low to find five _____ & _____ names that sounded good, made sense with the color combinations and fit the mechanics of the card.

The task was so complicated that during Apocalypse design, I decided it was easier to start by naming the cards based on the color combinations and then design the mechanics to fit the names. Quick trivia aside – because I wanted to use the names to represent the conflict between the enemy colors, I picked classic ____ & ____ expressions that talked about the conflict. (Stuff like Life & Death and Fire & Ice.) The most classic of them ended up working backwards from the order convention I named above (I mean who wants Death & Life or Ice & Fire), so I just swapped the direction of all five of them. This is why all five are backwards from normal.

I knew when we started the split card design for Dissension that we had to use the same technique. The team started brainstorming names. Then we discovered the advantage of having a guy who makes word puzzles on the side (that would be Mark Gottlieb). Mark came into the next meeting with a printing of every combination of _____ & _____ that existed. The team went through the list and circled every one we thought sounded like it could work for a Magic card.

Next we had to match up names we liked with particular cards. The trick was finding a name where each half fit a particular color combination. Even with the long list we started with, we found that certain combinations only had one or two names that at all worked. Luckily, the team managed to find names for all ten that we liked (including one pair called Research & Development; see above for the line about “being king”).

Once that was done, it was time for the real challenge. For each card we needed to find the following - two effects that matched the proper color combination, that fit the philosophy of the guild, that didn't repeat something done earlier in the block, that matched the names and that, on some level, made sense appearing on the same split card. I often say that restrictions breed creativity. Let's just say that this exercise brought out a lot of creativity.

To give you a little glimpse into what that process was like, let's take a look at Odds & Ends. (Yes, yes all the official spoilers are going to call it Odds // Ends, but at least one place can use the names as they were actually envisioned. Why has the ampersand been forsaken so?) To begin, let's look at the card again.

Still cool, isn't it? Okay. So we had two parts. A U/R half called Odds and a R/W half called Ends. Let's start with Odds. One of the tricks I've always enjoyed on split cards is having the name on an individual mini-card mean a different definition than it means in the combination name. Odds in Odds & Ends is talking about bits and pieces. But Odds can also mean probabilities. You know, like coin flipping. Coin flipping, though, is red. This meant that we had to come up with two effects that felt blue. After a little searching, we realized that a copy effect was cool as it overlapped Blue/Red. Once we knew we had a copy effect, it quickly became obvious that the logical companion spell must be a counterspell. This way the spell either stopped the threat or gave you your own copy of it. The best part of this card is that it played perfectly into the philosophy we set up for the Izzet. It was spell oriented and chaotic. It had a touch of the dramatic and an element of the unknown. A perfect Odds.

This leaves Ends. We tried the same trick as last time. Ends in Odds & Ends means pieces. But Ends is also a verb that means to bring to a conclusion, or in the case of a life form, to stop it. Odds already had a very natural defensive quality to it. What if Ends could also be used defensively? Possibly to end the threat of a creature or two. By affecting attackers, we brought the mechanic back around to the Boros. Their dominance is in combat. What better for a Boros spell than an instant that removes attackers. (While the rules do allow split cards to be different card types, we've found that doing so unnecessarily confuses people and we avoid doing it.)

Whew! And we had to do that ten times. I thrive on doing crazy stuff like that, so I had a blast designing the split cards (as did the rest of the Dissension team). I'm quite proud of how the split cards came out. I'm very curious to see what you all think of them.

The Splits

That's how it happened. A little more than you expected walking into this article, huh? It's interesting how hard it is to judge what pieces of the design will cause the most trouble when you start.

Join me next week when I might not talk about design.

Until then, may you have fun looking at your cards sideways.

Mark Rosewater