a series of interesting choices thoughts on game design from paul sottosanti


designing magic: planeswalkers & lorwyn

This is the fifth part of a series about working on Magic: the Gathering during my five years at Wizards of the Coast:

Part One (Joining WotC, playing in the FFL, developing Champions of Kamigawa)
Part Two (Developing Betrayers of Kamigawa and the Jitte mistake)
Part Three (Developing Ravnica, unleashing Friggorid, and designing MTGO Vanguard)
Part Four (Designing Ripple, the FFL during Time Spiral, designing Planar Chaos)

Future Sight

I wasn't involved in creating Future Sight, but I was a big contributor in a major debate that had started in one of our Tuesday Magic meetings, where everyone involved in Magic gets together to discuss the current major issues. The discussion revolved around the new card type that we were going to be debuting soon: Planeswalkers. The Future Sight team wanted to put a Planeswalker into their set, to increase excitement and foreshadow the "real" release of Planeswalkers coming in Lorwyn. I could see where they were coming from but it seemed like a terrible idea to me for several reasons:

  1. New things are exciting, but you only get that initial burst of excitement once. Why spend the Planeswalker excitement on Future Sight, a small set that already had a "hook" (these are cards from the future, some of which will become real later), when we could save it to sell our forthcoming large set? As much as people argued that the "real release" of Planeswalkers would still be Lorwyn, I knew that players wouldn't see it that way.
  2. Future Sight was already far above average in complexity and wackiness, thanks to a new card frame showcasing "future" cards like Steamflogger Boss. The set really didn't need a whole new card type for players to have to learn.
  3. The Planeswalker team needed more time. No one was completely happy with how they worked yet, and the balancing of a new card type was proving tricky. We had no idea of their power level and could've risked printing a completely unplayable Planeswalker or one that was far too strong.

In the end we compromised with referencing Planeswalkers on Tarmogoyf, a move that I think ended up being the best of both worlds. We foreshadowed Planeswalkers (and Tribal, for that matter) without actually showing our hand, which drove a lot of hype for the upcoming set. Of course, at that point no one had any idea that it would go on to become arguably the best green creature of all time.

Planeswalker Design

Around this time, Magic was going through somewhat of a creative reboot to make it more accessible and also better suited to supporting other potential media like books, graphic novels, or maybe even movies and TV. The goal was to have recognizable characters that could show up in multiple expansions (despite the different settings) that were relatable for players. Luckily, the IP already had an answer for this: Planeswalkers. And since the only way to get most Magic players to care about a character is to put a card depicting that character into their hands, it was time to add them to the game.

I was a part of the team tasked with figuring out what this new card type was going to do. Unfortunately I don't remember much about the process, but I do remember that one of the original designs had them following a list of instructions. They triggered on your upkeep I believe, and they would perform the first ability in the list on the first upkeep, then the second, then the third, and then they would wrap around and do the first ability again. We arrived at this because it didn't necessarily make sense to let you control another Planeswalker; it'd be like asking your friend to come play basketball with you and then having to direct their every move. With the ability list we encountered the opposite problem: they felt less like sentient beings and more like robots, especially when they performed an ability that didn't make sense with the game state, like giving a creature +4/+4 when you controlled no creatures.

We tried to get around this by giving them abilities that flowed into one another, like letting them create a creature before empowering it, but situations where the creature died in the meantime kept cropping up. Plus, it was sometimes hard to remember what step you were on, and marking them with tokens didn't work since you were already keeping track of their loyalty. Eventually someone hit on the idea of charging or granting loyalty for the abilities, and we changed it to let you choose which ability was activated. Then we let you activate them the turn you played them, which was a huge upgrade in terms of constructed quality and also just the overall feel. Suddenly playing a Planeswalker and having your opponent kill it on their turn didn't feel that bad.

Overall I think the Planeswalker card type was a huge success. I know that Richard Garfield has expressed concern about the amount of complexity that was added, and I agree with him that it's unfortunate that a new player can get a card that works so differently from everything else. At the end of the day, though, they provide some great gameplay, your opponents can interact with them (through burn spells or attacking them), and I think they've contributed to the recent success of Magic.


The Lorwyn design team was Aaron Forsythe, Mark Rosewater, Brady Dommermuth, and Andrew Finch, with Nate Heiss replacing Andrew partway through. I can't say enough about how much I enjoyed working under Aaron. He's as well rounded as you could hope for: strong design skills, with the playskill and understanding of a developer, a huge knowledge of the game's history, and enough business sense that he's now been the director of Magic R&D for almost five years. What I enjoyed the most though is that he's one of those people who, at the end of the day, just wants what's best for the game. When everyone on a design team puts their egos aside and works as a unit towards that singular goal it can be quite the magical experience.

So much went on during Lorwyn design that I don't really know where to start, but I'll try to break it down into cohesive sections.

The Setting

Lorwyn takes place in an idyllic world filled with rolling pastures and beautiful forests. Where possible, we wanted to convey that through the cards and make the player feel the happiness and contentment of the world. One of the earliest ideas I had was to have a world without killing. Black wouldn't get any "destroy target creature" cards, and instead removal in this world would be themed as tricks and pranks, with the occasional application of -1/-1 counters. Using -1/-1 counters was potentially important because we knew that the last two sets in the block (Shadowmoor and Eventide) were going to be an evil version of the Lorwyn world, after a Cataclysm-type event, and at the time all of the sets in the same block had to use the exact same type of counters. There was just too much confusion from trying to support multiple different types of counters on the same card, and by keeping them out of the same block we at least kept those situations out of most sealed deck and draft experiences.

This turned out to be a terrible idea. Not only did it have the anticipated problems of causing frequent stalemates (we thought we could solve that by seeding certain types of cards into the set), but the slow application of -1/-1 counters actually made the set feel more evil than most. You weren't just cleanly killing creatures, you were torturing them to death over multiple turns. We went back to the drawing board, and soon found ourselves bringing up the problem at another one of the Tuesday Magic meetings. We lamented the fact that we couldn't use +1/+1 counters, and people started throwing out suggestions, but none of them seemed viable.

Eventually, as a joke, I said that this would be so much simpler if +1/+1 and -1/-1 counters would just annihilate each other, like matter and anti-matter. In hindsight, I guess there's no great reason why I thought it could never work, but as a designer you tend to quickly get used to the editor and rules manager telling you that ideas just can't possibly work within the rules. It's not that they're not good at their jobs, or that they're trying to stifle innovation, it's just that they're the ones who have to deal with all of the corner cases and who have to clean up the mess if anything breaks, and they're naturally a bit conservative. To my surprise, though, I saw them seriously considering it, and pretty quickly they said that it could function as a state-based effect. It didn't take long for the room to agree that with the support of that rules change, Lorwyn and Morningtide could use +1/+1 counters, and Shadowmoor and Eventide could use -1/-1 counters. Suddenly everything was so much easier.

Race and Class

Early in the process we had a meeting to decide the plan for the first half of the block. Specifically, what would we leave for Morningtide so that it would feel like it's own set and not just an extension of Lorwyn? It was a unique situation because there were only two sets to worry about; Shadowmoor and Eventide were doing their own thing (focusing on color) even though they would take place in the same physical location. I started thinking about how we had just recently done the race/class update, where all cards now had a race (a card that used to be just "Cleric" would now be "Human Cleric") and all sentient races now had a class (instead of just "Elf" it would be "Elf Warrior"). I knew we should take advantage of that somehow, and quickly I hit on the idea of having Lorwyn focus on the races like a traditional tribal set and Morningtide focus on the classes, which hadn't specifically been done before. Since the cards all had two types, and each race would be confined to only a few classes, new possibilities would open up for existing decks when the class tribal cards showed up in Morningtide. The team was on board with the plan and it never changed again.

The Tribal Supertype

We now knew how Morningtide would set itself apart, but that still left us to figure out how Lorwyn was different. We had all played Magic during Onslaught and enjoyed it, and we knew that just printing a bunch of cards with specific races and some other cards that enhanced them wouldn't be nearly as awesome the second time around. One idea that was rolling around, that I absolutely loved, was to apply the creature types to other types of spells. Essentially, not only would we have Elves, but we'd have Elven sorceries, Elven enchantments, Elven artifacts, and even Elven lands. This would solve one of those age-old problems of tribal decks, that you couldn't play any spells at all without diluting the tribal qualities of your deck. If you wanted 40 Goblins, you had to have 40 creatures.

We set out trying to figure out how this would work. What did it mean to be an Elf Instant? Certainly we could make cards that said "Whenever you play an Elf, get X" but it felt like the Elf instants themselves should be tied to Elves somehow. Some options I put forth:

  • As an additional cost to play this spell, tap an untapped Elf you control.
  • Play this spell only if you control an Elf.
  • Tribalcycling (, discard this card: Search your library for a card that shares a subtype with this one, reveal it, and put it into your hand. Then shuffle your library.)
  • If you tap an untapped Elf you control as you play this, use the Tribal effect instead. (e.g. +2/+2 or Tribal: +4/+4)

We went through the pros and cons of each of the team's suggestions, and eventually settled on the final one from the above list and calling it Mojo. It allowed you to play the spells even if you hadn't drawn any creatures yet still rewarded you for having the right type in play.

Next we moved forward with designing Mojo cards, at least until we hit the next snag. Mark Gottlieb came to us and told us that "Instant - Elf", which was how we were planning on templating the cards, wouldn't work. The game rules didn't allow creature subtypes on anything other than creatures. We tried debating, but there were a lot of subtleties he was worried about, like old cards that said something like "put an Elf from your hand into play." You can read about more about this process in Aaron's article about Lorwyn.

Aaron fought bravely for it but eventually saw the writing on the wall and gave in. Back to the drawing board.

Of course, as you all know the set did in fact ship with Tribal cards. What happened? Well, a couple things. We tried some other options, like creating a little icon for each tribe and then putting those on the sorceries, but it forced awkward "whenever you play an Elf or [elf icon] spell" wording that everyone hated, so we abandoned that. I found the set far less exciting without our marquee mechanic, and ended up sending at least one impassioned email to Aaron to try to find a way to fit it back in somehow. Aaron asked Mark to keep trying, and he soon came through for us and found a way to make it work. If we created a new supertype called Tribal, and one of the qualities of being a Tribal card was that you could have the creature subtypes, the rules suddenly came together much more cleanly. We were back in business.

In the meantime we had decided that the Tribal spells didn't need a specific mechanic tying them together, and just triggering all of your effects that cared about creature types was enough. So Mojo was killed from the set.


We also knew that the set needed something else to carry it, something that was outside of the tribal theme. Early on Aaron suggested a "treasure" mechanic. Basically there were cards that gave you treasure, and then you used your creatures to "explore" and find the treasure, which put cards into your hand. Cards like:

Torch and 10’ Pole

Artifact — Equipment
Equipped creature has “: Explore. (Reveal a random card from your treasure pile. You may pay that card's converted mana cost. If you do, put it into your hand. Otherwise, return it to your treasure pile.)

Phlogiston Vial

, Sacrifice CARDNAME: CARDNAME deals 2 damage to target creature or player.
If CARDNAME is revealed while exploring, it deals 2 damage to the exploring creature.

One problem was that it was simply too high of a bar to have to assemble both cards that gave you treasure and also cards that could explore. Explore as a specific ability also made the mechanic parasitic, in that it only worked with cards from the set, which has become a death knell for many potential mechanics ever since the Kamigawa block. We worked through a bunch of problems with treasure, getting it to a pretty good place, but there was a lot of internal debate over the complexity of the mechanic. It really needed to show up heavily at common to be meaningful and interesting, which created a lot of pressure to simplify it. Eventually it was cut from the set instead, but not before inspiring the Hideaway lands.


I won't spend much time talking about Transcendence, because it wasn't around for long, but essentially it was borne out of the idea that one way to show the tranquility and happiness of the setting was to support an alternate path to victory from killing your opponent. I settled on a simple mechanic that was essentially the opposite of poison, where you accumulated counters, and if you got enough of them you won the game outright rather than losing it. This led to cards like:

Salvation Army

Creature – Merfolk
, discard three cards: Gain a transcendence counter. (A player with five or more transcendence counters wins the game.)

Creeping Fungus

Sorcery – Elf
Destroy target land.
Mojo – If you control more lands than that land's controller, also gain a transcendence counter. (A player with five or more transcendence counters wins the game.)

I liked it, but others felt it wasn't interactive enough, and Aaron decided to pursue other directions.

Theming the Tribes

Wizards employees aren't allowed to play in sanctioned Magic tournaments, so to scratch the competitive itch I had been playing a lot of the VS System card game from Upper Deck Entertainment. The Avengers set had recently come out, and one of the teams in that set was the Squadron Supreme. They were based around a mechanic where they grew incredibly strong when you managed to empty your hand (which was less crippling in that game than it was in Magic because you could put plot twists and characters in your resource row), and as such had a plethora of cards based around discarding. In a vacuum those cards would be quite poor, but in a well crafted Squadron Supreme deck they were devastating. While I'm not a fan of the VS System design in general, some of the most fun I've ever had playing a card game came out of drafting Squadron Supreme decks. I loved that they had such a strong identity and I greatly enjoyed the discovery process of figuring out how to make the deck tick. Many of the other teams in VS had their own theme that came through in the cards as well. One of my first pitches to Aaron was that we should use that as inspiration to make each of the tribes in Lorwyn feel unique and stand on their own.

I'll include a few of my early card submissions for each race below to provide a better sense of their proposed identity.


My initial proposal for Kithkin was that they should get bonuses for attacking and blocking together, comes into play effects that triggered if you'd already played a creature this turn, and some form of evasion (inspired by their great ancestor, Amrou Kithkin from Legends). We eventually simplified this down and focused primarily on "attack with three or more", which actually showed up later in the Great Designer Search 2 as one of Shawn Main's mechanics. The problem we found with it was that it was either a trickle or a deluge. If you only had two creatures who could attack, you sat there doing nothing and would likely lose, but if you found that third creature, suddenly +1/+1s, regeneration, and other keywords all turned on and your combat steps would be a complete blowout. There was no middle ground and therefore few interesting games. They lost that mechanic but they kept an overall identity of small boosts, large numbers, and evasion, and attacking with lots of creatures lived on in Cenn's Heir.

Kithkin Commander

Creature – Kithkin Soldier
When CARDNAME attacks, if there are three or more attacking creatures, creatures you control get first strike until end of turn.

Little Follower

Creature – Kithkin Soldier
When CARDNAME comes into play, if you've already played a creature this turn, put two +1/+1 counters on it.


Instant - Kithkin
Attacking creatures can't be the target of spells or abilities this turn.
Mojo - Attacking creatures also can't be blocked by creatures with power 3 or greater this turn.


My proposal for Merfolk was that they were resource traders. Essentially, we would make a bunch of engine cards that turned one resource (life, mana, untapped creatures) into another, and if you assembled the right ones you could create machines that would slowly but surely give you an incremental advantage. This slowly converged to an emphasis on tapping and untapping, with rewards like milling your opponent, gaining life or drawing cards. We also added a subtheme of Islandwalk plus the ability to turn your opponent's lands into Islands. I love how the Merfolk turned out and drafted them at every opportunity when the set was finally released. Summon the School was a powerful card that made the whole deck come together perfectly.

Merfolk Tinkerer

Creature – Merfolk Wizard
Whenever you draw a card, you may untap target permanent.

Merfolk Healer

Creature – Merfolk Cleric
, discard a card: gain 4 life.

Merfolk Opportunist

Creature – Merfolk Rogue
Discard two cards: draw a card.


We faced numerous challenges here. This was the first time Goblins had been primarily black, and aggressive Goblin decks had been extremely strong in Onslaught. We didn't want to add too much to those decks or have a similar metagame in Lorwyn. To that end I proposed that these guys would be the tricksters of the Lorwyn world. They wouldn't necessarily hurt you outright but they'd annoy you to no end by denying you resources or weakening your creatures. This didn't entirely work out, and we ended up using other tricks to keep them less aggressive, making cards like Mudbutton Torchrunner that were better on defense or with sacrifice effects than on offense, and other cards like Wort, Boggart Auntie that focused on bringing them back. The recurring aspect treaded on Zombie territory a little bit in exchange for some fun gameplay.

Voodoo Guy

Creature – Goblin Shaman
: Target creature loses all abilities until end of turn.

Hired Killer

Creature – Goblin Assassin
, : Destroy target creature that's already been targeted this turn.

Scary Voices

Instant - Goblin
Until end of turn, target player can't draw cards.
Mojo - That player also discards a card.


Elementals were the only tribe that would be in all five colors, with the sentient "Flamekin" in Red and non-humanoid elementals in all five colors. This is where I initially suggested the "becoming tapped" mechanic should live, as at the time we had Mojo spells that played nicely with it. That got moved to Merfolk, and we eventually landed on Elementals being a combination of Evoke and activated abilities, with a subtheme of "use this ability three times in a turn". Evoke let us print a lot of expensive creatures that still had utility early in the game, and the activated abilities gave them a nice outlet for extra mana which made cards like Smokebraider and Soulbright Flamekin quite exciting. All in all, a success.

Anger Elemental

Creature – Elemental
Whenever CARDNAME becomes tapped, you may have it deal 1 damage to target creature or player.

Mist Elemental

Creature – Elemental
Whenever CARDNAME becomes tapped, you may untap another target creature.


Instant – Elemental
Tap target non-flying creature.
Mojo - Tap all non-flying creatures.


This was a tough one. Elves hadn't ever been in Black before, and one of their more iconic cards was all about lifegain (Wellwisher). Lorwyn elves in particular were accomplished hunters. My first proposal was for them to be "more of the same with a slightly more evil twist" but late in design I hit upon a hunting mechanic that I liked. Many of the Elves would mark a creature (put a "hunted" counter on it) when they came into play, and then they would each have an ability that referred to hunted creatures, from outright killing them to whittling them down. We didn't use it though, possibly because of the counter problem (we were already using +1/+1 counters), and Elves ended up mostly focusing on creature tokens and +1/+1 counters. The hunting idea did, however, inspire Hunter of Eyeblights.

Elven Cheerleading Squad

Creature – Elf Scout
, : Target creature gets +2/+2 until end of turn. Play this ability only during your upkeep.

Dark Priest

Creature – Elf Shaman
, sacrifice an Elf: Put a 1/1 black Elf Horror token into play.

Mana from Heaven

Sorcery – Elf
Creatures you control gain ": Add to your mana pool" until end of turn.
Tribal - Elves you control gain ": Add to your mana pool" until end of turn.

The Minor Tribes

There was another team in VS, X-Statix, that had some interesting gameplay. They were all about having exactly one creature in play, and many of their creatures had ways to sacrifice themselves or would allow you to kill off your other creatures. While I didn't want to be quite as aggressive about it, I loved the idea of Giants in Lorwyn as loners who thrived by themselves, and it fit well with the evocative gameplay of Giants flinging other creatures to deal some damage before falling to their deaths. We didn't end up going that route, instead rewarding players heavily for having multiple Giants with creatures like Thundercloud Shaman to mitigate the fact that they were so expensive.

Cave Hermit

Creature – Giant Warrior
CARDNAME gets -1/-1 for each other creature you control.


Creature – Giant Warrior
While you control no other creatures, CARDNAME has Vigilance and can block two creatures at once.

Faeries ended up being largely defined by a card called Sickening Faerie in design. I left a comment on the record in Multiverse saying, "Very good but I think he's at the right level to be the premier common Faerie creature". Sure enough, the card ended up being printed as-is as Dreamspoiler Witches and provided many drafters with a reason to pick up other Faeries like Spellstutter Sprite. The theme of flying creatures with Flash fit them perfectly. They also picked up some of the trickery that we moved away from the Goblins.

Tricky Faerie

Creature – Faerie Wizard
: Until end of turn, you may play creature spells whenever you could play an instant.

Everybody Jump

Instant – Faerie
Target creature you control gains flying until end of turn.
Mojo - All creatures you control gain flying until end of turn.

The obvious qualities for Treefolk were huge amounts of toughness, keying off of Forests, and the Vigilance keyword. We gave them liberal doses of all three of these as well as some intriguing one-off designs like Doran, the Siege Tower and Lignify. One of the easiest tribes to design because their flavor fits so perfectly into Magic.


Creature – Treefolk Assassin
When CARDNAME deals combat damage to a player, if it's untapped, you may tap it to destroy target creature.

Stirring of the Oaks

Instant – Treefolk
Untap all creatures you control.
Mojo - Creatures you control get +1/+1 until end of turn.


These were created to solve a simple problem. With 8 different tribes, even with Tribal cards, there would only be 1-2 cards for each tribe in a given pack. This was rough on both casual players with a favorite tribe and drafters, so the Changelings were created as a way to let everyone fill in their decks with some additional options. Many of the power/toughness changing spells in the set were also themed as Shapeshifter magic, letting anyone use them for their creature type triggers. At this point I knew that I wanted a subtheme of "sharing creature types" (e.g. destroy two creatures that don't share a creature type, or move an aura between two creatures that do) in Morningtide, so I was excited to have them in the environment.


This was added very late in design after some all day offsite meetings at Mark Rosewater's house where we struggled to figure out what the non-tribal aspect of the set should be. Clash was nice in that it subtly rewarded Timmies for playing with expensive spells, yet everyone always had a chance to win the clash since their opponent could reveal a land. Yet Spikes strongly disliked the mechanic, because at the time you were told where to put the card, so there were no choices involved. We tried putting it on top, but it felt horrible when you needed a land, revealed a spell to a clash, and then had to wait to draw the useless spell. So we tried putting it on the bottom, but then everyone was sad when they were revealed a spell they were excited about and had to immediately lose it. The breakthrough was allowing the player to choose, giving it an element of Scry and appealing to the Spikes who could now use it to smooth out their draws.


I won't bother trying to identify specific cards that I designed; in fact, at this point I had stopped even keeping track. Not because I designed so many that I couldn't remember them all, but just because once you were heavily involved with the overall shape of a set the individual pieces seemed less important. I did notice, going through my old files, that I submitted the first two abilities for Chandra Nalaar, which is particularly cool because of how hard Planeswalkers are to design (most of them end up getting re-made during development). The printed ultimate ability is far better than mine though; I submitted "Destroy all lands." I didn't yet have a grasp of just how powerful the planeswalker ultimates could (and should) be.

And that, in a nutshell, was the process of Lorwyn design. If you have any questions about the design, or about specific cards, feel free to leave them in the comments and I'll answer what I can. The next post (the final one in the series) will be about Morningtide and the two digital games that I designed: ArtFight, and Alara Explorer. Thanks for reading!

Comments (0) Trackbacks (1)
  1. Great article. I love hearing about the origins of cards.

  2. I’ve always wondered: Was Jaya Ballard supposed to be the Planeswalker in Future Sight? And frankly, how did the Planeswalkers as they came to be really fit into Lorwyn at all flavorwise? It really felt like they were supposed to show up in Time Spiral block since, unlike later sets, they didn’t seem to have anything to do with the rest of the set at all?

    • I don’t remember it being Jaya Ballard. I read somewhere that it was supposed to be Garruk, and when he was removed the hole for a green rare was filled by Tarmogoyf, which seems plausible. But I’d have to ask Mike Turian to be sure.

      As for Planeswalkers being in Lorwyn, it’s a pretty frequent occurrence that the game design or other goals drive a decision and then the creative team has to figure out how to make it work. Luckily they do a great job and the story is in the background enough that most people don’t think twice about it. I vaguely recall a backstory about the Planeswalkers being stuck on Lorwyn because of how remote it is. Doug Beyer wrote this in an article:

      “Lorwyn is remote from many large, established planar hubs and is poorly documented in the parageographical literature. Furthermore, Lorwyn does not stand out in the void; it resides in a region of dream and æther that obscures it from most travelers. However, some planeswalkers have had success traveling there by first preparing spells that attune the mind to patterns of nature.”

      So I believe they were trying to find a way off of it during the Lorwyn story arc, but since they were stuck there they ended up getting involved with what was going on. I could just be making that up though; for better or worse the creative team and the design team didn’t work all that closely with one another. Aaron’s inclusion of Brady on the team was an attempt to fix that since we needed to work more closely with creative for a tribal set.

  3. Hello there! Would you mind if I share your blog with
    my zynga group? There’s a lot of people that I think would really appreciate your content. Please let me know. Thank you

Leave a comment