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


exploring penumbria

Jonathon Loucks had a troubled trip through the Great Designer Search 2. He set himself up with the difficult task of fitting the concept of "light vs dark" into Magic, which sounds simple in theory but is actually quite difficult in practice. His submissions were often too complex and he was often called out by the judges for submitting cards that players of his caliber would enjoy but that would be too much for more casual players. His final submission abandoned one of his more complicated mechanics, Illuminate, in favor of a new "all colors vs colorless" theme that the judges (correctly, I think) felt was less an embodiment of "light vs dark" than a replacement of it.

So how would I fix it?

One issue that makes light and dark so tricky to represent in a Magic set is that ideally you need your representations to work within all of the colors. For this reason I liked Jonathan's original direction, which was that darkness is represented by hidden information (showcased by Morph) and light was represented by illumination (though, like the judges, I disagreed with the specific implementation of Illuminate). That said, there are some challenges inherent in trying to make revealing cards a theme of your set:

(Morningtide, a set that I lead designed, flirted quite heavily with including a reveal subtheme. More on that in a later blog post.)

  1. One of the best aspects of Magic is the hidden information. The secrecy of your opponent's hand and draw step adds uncertainty, which makes the game interesting in countless ways. (Try playing Magic with open hands sometime and see how much worse it is.) This is the main reason that Wizards has pretty much never printed a constructed playable card that permanently reveals the opponent's hand.
  2. Players want to cast their spells. A reveal subtheme naturally leads you to having reveal as a cost, which requires players to hold spells back in order to reveal them at a later date.
  3. Revealing adds a large memory burden onto the players. I either have to write down what I saw so that I won't forget, or risk walking into it a few turns later and feeling really stupid because I should have known better.

So why do I still want to use revealing?

I think it's because if there is ever a time to make a set themed around revealing, this is it. The light theme is perfect, and I think it needs to be the first set in the block so that everything can be set up to support it. One of the problems in Morningtide was that much of the environment was still made up of Lorwyn cards, and that set wasn't built with a reveal subtheme in mind.

Let's continue down this path, although it's fraught with peril. Here's what I'd do:

  1. Denote a number of cards in the set as representing darkness or light. White would have more light cards, of course, but would also have some darkness cards as well. I'd start with White at 80/20, Green at 60/40, Blue at 50/50, Red at 40/60, and Black at 20/80. There would a roughly equal number of cards in each color that are unaligned.
  2. Light cards would revolve around revealing hidden information (but in limited amounts) and then taking advantage of that information. For example, a light-themed spell might ask you to target a facedown card or a card in your opponent's hand and guess the name of it, exiling the card if you're correct. Or another might ask you to choose a card type and you draw cards equal to the number of cards of that type your opponent has in their hand. Predict would be a natural reprint.
  3. Darkness cards would revolve around Morph and removing or re-hiding known information. Examples would be looting (draw a card and discard a card, allowing you to ditch a card they've seen if you choose), changing your morphs (pick up a facedown card, then put a card facedown from your hand), and shuffling your library.
  4. There would be reveal triggers sprinkled throughout the set that would do something when the card is revealed. For example, a cycle of common lands that come into play tapped and draw a card for the owner when they're revealed. (You'd probably also have to pay a mana so that they're not broken in half with multi-reveal cards in older formats.) There would also be one or two morph cards that let their owner do something good when they're revealed, so that there's a bit of a risk when using light cards to peek at the darkness morphs. These would also trigger when you flip the morph over naturally.
  5. I would keep Dig as a mana smoothing mechanic. It fits the world nicely. I like the "bottom of library in a random order" wording too.
  6. All other mechanics at common would be kept quite simple. Energize (gain a bonus whenever you play a noncreature spell), one of Jonathan's mechanics, would be a good start for light. Shroud is an obvious fit for darkness, although not something you want to use too often. It could be ramped up slightly if there were more non-targeted spells that interact with permanents than usual, though (like edicts).

The goal here would be to provide an ebb and flow of information as the light player reveals cards and the dark player endeavors to use them or hide them again. There also might be space for some darkness cards that reward the player for keeping secrets. Perhaps darkness has a cycle of cards that count the number of facedown cards you control and does something based on that? And then light has a Break Open variant that also deals 3 damage to the creature. (I've always thought that would've been an awesome card for Onslaught block. Usually it's removal, but once in awhile it backfires spectacularly. It's even better in this set since you can set it up with reveal cards sometimes and you only have to cast it blind if you're desperate or want to get lucky.)

Anyway. Am I sure that this is the right direction? Not exactly. I'm starting to wonder if this would work better in an online game, where the computer could keep track of the revealed cards for you. But I do think there's something compelling here. It's part information warfare, part structure for setting up fun combos, and part an excuse to revisit the world of morphs, but this time with a little more interactivity beyond just guessing based on how much mana they have available. I would be excited to try it out!


why i’m skeptical about active reload in SpyParty

Tonight I'm heading over to Chris Hecker's house to playtest a new feature in SpyParty, active reload (in this case, "Action Testing"). SpyParty is an incredible game that I've had the pleasure of playtesting for many hours at this point. If you haven't heard about it, here's a quote from the official site: "SpyParty is an asymmetric multiplayer espionage game, dealing with the subtlety of human behavior, character, personality, and social mores, instead of the usual spy game explosions and car chases."

The easiest way to explain how it plays is to compare it to a reverse Turing test, where instead of an AI pretending to be human, the Spy is pretending to be an AI. The job of the Sniper is to watch for discrepancies in behavior between the AIs who inhabit the party and the Spy who is walking in their midst. This could come through simple mistakes (walking erratically, bumping into things), more subtle behaviors (that guy is spending a lot of time at bookshelves), or seeing mission completions (that guy just reached into a book and removed some microfilm!). These can be categorized as "soft tells" that make you more suspicious but don't confirm anything, and "hard tells" that give the Spy away completely.

Chris has already written a blog on the new feature. Here is his primary reason for wanting to try out the mechanic:

The flaw in the current game is if a merely good Sniper is playing an elite Spy, the Spy player can’t accomplish any missions if the Sniper is looking directly at the Spy, even though he or she is much more skilled. The Spy will still win most of the time, because the merely good Sniper won’t be able to tell which partygoer to watch if the Spy is good enough, but I still consider this a flaw because an elite Spy should be able to perform missions right in front of the less skilled Sniper due to the skill differential.

The source of my skepticism starts here, because I actually disagree with this premise. There are essentially four major levels of suspicion between the Sniper and Spy in the game (with more gradations in between):

  1. The Sniper has been completely thrown off track and has either marked the Spy as "not suspicious" or is focusing on another suspect closely.
  2. The Sniper has no idea who the Spy is and is either surveying the whole party or watching a particular objective.
  3. The Sniper is suspicious of the Spy and is keeping a very close eye on him, either from one or more soft tells, or process of elimination, or a gut feeling.
  4. The Sniper is "sure" about the Spy, likely from having seen a hard tell.

Once the Sniper reaches the fourth level, the game is essentially lost for the Spy. He is about to shoot. (Although I have won a couple close ones by performing the final mission in full view of the Sniper and then doing my best to hide behind and weave through other partygoers for the final ten seconds, but that won't work against strong players.)

So let's move on to the third level, which is what this mechanic is intended to address. Chris's point is that elite Spies in this situation should be able to perform missions right under the nose of a merely good Sniper, but I disagree. The interesting interplay of the Spy and Sniper is about two things: not letting it get to this point in the first place, and then recognizing when it does and going into damage control mode, blending back into the flow of the party.

There's nothing worse as the Sniper than having a hunch and following it, watching a certain person closely for a good 30 seconds, and nothing happening. A growing panic forms as you wonder if you were correct in the first place, and you start to worry about what missions might have been completed when you weren't paying attention to the others. You have to decide whether or not to cut your losses and abort your close observation or just continue following and hope for the best.

(As an aside, if you're wondering why a Sniper can't just watch the entire party at all times, one integral piece of the design is that there's simply too much going on at any one time to focus on all of it. At least, I certainly can't. If there are Snipers who can I think that'll create even bigger problems, but luckily there's an easy tuning knob, the total number of people at the party, that you can turn to increase the information density.)

Unfortunately, the Action Testing undermines these interactions. Now you really don't have anything to go on, because you might be watching the actual Spy and he's finishing missions right under your nose! My theory is that this will make losing a lot less fun, because it introduces that feeling of "there was nothing I could do." You could be watching the actual Spy for the entire party and still lose. Yeah, even the awesome results are still theoretically noticeable, but remember that the stated goal is to allow the Spy to win while being watched, so I can't imagine it'll be easy. There's a question I've been thinking about a lot recently: "Is your game fun to lose?" I think adding Action Testing risks hurting SpyParty a lot here.

You'll also have the games where the Spy flubs an Action Test for whatever reason. At that point, there should be a good chance that you notice him, and win the game on the spot. Here's another relevant question: "Is your game satisfying to win?" Sure, it's always fun to win, but the brilliance of current SpyParty comes in forcing me to interact competitively in novel ways. When I win it's because I picked up a subtle clue, or outthought my opponent. Winning because my opponent or I were good or bad at a reaction/timing mini-game is something I can find in countless other games.

And finally, there are two aspects of human nature that concern me here:

  1. Humans aren't that good at properly evaluating risk, especially when randomness is involved.
  2. Humans tend to overvalue their own abilities.

Both of these point to people trying the Action Tests far more often than they should. The design intent might be to provide a tool for someone under close scrutiny that shouldn't be used otherwise, but my guess is that people are going to use it far more frequently than that. The more people use it, the more the game becomes about the Action Testing and the less it's about its other novel aspects.

With all that said, I love Chris's specific implementation of the active reload system. Randomness in multiple axes should go a long way towards preventing people from mastering it, which was another initial concern. And I'm certainly hoping that I'll be pleasantly surprised tonight. We'll see!

edit: My post-playtest writeup is now up: everything went better than expected.


tiny adventures: depth

Though accessibility is key to social games on Facebook, you can't neglect depth, because you want to keep your players interested in the long run. On the surface, accessibility and depth are at odds with one another, but you can have both if you build your systems carefully. A simple example of how we did this with Tiny Adventures was in giving primary stats to the terrains.

If you recall the way the adventure system worked, each adventure pulled from a set listing of terrains in a specified order. So, we knew that players who had seen an adventure before could figure out what terrains they'd be traveling through. We also knew that with a fully random system, this knowledge wouldn't help players out at all.

To make the terrain types matter, we assigned two primary stats to each terrain. Encounters for those stats would be twice as likely as encounters for any other stat. Savvy players who realized this could pick adventures that were well suited to their character. For example, I would tend to favor adventures with a lot of Dungeon encounters when I played a Wizard, because Dungeons favored Attack Bonus and Intelligence. Within adventures, players could also switch up their equipment based on the upcoming terrain.

A couple reasons why the primary terrain stats worked well:

  1. It cost us literally no development time. All we had to do was skew the distribution of encounters that we were writing. So, rather than ask a writer to make five encounters for each stat, we'd ask them to write eight for Attack Bonus and Intelligence and four for each of the others. Then we just threw them all into a random pool with equal selection chance and everything worked out automatically. (If we'd written an equal number of each and just skewed the probabilities, the primary stat encounters would have come up twice as often, and we'd have more frequent repeats.)
  2. Players who didn't want additional depth could completely ignore it. Many systems that add depth also add complexity. While on many platforms that can be fine, in the social space you want to be careful about overloading players who aren't ready for it. The best methods for adding depth are often under the surface in a way that will only be found by players who dig for it.
  3. It gave players something to discover. There's a lot to be said for straight randomness, but I'm a fan of building in patterns for players to discover over time. I would have been sad if the players had gone through all the trouble to create a stats versus terrains matrix if there hadn't been anything to find. Uncovering the secrets of a system can be incredibly satisfying.
  4. It gave players who were going for high scores a subtle way to improve their odds. I'll be talking about the scoring system later, but taking advantage of the terrain knowledge was one of the ways that hardcore players could improve their scores.

I recognize that this system only added something for a small percentage of players, but those are the types of players that will come back day after day, and spread the word about your game.

What else did we do? We had the generations system, which added character skills to the game, but only for players who'd already retired a couple characters (and therefore were ready for additional complexity). We had the scoring system and leaderboards, which gave players incentive to optimize their characters and show off their skills. We had the fixed story encounters, which were a more obvious way for players to optimize, since the actual encounter was the same each time. I think this functioned nicely as a bridge to the primary terrain stats, in that players would quickly realize they could take advantage of the story encounter knowledge, and then start digging more deeply for other opportunities to get ahead.

Little touches like these helped Tiny Adventures become the eighth most engaging app on Facebook when it was released (and the second most engaging game, pretty much).

Takeaway: Don't neglect depth even when you're building for a more casual platform. Look for systems that add depth in a way that players who don't want the added complexity can safely ignore. If the system is too visible, players will feel compelled to explore it, even if they aren't ready.

Next topic: The scoring system and leaderboards.


great designer search: green in epolith

Taking a quick break from the Tiny Adventures posts to talk about the Great Designer Search 2 again. If you're new to the blog, I worked at Wizards of the Coast for five years, and was heavily involved in Magic design during a portion of that time. Reading through Ethan Fleischer's assignment for this week I felt compelled to blog about my reactions to some of the cards from his set, Epolith. I want to stress that I haven't read any of the article comments or any other thoughts on the topic, so apologies in advance if I'm repeating what's been said elsewhere.


Creature - Horse
Evolve - Whenever a creature with a power greater than CARDNAME's power enters the battlefield under your control, put a +1/+1 counter on CARDNAME.

This is a sexy one drop. Great card to start the set off with, and I'm loving the Evolve mechanic. Rewards you for playing more and bigger creatures but has a built in cap so that it can't get out of control. This is one of my favorite mechanics from the contest so far. The "a" from "a power" can be removed though, and I'd add "you may" to avoid lots of missed trigger disasters during tournaments.

Wandering Elf

Creature — Elf Nomad
When CARDNAME enters the battlefield, you may search your library for a basic land card, reveal it, put it into your hand, then shuffle your library.

Sure. It's unfortunate that there's no obvious reason for this reprint to be in this set over any other (it's more interesting when the card's value changes in the new set or it feels like it naturally fits in somehow), but it's a fine choice regardless.

Cave Bear

Creature - Bear
Evolve - Whenever a creature with a power greater than CARDNAME's power enters the battlefield under your control, put a +1/+1 counter on CARDNAME.

This seems reasonable on the surface, but a tad powerful for common. This is a two drop that's going to grow to 4/4 fairly easily. That said, I'm a fan of giving green commons that you're happy to first pick. It's all too easy for red and black to have all the first picks because removal is so obviously a must have.

My other concern is that a plethora of evolve creatures is going to put a huge emphasis on curving out (playing a one mana creature on turn one, a two mana creature on turn two, etc). Missing your four drop could cost you a 4/3 and three +1/+1 counters, which is pretty swingy. One of the worst things about the VS card game was how heavily the design punished players for missing on-curve drops. They ended up having to print a bunch of cheap creatures that let you tutor so that games weren't decided by which player curved out more effectively. But I digress.

Fertility Shaman

Creature - Elf Shaman
Whenever you cast an Enchantment spell, put a 1/1 green Elf creature token onto the battlefield.

Seems like the set has an enchantment subtheme. It's nice that this creature helps recover the potential card disadvantage of auras in a way that's reasonable at common.

Mischievous Monkey

Creature - Ape
When CARDNAME enters the battlefield, put target artifact into its owner's library third from the top.

The "third from the top" text is cute, but I like it more in blue or white, since they often focus on dealing with problems temporarily and also on planning for the future. Green's permanent destruction should be more final I think. Wizards did show a willingness to do this in black with Lost Hours, but that was in Future Sight, a set with a focus on time and the future.

Leaping Ornitholestes

Creature - Lizard
G, Sacrifice CARDNAME: CARDNAME deals 3 damage to target creature with flying.

Ornitholestes is a mouthful, but it's an actual dinosaur name, so I appreciate that Ethan put in some research and is tying his world back to ours. I enjoy the flavor of a small, agile lizard sacrificing itself in an epic leap to take down a small flier, and it's elegant to line up the damage with the power of the creature.

Wing-grabber Tree Spirit

Creature - Plant Spirit
Evolve - Whenever a creature with flying enters the battlefield under an opponent's control, put a +1/+1 counter on CARDNAME.

This is the point where I realized that Evolve didn't always trigger off of greater power, and to be honest, I found it disappointing. There is something nice about the flavor here, where your creature is evolving to keep up with what's going on across the table, but I'm not looking forward to stalled board states where I have to keep track of five different triggers on my variety of Evolve creatures. Honestly, the flavor of my weaker creatures evolving to catch up with my stronger ones is good enough. I would strongly consider focusing this mechanic on that one trigger.

Elven Memory Keeper

Creature - Elf Shaman
When CARDNAME comes into play, you may return an enchantment card from your graveyard to your hand.

Traditionally this line of text belongs on white cards, but it's not unreasonable to put it in green. Rofellos's Gift and Nature's Spiral have paved the way somewhat. I do wonder how White will interact with enchantments in this set however.

Sluggish Sauropod

Creature - Lizard
Whenever CARDNAME attacks, it doesn't untap during its controller's next untap step.

I think this creature should probably be a 6/5. There aren't any other green commons with toughness above 4, and the drawback is quite significant. I do like the combination of high power with an ability that encourages the opponent to not block and just worry about it later.

Velociraptor Pack

Creature - Lizard

VELOCIRAPTOR ENTRY POINT! A fine card. No complaints.


Creature - Lizard
Evolve - Whenever a land enters the battlefield under your control, put a +1/+1 counter on CARDNAME.
CARDNAME has trample as long as it has a +1/+1 counter on it.

Not a fan of using Evolve as a Landfall retread, and this makes me dislike this implementation even more. It's becoming more clear that Evolve in this set just means "trigger - get a +1/+1 counter", but that's not tight enough to be a compelling mechanic. Also, what is the flavor of this evolving in response to more friendly lands showing up?


Creature - Lizard

I like it. Craw Wurm has been obsolete for years now, so it's fine to print a creature with strictly better stats. This guy is also almost guaranteed to trigger your greater power Evolve creatures, so he'll be better than he looks at first glance.

Nature's Sacrifice

Destroy target artifact or enchantment. Its controller may put a +1/+1 counter on a creature he or she controls.

Great card. It fits in the environment, it's possibly exciting to some number of Legacy or Vintage players, and in a pinch it can be used as a combat trick by blowing up one of your own permanents.

Selection Pressure

Put two +1/+1 counters on target creature.

Seems like a good spot for the common green pump spell.

Heavy Shell

Enchantment - Aura
Enchant creature
You may remove a +1/+1 counter from CARDNAME's target rather than pay CARDNAME's mana cost.
Enchanted creature can't be the target of spells and abilities your opponents control.
: Regenerate enchanted creature.

Great use of an alternate casting cost. Fits the set's theme, isn't anything that's going to be broken in constructed, and has an effect that benefits greatly from occasionally being free. My only critique is that it's too bad it's not slightly more expensive up front so that the alternate cost comes up more often. It might have to do more to justify the added cost though.

...Wait a minute, I just realized this card doesn't have Flash and has no way to get it. My brain simply filled it in because it seems like such a clear fit. I would either add Flash to the card straight up or alternatively just give it Flash if you pay the alternate cost.

Primordial Fog

Prevent all combat damage that would be dealt this turn.
Gain 1 life for each +1/+1 counter on creatures you control.

Not loving this card. I'm on board with having a common green Fog, but the clause feels tacked on to me and some inconsistent life gain honestly isn't worth an extra two mana. I'd consider adding a mana and having it put a +1/+1 counter on each of your creatures instead, as I think that would a pretty interesting spell that (unlike most Fogs) would be playable. Some other changes would have to be made to the set to make that work, such as moving the life gain to the card below.


Search your library for a basic land and put it onto the battlefield tapped, then shuffle your library. You may put a +1/+1 counter on a creature you control.

I would try changing the clause here to "Gain a life for each land you control." This ties in the rider to the initial effect and makes for a card that you're happy to draw late game as well as early.

Venom Glands

Enchantment - Aura
Enchant Creature
You may remove two +1/+1 counters from CARDNAME's target rather than pay CARDNAME's mana cost.
Enchanted creature has deathtouch.

In contrast to the other potentially free Aura, this card needs a lot of work. Let me begin:

  1. Deathtouch is mostly useful on creatures that are smaller than the creature they're fighting. So in the best case, I'm putting this on a small creature and fighting a large creature. But even in that best case, I've just 2 for 1'd myself. This card doesn't have a lot of upside.
  2. This card costs four mana to give my creature Deathtouch. That effect is probably worth around one mana. In the past, for four mana, I could both cast a Gorgon Flail and equip it, getting deathtouch, +1/+1, and an equipment.
  3. The alternate cost is unnecessarily steep. Removing two or more counters should likely be saved for uncommons and rares with strong effects.
  4. The alternate cost should specify "creature you control" somehow. I don't think you want this and other auras like it being used as strange combat tricks that shrink your opponent's creature.

There are some definite mistakes, but the Green commons are in a solid place overall. I'll be surprised if Ethan isn't still in the competition after this week. And finally, here are some quick thoughts about the other competitor's entries:

Jonathan Loucks - Living Reflection could use a new name. It's too much of a mouthful for a mechanic that's going to be used often. Also, Illuminate cards tend to read very poorly. There's so much text before you get to the payoff, which is a bad sign for a common.

Shawn Main - I love Blight, it's my other favorite mechanic from this round of assignements. It's flavorful and should add a lot of tension and uncertainty to the gameplay.

Devon Rule - Gold is cute, but it's a name you probably don't want to use (thanks to Gold being commonly used to refer to multicolor), and basing an entire color around life payments makes me nervous even if you have a mechanic to mitigate it.

Jay Treat - "You and your allies" is a lot of extra text for something that will only really help in Two Headed Giant. The cards that have it do things like granting first strike until end of turn, which won't help your friends unless they're sharing your attack step. If you do want to go that route, you can probably cut the "you or" and just define "your allies" as including yourself.

Scott Van Essen - Not a big fan of Countercast at first glance. Cost reduction mechanics generally only appeal to Spike. I do like that it interfaces with previous sets and seems exciting with cards like Vivid Grove though.

Daniel Williams - I like the fixed version of Showdown. I think Howling Coyote needs a once per turn restriction or more likely a mana cost on the ability. Right now it disables blocking for your opponent entirely.

Jonathan Woodward - The mechanics seem solid. It saddens me that Graforman Explorer is just Sakura-Tribe Elder with a tap restriction. Maybe the land should come into play untapped.

Disagree with something I said? Let me know in the comments!


tiny adventures: accessibility

Even though a game with the Dungeons & Dragons IP is unlikely to break into the mainstream, any game on the Facebook platform needs to have accessibility as a huge priority. People generally aren't looking for hardcore gaming experiences on Facebook. They want something they can ease into, with low time commitment and a gradual learning curve. And since Facebook games are free, it doesn't take much for a player to quit. When a player hasn't invested anything into the game (with a traditional purchased game, they've invested money up front), it's far easier for them to give up on it quickly. On the other hand, social mechanics work in your favor and encourage players to persist through difficulties.

Character creation, something that seems natural to include in any RPG, can actually be a large barrier to accessibility. It asks you to make hugely important decisions at a point when you understand the game the least. That said, it works fairly well for pen and paper D&D, because the DM or other players in the party will often help a new player through the process. But for a single player experience like Tiny Adventures, we decided to just give people a selection of reasonably balanced premade characters that they could choose between. This way, few players ended up making a choice that they later come to regret.

We kept adventuring simple as well. Players just had to choose an adventure and then return in a few hours and read the results. While there was a definite advantage to knowing D&D rules, it wasn't at all necessary. Players weren't asked to choose what skill to use, or choose which enemy to target, or anything along those lines. We were aided by the new 4th edition rules, where characters each focus on a single stat for their abilities, so we just gave every character a single primary stat that increased their attack bonus. Finally, we added tooltips to everything in the character screen that explained how the various bonuses worked.

You're probably thinking right about now that it sounds like we made the game too simple, but there's almost no such thing when it comes to Facebook, as long as you also have depth. I'll be talking about that soon.

Takeaway: Accessibility is key when it comes to Facebook games, and free games in general. Strong accessibility plus solid social mechanics will create the growth that you want.

Next topic: Adding depth while still preserving accessibility.


tiny adventures: the adventure system

With Tiny Adventures we were faced with a question: how do we tell stories, but not tell the exact same ones every time? While we could have simply thrown an endless series of random encounters at the players, part of what makes the D&D experience memorable is the campaign storyline that ties everything together. Yet, we clearly didn't have the resources to emulate a human Dungeon Master.

We chose to create four adventures per level (we had chosen to focus on heroic tier characters only, i.e. level 1 through 10), for a total of forty adventures. Each adventure consisted of a series of encounters, some random and some story driven. To make the random ones feel like they belonged, we chose eight terrain types that essentially functioned as encounter pools. Then we wrote transition text that let the player know when their adventurer was crossing from one terrain type to another. For example, here's the layout for a level 1 adventure:

Difficulty: Level 1
NAME has heard reports of a cabal of dark elves robbing travelers in the nearby Coilspine Mountains. Because of the thick iron door and dangers inside their stronghold, no one has yet cleared them out of there.

The adventure begins: NAME hiked into the Coilspine Mountains to find the Stronghold of the Drow.
Encounter 1: Random level 1 Mountain encounter
Encounter 2: Random level 1 Mountain encounter
Encounter 3: Random level 1 Mountain encounter
The adventure continues: NAME found an enormous iron door set into a rocky cliff not far from the mountain pass where travelers reported the drow raiders. While watching from a safe distance, he/she saw one of the dark elves wriggle out from behind a small rock several meters to one side of the door. After the elf was gone, NAME snuck up to the rock and entered the Stronghold of the Drow using this secret entrance.
Encounter 4: Random level 1 Dungeon encounter
Encounter 5: Random level 1-2 Dungeon encounter
Encounter 6: Random level 1-2 Dungeon encounter
The adventure continues: Deep in the stronghold NAME heard elvish voices. He/She crept up to a door from which they were coming and peeked in through the small window in the top part of the door. Inside he/she saw several drow standing around a table. The tallest one had his back to the door, and he was gesturing while he talked, repeatedly pointing to a map laid out on the table.
Encounter 7: Final story encounter

This adventure consists of three fixed pieces of text, six random encounters with two possible outcomes each, and one final story encounter, also with two possible outcomes. We liked this because the fixed elements succeeded in telling a consistent tale, but the random encounters ensured that the tale would play out differently each time. The terrain types made it feel like the random encounters belonged to that adventure, when in fact they were part of a general terrain pool, because that was the only way we could possibly create enough content to get the variety that we wanted. I felt like we had succeeded with our goal of making the world and the adventures evocative when I saw that one player, Thomas Denagh, had drawn a map of the world based off of the locations that are mentioned in the game.

The final story encounter gave the player closure by wrapping up the story we had set up in the adventure description. In this case, the final story encounter was always an Attack Bonus check with a Magic subtype (there were items that gave bonuses in Magic encounters). If the adventurer succeeded on the check, this text would appear:

NAME ATTACKED the drow in his back with his/her WEAPON. Then he/she kicked the table over, knocking down the drow witch and disrupting her spell. NAME's surprise attack caught the rest of the dark elves flatfooted and he/she easily ATTACKED them with his/her WEAPONTYPE as well.

And if they failed:

NAME ATTACKED with his/her WEAPON but missed, damaging nothing but the map on the table. When the nearest drow stepped aside, he revealed the drow witch across the table from him. She cast a curse on NAME, bringing him/her to his/her knees writhing in agony. They tied NAME up and took him/her to a cell down the hall. After the drow left, NAME picked the lock and escaped. He/She checked the room where they had been meeting, but it was empty -- save for a few gold coins scattered in the corners.

As you can see, encounter text contained numerous variables so that it could be personalized to the adventurer. All gender-specific words had both forms handwritten inside brackets. Each weapon had a specific verb attached to it, as well as a short name because repeating the full name too many times tended to sound bad. The verbs gave us a chance to make the weapons feel distinct, and allowed us to make the powerful high level weapons feel even more powerful by giving them verbs like "decapitated". Finally, they allowed us to at least give a nod to spellcasting without having to branch a lot of the text; we simply used spell names as verbs (like "fireballed") for the weapons that wizards would tend to equip.

We flagged each encounter as a specific level and only pulled in encounters that were appropriate to the adventure (as you can see above, each random encounter was tagged with a specific level range). We could have built generic encounters and scaled them to the appropriate level, but we bit the bullet and went with fixed levels for three major reasons:

  1. It would be lame to get an encounter early in your adventurer's career and then get it again, with the same text but a more difficult check, five levels later. Because we had safeguards against getting the same encounter twice in the same adventure, it was pretty rare to see the same encounter twice in the same playthrough.
  2. We wanted to show character progression, and one of the most effective ways of doing that was having the character encounter goblins and kobolds early on, and storm giants and dragons in the later adventures.
  3. We could write in specific rewards that were level appropriate. So, if the elf prince that your adventurer just saved was wielding a mithral dagger, he could give it to you at the end of the encounter because it's appropriate for that level. If we were scaling the encounter level we would've had to do away with specific rewards.

The hardest thing about writing all of this text? Figuring out how to work a reward into almost every block of text. Because rewards gained through encounters determined an adventurer's score, we needed all of the successes, as well as many failures, to give a reward of either gold and items. It felt disconnected to just hand out a reward if the story didn't mention it, so we had adventurers looting corpses, looking under rocks, grabbing items off of tables, and who knows what else to justify the payouts that we gave them. In the end, though, it was a blast to work on creating the stories, and one of the most gratifying threads on our boards was a huge collection of quotes that players posted that had made them laugh. I wish it were still around so that I could link to it.

Takeaway: With the right mixture of structure and randomness, you can create experiences that both tell a consistent story and still feel fresh when encountered multiple times.

Next topic: Accessibility


tiny adventures: female character choices

One of the goals of the small strike team that created Tiny Adventures was to remain completely free of outside dependencies. Digital projects at Wizards had shown a tendency to get bogged down because they relied on other departments or contractors for various critical components, and since our goal was to create a game a month, we had to keep that to an absolute minimum. On the team we had two designers, a programmer, two producers, a quarter of an art director and a third of an editor (a lot of people at Wizards split their time between multiple projects). Like I mentioned in yesterday's post, we also utilized some contract writers for Tiny Adventures, but that was directly controlled by us and Brandon Bozzi did a great job of keeping everything moving.

Because of the restricted time schedule and insular nature of the team, we had zero ability to generate new art assets. One feature that we wanted to have was male and female versions of all of the character classes. While it would've been trivial from the technical side, it turned out that almost all of the decent D&D art that we had at our disposal showed off male heroes. We had three options: delay the app, put up subpar art, or only give one gender option for each class (with something like 6 male characters and 2 female characters). We chose the latter, foolishly thinking it wouldn't be that big of a deal.

Well, it ended up being probably the number one requested feature, and annoyed people to no end. Because you couldn't choose race and class separately, or choose how to spend your stat points, even players who didn't care about the gender of their character wanted the extra options. To the players who did, the imbalance implied sexism, since players tend not to think about production difficulties, and even if they did, it's hard to imagine a company like Wizards having trouble creating a few small pieces of art.

(As an aside, it's very clever that almost every app puts Alpha or Beta on the end of their title these days. It makes players be a little bit more forgiving, they feel like they're getting in early, and it costs you nothing, especially since it's become accepted to allow purchases during beta.)

Fortunately, the team that was tasked with maintaining the app eventually fixed the imbalance by digging through the archives and finding some quasi-reasonable options, so every class had both a male and a female option. As a bonus, this meant that a lot more races got to share in the spotlight as well, which meant that it was a lot easier for everyone to find a combination that they liked.

Takeaway: If at all possible, have gender balanced character choices. It's easy to underestimate how important this is to a lot of your players. Also, players don't care what hurdles you had to overcome or what prevented you from putting in a feature. They only care about what they're interacting with, which is the end result.

Next topic: The Adventure System, or "How do you tell a story without telling the same one every time?"


tiny adventures: the end game

Tiny Adventures was a content-based game, which was a risky decision for such a small team on such a short schedule. In Nik Davidson's original pitch, the adventurer was sending back postcards to let the player know how she was faring. We all thought this was an evocative way of letting the player interact with their character, especially given the strength of D&D and other tabletop RPGs in telling compelling stories. In the end, it required a creative coordinator, an editor, and a small stable of contract writers (plus other members of the team pitching in heavily) but we managed to create almost a novel's worth of stories that accommodated the hero's gender, class and weapon.

Of course, the scourge of all content based games was still a factor. We needed a plan for the end game. No matter how efficiently you create new content, players will devour it more quickly, even in a game with built in delays like Tiny Adventures. What do players do once they reach maximum level? Fortunately I had been playing a game recently where the player retired his adventurers and each one passed on a single item to the next generation (can't remember the specific game, unfortunately). This seemed like a perfect fit, and thus we sidestepped the classic end game problem by removing it in favor of repeat playthroughs of the main game.

There was only one remaining question, and it concerned me greatly: why should players keep playing?

  1. To see all of the content. We had two plans here. First, we tuned the leveling curve so that we had more adventures, encounters, and items than a player could see in a single playthrough. Second, we added rare encounters and rare artifacts so that even once a player had seen all of the common content, there was still more out there.
  2. To try out the different classes. The classes were decidedly similar at this point, but they did differ in terms of their primary stat and weapon/armor proficiencies, which made them want different equipment. We also added some differentiation with a system that I'll talk about below.
  3. To try to get a high score. More on the scoring system later in the week.

Yet, this still didn't seem like enough. We wanted to add rewards that would stack up as the player retired more and more characters. The simplest solution would be to allow people to take an additional item each time they retired, but this would've quickly made most of the content trivial and would've diminished one of the most exciting aspects of the game (item upgrades). Any system that reduces the number of times a player can find an upgrade has a high bar to cross before it should be added.

The proverbial ace up our sleeve was the generations system, a series of unlocks that were linked to the number of retirements. We added a page that showed the unlocked bonuses and the points at which you'd receive your future unlocks, so that players were always anticipating something. We then designed two abilities for each class that would unlock at the 3rd and 6th generation. The first one was usually a simple attribute boost or heal, and the second could change the game somewhat and allow you to take advantage of advanced knowledge.

The original plan was for the unlocks to extend up to generation 25 (with significant gaps in between), but we never found the time to implement a couple of the later ones. Still, the system seemed to be a success, and was a strong contributing factor to players sticking around through many different characters. The sense of overall progression that it gave was perfect for ameliorating the potential negative feelings from forced character retirement.

Takeaway: With a small team, don't build a content based game if you can help it, but if you do, give the game a definite ending and give players compelling reasons to play through again and again.

Next topic: Female character options.


kinship and the great designer search

If you follow Magic closely, you're probably aware that The Great Designer Search 2 has just announced its eight finalists. If not, you're probably wondering what The Great Designer Search is. Essentially, it's an online "reality show" design competition that Wizards of the Coast runs when they're light on designers. When the first one completed, all four of the top contestants went on to be hired full time by WotC:

Mark Globus – It was a pleasure to work with him on Uncivilized, and he’s a great manager as well. It’s my understanding that he manages a lot of the Magic designers and developers these days, which frees up Aaron Forsythe for work that he finds more interesting.
Graeme Hopkins – A lot of credit for the success of Tiny Adventures belongs to Graeme. He not only coded most of the app singlehandedly, but was invaluable for making little tweaks to the UI that made the player experience better.
Ken Nagle – By the time that Mark Rosewater finally moves on from his head designer position (I'd guess this is at least five years out), I’ll be surprised if Ken isn’t high on the list to take over. It's clear that this man loves Magic.
Alexis Janson – Out of the four, I’ve worked the least with the winner, Alexis, but she’s a competent Magic player and programmer as well.

So, it’s fair to say that the first GDS was a huge success for WotC, and all eight of the new contestants have a good shot at finding employment as a result of this.

The first round was a series of essays, and since I can’t imagine WotC wanted to read all of them, I’m assuming this was mostly to weed out the people who weren’t serious about the whole thing. This was so that they didn’t pollute the next round, which was a multiple choice test that was automatically graded, with some percentage of the top scores moving on. I question the validity of that test in terms of finding strong designers (and there’ve been plenty of complaints about specific questions), but it did weed out people who don’t know Magic well, and some concessions have to be made to make the whole process realistic.

Now, Wizards has posted the essays for the finalists, and while I haven’t read them all, it did catch my eye that Jonathan Woodward’s answer for “worst mechanic in Extended” was kinship. As the designer of kinship and the lead designer of Morningtide (the set where it premiered), this piqued my interest. I’ve always had a soft spot for kinship and thought it was a reasonably successful mechanic for the “class matters” set that Morningtide was. Anyway, here was Jonathan’s reasoning:

"Of all of the mechanics currently in Extended, kinship is one of the worst. If the top card of the controller's library matches a creature type with the kinship creature, the controller gets a beneficial effect. In theory, this is exciting! In practice, however, there are two problems. The first problem is that the mechanic is very linear; instead of playing cards because they're fun, or good, a player needs to play them because they share a creature type with his or her kinship cards."

To me, this seems akin to saying that kicker is a bad mechanic because it’s very modular and therefore doesn’t give deckbuilders any guidance. Linear versus modular is a spectrum where both sides have their place, depending on the goal of that mechanic within the set. Both have advantages and both have disadvantages, but claiming that a mechanic’s linearity is an inherent problem requires either some specific context or the conclusion that all linear mechanics are bad.

"The second problem with kinship is that the card checked is the same card the controller is about to draw. Assuming that the controller has accepted the linear nature of the mechanic, most of the cards in his or her deck that don't share the kinship creature type are likely to be lands. Therefore, once the player has enough lands, failing to have kinship trigger is likely to be followed by failing to draw a useful spell. This leads to players feeling bad; they see the lands they are about to draw, but can't do anything about it."

While it’s absolutely true that kinship makes the highs a little higher and the lows a little lower when you draw your card for the turn, in my experience, the situation that Jonathan describes doesn’t really happen. Because we intentionally moved the trigger to upkeep, players naturally resolve kinship almost as a side effect of their draw. You look at your card, and either reveal it for kinship, or simply put it into your hand if your Kinship triggers missed. If the trigger was during, say, the attack step (as it was for awhile in development), there would be a lot of time to think about the upcoming land, but with it during upkeep I think it works out nicely.

And, for completeness, here’s a bit of history on kinship since I don’t think I wrote about it while at WotC. Sometime during Lorwyn design, we were looking for a way to separate Lorwyn and Morningtide so that each would have a compelling hook. The fact that most Magic creatures had both a race and a class was fairly new at that point, and I suggested that Lorwyn should focus on race interactions and Morningtide should focus on class. Everyone thought this had potential, so we went forward with that plan, and agreed that we would save themes like “sharing a creature type” for Morningtide. Triggering off of sharing a creature type was a great way to make it suddenly matter that your Goblin is also a Warrior, since it shares a type with that Elf Warrior that you also happen to have.

Kinship was originally designed as something like this:

Synergy – CARDNAME gets +1/+1 until end of turn. (Whenever this creature attacks, reveal the top card of your library. The synergy effect triggers once for each creature type it shares with this creature.)

There were quite a few problems:

  • The templating for the mechanic didn’t really work, and if I remember correctly, it wasn’t entirely easy to fix.
  • In order to trigger synergy, you had to attack, but in order to be able to attack, you needed the synergy to work. This often led to players just not attacking since they couldn’t be sure.
  • It was cool that it rewarded you for matching both type and class, but this was at odds with something that I mentioned in passing above, which is that we wanted you to feel just as good about matching up Warriors as you would with matching up Elves. This made it twice as good to match up both race and class, which took the emphasis away from class sharing across races.
  • At first I thought that Changelings revealing other Changelings was awesome, but it became clear that it was awkward that players didn’t really know exactly how many triggers that created (it was equal to the total number of supported creature types, but few people know that number offhand). It also would’ve been an issue for MTGO when that many triggers suddenly popped onto the stack.
  • The aforementioned problem of seeing a land during your attack step and then having to “look forward” to drawing it during your opponent’s turn.

We played with this for awhile, and eventually realized that we needed to simplify the mechanic down to more of an all or nothing based on whether or not you shared any creature types at all. This was a lot better, but of course it didn’t solve everything. The next major breakthrough was to move the trigger to upkeep (I have a vivid memory of discussing this with Forsythe in the parking lot of an IHOP…not sure why I remember that), which put the mechanic into a satisfying place. Finally, the name was changed to the much more appropriate kinship, as synergy was flavorless and could describe hundreds of other mechanics. (I’m mediocre at best at naming. My only success at naming a mechanic that I designed was ripple, which somehow remained unchanged in both effect and name throughout development. Of course, that one deservedly gets some hate as well.)

The final interesting point about kinship was that it was one of those mechanics where I ended up pushing for it despite opposition at various points. I know that Rosewater always has epic stories about how he was the only believer in something, and everyone else hated it, and then it all worked out in the end, but this wasn't really like that. There was just some skepticism amongst the team at various points about whether or not the mechanic would end up working out. I always do my best to be objective in these situations, and there were definitely times that I was worried I was clinging to something that wasn't going anywhere, but once everything fell into place I was glad that I did.

Of course, sometimes you have to give up on something even when you still believe in it. My old article The Color Purple is a good example of that happening during Planar Chaos design.

Well, I know I promised more on Tiny Adventures, but life got in the way and then I wanted to comment on the Great Designer Search. Tiny Adventures part two is still coming soon. Oh, and I don’t mean to hate on Jonathan, he’s just the one whose answers I happened to read. I wish him and the rest of the competitors the best of luck. Can’t wait to follow along!


the social mechanics of d&d: tiny adventures

Towards the end of my tenure at Wizards of the Coast, I had the pleasure of being the lead designer on a Facebook app called Dungeons & Dragons: Tiny Adventures. The game was made by a small strike team that had set out to make a game every month. While we didn't quite hit that on D&D:TA, it only took eight weeks for a team of around six core contributors to bring the game from conception to launch.

With no advertising whatsoever, the game rapidly reached numbers of over 350,000 players, exceeding our wildest expectations given the relative lack of mass market appeal of the D&D brand. It was one of the top ten most engaging apps on Facebook (where engagement is defined as daily users divided by monthly users) as 40% of players came back every day to check on their adventurer. Sadly, due to legal issues surrounding the D&D digital rights, the app is currently not available.

In sharp contrast to many digital game efforts, there were so many things that went right with this project. I'd like to talk about the social mechanics, because I think they were unique and interesting in ways that I haven't seen since. Let me note that I can't take full credit for these; I don't think any of us realized all of the strengths of the system until after the fact, and honestly this was a team effort. We had a playable game up and running in a matter of a couple weeks and we tweaked the mechanics until they felt good.

The social mechanics in Tiny Adventures were fairly simple. Whenever you were on an adventure, you could be buffed by any number of friends. Each buff gave you a +1 to your roll for the next three encounters - however, there was a maximum bonus of +2 on a given encounter (in a d20 based game, this meant a 10% bonus that was significant but not gamebreaking). Between each of your adventures, you could be healed by each of your friends once. Normally players would have to wait to recover their hit points over time, but friends could circumvent that and with enough heals you could head out again immediately.

Having more friends playing Tiny Adventures was always good.

In some social games, you'll eventually find yourself in a position where there's no reason to add more friends, because you've already unlocked all the necessary rewards. In Tiny Adventures, even though only two friends could profitably buff you at once, it was always useful to have more active friends because it increased the percentage of the time that you were at the buff cap. For example, with only two friends who played, you could probably expect to have two buffs on you maybe 10% of the time. Add another friend and that percentage increases significantly, but you'll still need 10 or maybe even 20 friends playing actively before you can expect to be at +2 most of the time. Even once you have a ton of active friends, adding more always increases the uptime just a little bit more, especially when you consider that adventures often run overnight, and it's tough to find people who are awake to buff during those periods.

The buffs were frequent but not too frequent.

Since buffs only lasted for three encounters, and encounters were around 10 minutes apart, players could buff each other about twice an hour. This is often enough that there were usually buffs to refresh if you're checking in frequently, and it contributed to the need to have a lot of friends in order to maintain maximum buff uptime. At the same time, it wasn't frequent enough to be overly annoying, like single encounter buffs would have been.

Players couldn't see how many buffs their friend already had.

The reason this worked well was that it removed a potential source of discouragement. If you could see that your friend already had five active buffs, you might not bother adding another one, since you'd know it wouldn't do anything. But since you couldn't, players got in the habit of buffing all of their friends.

However, players could see who had buffed and healed them rather easily.

The recipients of these actions feel great because they see how many of their friends have taken the time to help them out. This led to feelings of jen being commonplace in the game. Jen is a Confucian concept that Jane McGonigal brought up in a GDC microtalk, meaning "a mixture of humanity, benevolence, and kindness not well captured by any word or phrase in the English language." Designing mechanics and interactions that encourage it (like Left 4 Dead's reliance on your friends to save you from special infected) is a great way to make players come away from your game feeling better about the world.

We intentionally didn't add "buff all" or "heal all" buttons.

Even though this was a frequently requested feature, we chose not to implement it. Instead, we made it incredibly simple to buff each one of your friends, but still made you do it individually. We felt that this was important for making it feel like a directed act of kindness rather than a mechanical action that everyone would ritually click without thinking about it, which would in turn cause them to expect it from others without thinking of it as something special. Additionally, the individual buffs and heals meant that you spent more time on your friends page, which showed information like the name of your friend's hero, their level, their current adventure, and their current score, and kept you invested in their progress.

It was important to keep your friends playing actively.

In some popular social games, all that matters is the number of friends you have who have accepted your friend request inside the game. Once that happens, it's irrelevant if they continue to play or not. It worked in Tiny Adventure's favor that you needed people to be playing actively in order for them to help you. When someone messaged you to ask for a buff, and you log in to give it to them, it's easy at that point to just go on another adventure and get right back into the game.

You didn't have to have any friends at all.

In the early days of games like Mafia Wars, you had to convince friends to play or you literally couldn't progress in the game at all. Thankfully, games have moved past that, but it's still common to cripple a player pretty heavily if they don't feel like spamming their friends with invites. I think we struck a nice balance in Tiny Adventures, where players weren't forced to have friends who played but it did add convenience and a meaningful amount of power.

When you failed an encounter because of lacking buffs, you tended to blame your friends, not the game.

In many social games, when you are penalized due to not having enough friends, it's easy to become angry at the designers or the company for putting in the requirements and/or friend bonuses. But in Tiny Adventures, when you fail an encounter by 1 and you were missing a buff, you didn't tend to hate the game for that. Instead, you were annoyed at your friends for not buffing you, and you would make sure to remind them to keep the buffs coming.

There were distinct reasons to ask friends to help you out right now.

When players had a tough encounter coming up and they saw that they only had a single buff, they would often message a friend and ask to be buffed. Or, when players finished a tough adventure and needed some extra health to start the next one, they could message others and ask for heals. Rather than the app spamming their newsfeeds with posts about how their friends could get a little bit of gold by clicking, Tiny Adventures players were invested enough in the bonuses to ask for help personally. All it took was one or two encounters that were failed because of a missing buff, and everyone understood the importance of helping out.

That ended up being way longer than I intended, but that's my take on the success of the social mechanics in Tiny Adventures. Thanks for reading! There's so much to say about this game, so I'm sure I'll do another blog on this topic at some point in the near future.