Running my first game of Sleep is Death

As Tracy wrote, we played our first game of Sleep is Death last night. I think it went really well, and I definitely plan to tell more stories with it. (I also hope to find someone who will tell me a story with it, but this is admittedly pretty demanding.) You can read the whole thing here. There were some hiccups where I misused the software slightly or misjudged how long something would take me to do (it's really interesting but also fairly stressful to create art on-the-fly for a game) and these resulted in a few continuity errors, some more embarrassing than others. These are a part of the experience, though -- SiD is a performance on the part of the controller and the player, and the mistakes become a part of the performance. Some of the mistakes are even sort of interesting if you take them as part of the narrative -- others merely serve as reminders that there are people behind all of the images on the screen. (Before we go any further, if you need a refresher on what SiD is, start here.)

Broadly speaking, the experience was surprisingly emotionally engaging. I had been working for the past week on creating the assets I would need to tell the story, putting in half an hour here and an hour there building characters, rooms, objects, and other art. The longer I worked on it the harder it was to believe it would ever come together in a story. And there's a real risk with this game that it won't -- if you go on building your story's world forever, if you don't finally play a game with someone and make it all live and breathe for a moment, I imagine you can get lost in the details. The limitations (the time limit, the turn-taking structure, the needs and agency of your player) are what will make the story happen. They give you structure and sense. The goal should probably be to do world-building as quickly as you possibly can while doing a decent job (with an eye toward creating assets that you can use in many, many stories) so that you can get into the meat of player interaction.

Of course a big part of the story's success was Tracy's performance as a player. Her gifts for character and dialog as a writer not only created Mr. Sagara, they made him an engaging person. I felt real sadness when he began disconnecting the phones, and my decision to make him cut his hand while disconnecting the phone in the closet built on that feeling. There's a sense of isolation in the game -- a weird loneliness that comes from knowing this whole little world is being built and sustained by just two people, for a very limited time -- that can be, if modulated by the story correctly, quite devastating. I saw glimpses of that. There is also, of course, a real warmth in the fact that you're interacting so intimately with another person. It's hard sometimes when writing to remember how important it is to be generous to your reader. That never becomes a problem in SiD, because your "reader" is right there, with a very real voice in the story's creation; the kindness fostered between controller and player is a very good feeling.

I suspect you could learn a lot about writing, reading, and art in general by playing SiD.

Below the fold, I'll make some more specific observations about how SiD works, and how those who play it can build a better experience.


There are some principles I worked from while planning the game and others I discovered in playing it that I think will generally lead to more satisfying SiD stories.

• When building your world, don't try to create a convincing or complete environment. You don't want to spend the next two weeks building a game world that you'll play in for about an hour. If your story takes place in a house and you don't want it to leave the house, don't build a front door. If the kitchen and the bathroom won't figure in your story in an important way, leave out the kitchen and the bathroom. Your player will have enough on his or her mind that they probably won't even feel conscious of it -- and to the extent that they are conscious, they may be grateful for the limitations. Many players describe SiD as a somewhat stressful experience in that they're always worrying about making the "right" choice. You want to help them feel free in ways that reduce this anxiety, and one way to reduce that anxiety is to reduce the number of options. Too much freedom, or too many elements, can be overwhelming. 
• Relatedly, remember that you're playing a game, and in games incentives matter. If you put your player into certain mindsets, you'll find it counter-productive for storytelling purposes. For instance, one of the main styles of play to which gamers are accustomed is the "artwork acquirement" style. Think about how Gabe from Penny Arcade says he mainly plays video games to unlock more and more art. For some games, this is an effective system. You do NOT want your players approaching SiD from this perspective, though (you haven't got the time to make it worth their while), and unfortunately the form sort of inherently encourages it: one of the main tensions when a player is making a decision is the awareness that the controller either has or has not created art to represent his actions, which can lead to exploring or making decisions purely to see more art (or to test the controller). If you have a manageable environment with limited objects and art, good players will quickly adjust and assume a more storytelling-conducive stance.
• Limitations also make the player's decisions more meaningful. Players on sidtube often seem to approach the game from the perspective that since they can type anything into the verb box, they can actually do anything. This is a bad way to tell a story (and a real stress on the controller). Instead, encourage your player to see the world in terms of complicated potential interactions between a few simple objects wherever possible.
• Character should emerge from the player's decisions as much as possible, but you also want to spur the player to think creatively. Some controllers seem to begin games by asking the player's name in a way that encourages them to insert themselves into the story. This seems like a boring strategy! My preference is to force the player to imagine herself as someone outside her own body, mind, and experiences -- thus my decision to start Tracy out as a naked fat guy. This way you can still let the player define the character (his past, his name, his mood, his speaking style, etc.) while pushing her to imagine something beyond herself. This sort of collaborative characterization (creating a visual game world that interacts with the player's textual input and actions to define character) is one of the really exciting things about playing SiD.
• Give your player a set of competing impulses and choices from the very beginning of the story. Tracy's character Mr. Sagara starts the game nude in a room full of telephones. The telephones soon begin to ring. There are clothes scattered on the floor. Tracy has to negotiate several competing urges at once: does she pick up the phone, put on the clothes, or something else, first? Then what does she do? The choices she makes, and the order in which she makes them, define her character, create player agency, and set the tone for the game. I've seen too many stories on sidtube where the player goes for far too long without actually making any real choices.
• Give the player an out. There's a real possibility that the player will get bored before the game is supposed to finish. They may however feel it's rude to say so. (I know I would.) I was nervous about this while I played with Tracy, but I realized I had accidentally given her an out: the huge hole in what was once Mr. Sagara's bedroom provided a dramatically coherent way for Tracy to end the game essentially at will. All she had to do was go to the hole and jump inside: I would take this as a sign that I should cut the game short, but it would be an interesting enough decision and feel sufficiently relevant to the character that no one would have to feel cheated. This isn't probably always possible, and it needn't always be so dramatic, but where possible I think controllers should provide characters an option to end the game at will without violating the narrative logic of the game.
• Always have multiple endings in mind. An SiD story doesn't need to end with a bang or a demonstration of your prowess as a controller -- rather, it should emerge as the result of negotiation between player and controller. You want to have a few rough sketches for endgames in mind at the outset (and the relevant assets created) but in a way the entirety of a Sleep is Death game is all about negotiating the way it's going to end. The worst SiD stories I've seen were clearly meant to bring the player from point A to point B, which seems to rather defeat the point. 
• General level design advice: Don't always fill the whole screen with tiles. This makes it look like you are making fewer decisions than you could be making -- good, deliberate composition is tough in SiD, but the more you can make it clear that things look the way they do because you intended them that way, the better they'll feel for the player. The room editor seems to subtly discourage not filling the entire screen with tiles but negative space was extremely important to how I created my scenes and I think people would probably benefit from using it more consciously. By that same token, try to limit your color pallet for a given story as much as you can. This also makes the art + world look more deliberate, less accidental, and therefore more engaging.
• As a broader note, I think that probably surrealism is the natural mode of Sleep is Death. Could be wrong, but that's how it feels to me. Maybe more on that later.