Starry Heavens, the fourth and latest physical game I have created with Nathalie Pozzi, premiers at MoMA in a few days, in the ARCADE event that Kill Screen magazine curated. I am sure there will be plenty of documentation of the final game, so before all of that hits, I wanted to post something about the process of making the game.
The design process started about six months ago – as soon as we heard about the possibility of being part of the event. I knew it was going to be a crowded, one-night, party-like exhibition and I wanted to create a game that could work specifically for that social context. On the other hand, this was the Museum of Modern Art, so Nathalie and I wanted a game that would be sufficiently spectacular, in terms of space design, overall visuals, gameplay, and cultural smarts.
The first prototype was a tabletop boardgame. A random number generator created a list of numbers, which I read out loud. If you heard the number on which you were standing, you could move your piece. The goal was to surround other players and oust them from the board. The game felt a bit like a strategic version of Conway’s Game of Life – you were part of a process and flow, making temporary alliances with others to team up and kick out other players. It felt a little random, but also fun in a real-time Pachinko-machine kind of way. I was calling the game “Numbers.”
One of the impulses behind the design was to solve what I call the live-action time problem. In a large-scale physical game that is real-time (like a sport), even a non-contact game quickly becomes a contest of speed and reflexes, and direct or indirect body checking and rough physical play quickly becomes the way to win. In a turn-based game (in which each player takes his or her own turn), a careful or strategic player can take a very long time to decide on a move, slowing down the game for everyone.
The reason for the random number pulse was to set a steady pace for the game and keep things moving. The idea was that an audio recording (or potentially, a real-time output of a random number generator) would pace the game, keep players alert and listening, and give the overall experience a rhythmic beat. It was a middle path for solving the live-action time problem.
Playtests of this game worked fairly well. The steady beat of the numbers turned the game into a kind of dance, with some portion of players all moving on each beat. The early versions of the game didn’t have any win condition, and players simply presumed that their goal was to capture everyone else. I liked the idea that the strategically ambiguous game rules only provided a process and that the players themselves provided the winning condition.
There were some problems with the idea of players moving simultaneously. What happened when players moved into the same space? Could players “game” the count and wait to see where others were moving? We developed elaborate “bumping” rules that let players displace each other across the grid, but it always slowed down the game, which was going to be a problem if there was an automatic count going on. The game was supposed to be about flow, but it often locked up in local disputes about what was supposed to happen next. Everyone would fall behind an automatic count.
And something larger was missing from the game design. The game was too shallow – players needed short and long-term goals, and although I loved the authoritarian feeling of “waiting for your number to be called” it was more important for me that players were engaged and challenged with the game qua game rather than the game being intentionally un-engaging to serve a larger conceptual or aesthetic purpose. I think that’s what makes me a nerdy, classicist game designer, rather than an artist who is using games to make conceptual statements. The craft of my discipline is important to me. It’s possible for the twain to meet (the occasional Surrealist or Situationist game such as the original Exquisite Corpse exercise has “elegant design”) but it’s exceedingly rare. Most Fluxus games, for example, are so eager to dethrone the notion of traditional game authority that they happily dispense with the tightness required of a bona fide game design. Again, not to say that game-ish conceptual art activities shouldn’t be attempted – they’re just not my cup of tea.
The game continued to evolve – on tabletops and in large-scale playtests. Numbers become colors, and Nathalie and I called the game Black, White, and Gray. We tried many ways to incorporate deeper strategy and play – most of them involving additional game elements. Players might start the game with a handful of objects and try to win by getting rid of them during a capture – or in other versions they would begin empty-handed and try to collect a certain number of objects. The “winner” got to be the one to call out Black, White, or Gray – until the next player who reached the victory conditions took their place.
Oh yes – we were trying out using players as the counter instead of an automated count. We actually did try playing audio files during playtesting as well, but as I feared, it was too easy to fall behind the computer count, especially as players were learning the game. (And given our party context, there would always be new players joining, causing this same problem.) I liked the idea of getting rid of an automated computer/AV system in any case. All of the games I had made with Nathalie were completely non-digital, and it felt right to keep computers and electronics out of this game too.
But overall, things still weren’t clicking. The game felt casual and less-than-compelling, and we were reluctant to add a physical economy to the game. What were the objects that players were collecting, holding, and giving away? Where would they be stored onsite – in some kind of container? Where did they go after play? We wanted the physical space of the game to be clean and simple. And in addition, we were running into lots of problems with game flow – the action sequence (including bumping) was still awkward and ambiguous much of the time.
The big breakthrough came when dissatisfaction created a number of simultaneous rules changes for a playtest at the Eyebeam museum of digital art. I should say that we had been playtesting the game at every opportunity, at least once per week when I wasn’t traveling. This was challenging, since our game required many players (at least 8 or 10) and also a large physical space. Eyebeam has a monthly game designer playtest event (organized by Kaho Abe and Greg Trefry), Josh Debonis runs a monthly board game designers group, and we made good use of both (the photo above is from Eyebeam). Nathalie and I also organized our own playtests – at friends’ dinner events, at the NYU Game Center, any time we thought we’d have half a dozen or more people in a space where we could lay down some masking tape. Part of creating games like Starry Heaven is instrumentalizing your social life in order to create playtesting opportunities. I am blessed with a shamelessness that makes this possible.
The breakthrough playtest dispensed with object-passing, making players count their captures on their fingers instead. It also placed the “caller” player at the center of the playfield grid, and the overall goal was to make your way in to the center to replace the caller and become the caller, lasting there as long as possible until the next player arrived and took your place. Some of the lines in the grid had extra thickness, requiring players to have captured a certain number of other players before they could cross those lines.
This version of the game felt right. Yes, it was full of timing, layout, and balance issues, but it was the most enjoyable one yet. Players actually had goals. They could slowly accumulate captures, making their way gradually through the concentric “levels” of the grid, gradually earning their way into the center. It felt a bit like a slow-motion live-action videogame.
Just as important, the game supported players being kicked out (through capturing) and then re-entering at will, at any time. The ability for players to enter at any time (as opposed to the more traditional experience of waiting for a game to finish and then joining a new game when it started) was going to work really well for our party context. (It’s actually a quietly innovative aspect of our game design.) The older versions of the game supported this kind of continuous re-entry mechanic, but they never worked as well. In the new design, the central “caller” player now had a clear rolling timeframe – you call Black, White, or Gray until someone (inevitably) replaces you. So the idea of a flowing game structure, in which players passed through the game space, changing roles even as the roles themselves remain constant, made more sense.
It was around this time that Nathalie and I found out that the event in which our game was going to appear was moved up in time by a month. Suddenly, instead of two and a half months before our new game premiered, we only had about six weeks – and we were going to be traveling for fully half of that time.
Our playtesting went into overdrive. We squeezed in one more playtest before the trip, and planned to do more after we got back. During the trip, we worked to refine the game content and aesthetics, now that we had (or so we thought) the game design under control. Nathalie wanted to do more than a grid on the floor, and she convinced me that helium-filled weather balloons would let her activate the space dramatically. I am always afraid of balloons in games (since they can signify kids’ birthday parties) but I trusted Nathalie’s space design instincts. Our first idea was a single huge balloon (20 feet in diameter) that the caller player slowly pulled down, trying to reach it before being kicked out, although this idea gradually morphed into several balloons (including a 12 foot central balloon for the caller to pull) that helped lead new players into the space.
The balloons created a kind of “cloud cover” above the metallic game grid (Nathalie was planning on steel plates for the game spaces) and this prompted Nathalie to mention a Kant quotation: “The starry heavens above, the moral law within,” that gave rise to our title. We had already been exploring the idea of an abstract political/moral fable as the game content – the central player had become the “Ruler” and the other players were “Subjects.” The idea of an absurd monarch reaching upwards to the clouds while the citizens of the state circled like sharks, plotting the ruler’s removal, felt right. Starry Heavens stuck, and Nathalie began sourcing our physical materials in earnest – with only a month until the premiere.
Returning from our trip, we were going to have two and a half weeks until the show, and we launched into playtest overdrive. In ten days we held no less than five full-scale playtests, in addition to tabletop testing. I probably burned trhough my Twitter and Facebook playtester possibilities for the next several months, but it was worth it. Dozens of playtesters showed up and each time we played the game, it evolved for the better.
The pictures that go along with this blog post cover nearly half a year. In some ways, the initial concepts behind the game didn’t radically change – core design ideas around time, space, and collaboration were there from the start. Yet at the same time, so many aspects of the game went in for revision and tuning that it is hard to remember everything we tried. During this final push, we worked not only on the grid layout and the game logic, but also the game content, the play rituals (for example, how exactly you physically perform a capture), and the rules themselves.
About the written rules: Although the MoMA event was only going to last 3 hours, and we were going to have multiple assistants on hand to explain the game, we still wanted a supertight rules presentation that could explain the game without anyone else being present. Some players simply like to read the rules first, and if Starry Heavens was going to have a life beyond its premiere, it would likely be displayed in a context that would require more explanatory self-sufficiency on the part of the game. For each playtest, we had players read the rules first and try and figure out the game (asking repeat playtesters to pretend they didn’t know anything). Painfully witnessing how difficult it was to understand the game just from the rules helped us evolve them in terms of text and graphics. The rules were their own design problem that evolved in parallel with the game, easily going through 30 or 40 iterations.
These final playtests produced a kind of design high for me. As we refined the Ruler’s role, the way that captures accumulated, game terminology, ending conditions, and so many other things, I was struck by how active and important our players’ suggestions and ideas were for the game. Nathalie and I tried our own ideas, but we also tried out many players’ ideas too – often they were better than our ideas, and we incorporated them into the design.
During a playtest, a player might call out, “I don’t know what to do when this happens” and I would ask, “What do you think should happen? What feels right?” Nathalie and I would let players brainstorm solutions to design problems and would probe them to tell us what was and wasn’t working – what concepts felt confusing and unintuitive, as well as what could still possibly be fun after 60 or 90 minutes of playtesting. Many of our playtesters were designers themselves, but many were not. Both kinds were useful for identifying problems and solutions.
Mostly, this process really helped me understand one of the things I love most about being a designer. Yes, Nathalie and I were authors, conceiving and executing an original work that would embody our ideas and sensibility. Yet we also were very much collaborators with our players, keeping our egos down to zero as we saw our “brilliant” ideas dashed to bits on the merciless reality of game players demanding a working game.
Even after a game is made, it continues to be a collaboration with and among players, who will (hopefully) find ways to play that the game creators never anticipated. But this collaboration is even more true during the design process itself. Game designers that don’t take full advantage of the tough love of player collaboration are missing out – not only for their games, but for themselves.
A deep thanks to all my playtesters. See you at the premiere!
being playful 
