SpyParty dev details his Blizzard-inspired 'depth first' approach

Speaking to a packed room at New York University's Game Center last week, developer Chris Hecker -- a man perhaps best known for giving ... impassioned rants -- detailed his "depth first, accessibility later" approach to development of his latest project, SpyParty. The game is a twist on the Turing Test: one player is "the Spy" while the other is "the Sniper." The Spy must complete a set of objectives without being spotted, while the Sniper looks on and tries to pick out who the Spy is from a group of NPCs (and then murder that Spy) before the time runs out.

How does the sniper spot the spy? By paying attention to a variety of "tells" -- from the subtle (a human Spy's order of actions may differ from an NPC's) to the "hard" (catch the Spy covertly slip an object to an NPC). As Hecker is keen to point out, SpyParty is a game about human interaction. "You have to decide where you're going, go there and don't look back (basically). Of course, I also make the NPCs fidget occasionally, just to fuck with people," Hecker revealed to a laughing audience. "And that's interesting -- that interplay ... I mean, it's an inverse Turing Test at a certain level."
%Gallery-107897%

As the game is currently in production (with a couple of years left to go), Hecker broke down the current state of SpyParty into three boxes: "The Already Working," "The Has Potential" and "The Who-the-Hell-Knows." He characterizes the main Spy mechanic -- "playing it cool" -- as "The Already Working." Where things start getting hairy is high-level play. As Hecker explained, he wants "e-Sports level of player skill" in his game. The problem is that, when playing against a good enough Sniper, no Spy (no matter how good they are) could complete their objectives if the Sniper was looking directly at the Spy.

"The tells are too obvious -- they're not obvious when you're not looking for them, or you're looking in the wrong direction -- if the Sniper is looking right at the Spy, the Spy's not gonna be able to bug the ambassador (or whatever)," Hecker observed. "And this is a problem because one of the core things about the game is I want to turn player skill to 11." His solution, outside of tuning the subtlety (or lack thereof, in some instances) of player animations, is to potentially add a "small player-skill challenge to completing missions more subtly" -- a metagame, for instance, along the lines of Gears of War's active-reload mechanic. This possible answer "Has Potential," but is still very much in the concept phase, Hecker added.

Moving on to the "Who-the-Hell-Knows?" portion of his game, Hecker dissected "what [SpyParty] is saying through its interactivity." Primary among its messages: making consequential decisions with partial information. Beyond the obvious (shooting a party guest as the Sniper or attempting to complete an objective as the Spy), he went into the minutiae behind high-level player interaction. One example is the Spy's ability to engage a mission very early on that puts him at a high risk of getting caught: the Spy must speak with a double agent to hand off a package within a small social circle, causing both the Spy and Sniper to audibly hear the secret words, "banana bread." This reduces the number of possible spies greatly, but only if you're aware of what's going on. Players new to the Sniper are -- for obvious reasons -- often shocked upon hearing the words, and react as such, giving the Spy an early advantage.

This plays to the aforementioned "e-Sports" level of depth Hecker hopes to achieve with his game, repeatedly stating as much during his talk. "I'll eventually dial it back with handicapping ... you can always make things more accessible, but I wanna go as deep as I possibly can on this mechanic and this concept before I start actually dialing it back."

His approach to the game's development, he said, was inspired by Rob Pardo, Blizzard's chief game design executive. "Rob Pardo gave this talk at GDC Austin in 2008 [ed. note: he meant 2006] about their different kind of design sensibilities and just how Blizzard goes about doing things. And he made this point about 'depth first, accessibility later,' and I was right in the middle of Spore and we were doing it the other way. We were making sure it was very accessible -- and the fact of the matter is we didn't really ever get the depth there. The Creature Creator, actually, I think has depth -- but we actually did that depth first. We had a really impossibly hard creature creator, that could make arbitrary creatures, and then we backed off from that and made it accessible. Whereas the actual game mechanics, we didn't do that, and I think it shows in the game we ended up shipping. And so, I wanted to not make that mistake again. That laser-like focus is, in some sense, a reaction to that."

Before fielding questions from attendees, Hecker trounced the entire room in a live demonstration of SpyParty, as one audience representative did his best for us (unfortunately to no avail). But then, Hecker is a high-level player of the game. At least, that's what I imagine other participants were telling themselves to feel better about the loss, like I was.

NYU Game Center director Frank Lantz, who's also the creative director at Area/Code, developer of the wonderful Drop7, closed the evening out by inviting attendees to join him at -- where else? -- BabyCastles for the "after party." And though Hecker noted that this would be the last time we'd see his game in public for a while, he mentioned during his presentation a possible appearance at GDC in late February. With any luck, SpyParty won't stay incognito for too long.%Gallery-108049%