GDC: Indie Keynote - Championing Immediacy And Depth

Tiger Style co-founder Randy Smith (Spider: The Secret of Bryce Manor) delivered the keynote of the Indie Games Summit, encouraging indie developers to embrace a philosophy of immediacy and depth to hasten their popular ascendancy. Opening with a self-deprecating "I do have a worry that this presentation will not live up to expectations or achieve what it could," Smith outlined that his background may not be "indie" enough for the crowd -- having worked for EA and Ion Storm, and "in bed" with Apple thanks to the success of his iPhone hit Spider. Of course, he joked, the Kokoromi crew has been "coaching" him on how to be more indie. According to Smith, the indie scene is producing a lot of cool games that just aren't living to their potential. "So many games had these great ideas but weren't working for various reasons," says Smith. His tactic is "immediacy with depth", which he warns is "just one game design lens; not all games need it." In other words, a game should immediately communicate "why this game is so cool," and then it should reveal "why to keep playing after that wears off." When discussing immediacy, says Smith, "I think this invokes a specter or paranoia about dumbing things down. I want to be clear that I am not trying to changing indie games in that respect." He recognizes that it's a mainstream developer concept -- but as he has before, he espoused bringing some of these methods into indie development -- where they'll work. "What does it mean to bring these triple-A techniques into indie gaming? I'm sure I have a different answer than some folks," says Smith. 
 In mainstream development there is a sense that "you need to learn the rules so you know how to break them." However, this usually means "you need to stay inside the rules. There's a lot of paint-by-numbers game design." However, when it comes to using these concepts, "in indie you're very agnostic, you start where it makes sense for you. I don't think indie equates to bad game design, we take these rules where we're going to start and apply them where they're going to work." As he reiterated several times, "the emphasis is on a more powerful indie through better design." Smith went through a survey of many games that exemplified the characteristics he was looking for. For example, Canabalt has simple controls that make it very immediate, and "you fail a lot but it's easy to get right back in, so the cost for failure is pretty low." Firemint's popular Flight Control is also "very immediate" -- "there is really great feedback on screen." With this game, "this idea of how much fun you get out of the game, depending on how much you put into it, is really important," says Smith. Games have "affordances", says Smith -- defined as what you can do in them. Games with strong affordances are more appealing, says Smith. He adds that "you can qualify affordances according to immediacy and depth." This ties into the idea of "juicy gameplay" espoused by 2D Boy (World of Goo) developer Kyle Gabler. "This is a really awesome way to make your affordances more satisfying." Lots of output for minimal input "helps your interactive stuff feel much more alive," says Smith. Sound and visual effects are prime examples of how to make games "juicier" without affecting gameplay directly. So are things like "a frequent stream of rewards" during gameplay -- which, though Smith did not explicitly say so, are two major tenets of mainstream development. Though he likes the game a great deal, Smith called out Scribblenauts 20-plus minute tutorial as not good for immediacy. "They thought it was best that every single player went through this long training structure, but it's kind of the opposite -- as it has this barrier to entry problem." Instead, says Smith, you should explain the basic controls and then "try to set up the game so that the player learns the controls for themselves instead of telling them what to do." This style of development produces what Smith calls "the game-toy", a game that offers "the smallest possible package that's fun to play with, that's resting on a sea of depth." His own Spider is a simple platforming-style game that's easy to understand, but there's lots of depth, e.g. 38 levels, 24 achievements, other modes. Games developed in this style, such as I Dig It for the iPhone, are however notably criticized because "the core gameplay is always the same. This comment is also applicable to Spider, and it's one we agree with too. One of the shortcomings of the game-toy with depth approach is that gameplay tends to feel the same all the time." 
 Players being rewarded for failure is a "really smart mechanic," says Smith, as in Hook Champ, where resources you earn before death carry over to your next attempt at the level. Some games use the "depth on demand" approach, where there is plenty of content that isn't more difficult than what has been released before -- a number of levels, for example. Says Smith, "The threshold for success is pretty low... But if your player is interested in mastery and completion there's more to do." It's worth noting that "it's very important to invite the player into depth" by showing or telling them that there's more to achieve. Players didn't realize Spider had other gameplay modes because the initial version didn't do a good job of informing them, says Smith. They were told when they completed the game, but "less than 50% of the players actually win." One mechanism for adding depth is analyzing the game's "low and mid-level game loops", says Smith. "The low-level loops are the things you are doing all the time in the game. It's pretty tightly connected to the controls and the most basic affordances." On the other hand, "mid-level loops [are] how you think about what you want to do with those powers." In the case of Spider it's "not how to build a web, but which web you want to build. It elevates your mind up one level from the basic thing." One way to add depth is by delivering "simple rules that lead to interesting consequences" in gameplay -- such as in Galcon, which is easy to understand and, through its random levels, allows interactions to play out quickly. "Shuffled content means that there is no single level that you can learn, but you need to master the gameplay itself," says Smith. 
 Spider offers Xbox-style achievements. "On the design side we're very interested in making sure that these achievements worked toward expelling the different ways you can play our game" -- for example achievements that might rarely occur through normal gameplay point toward different ways to play, and also encourage attention to the achievement system. "Lots of different games in the indie space have simple, really good concepts but there's not a lot of good reasons to play them after 10 minutes," says Smith. Smith pointed out two "awesome indie game which could benefit from this analysis" -- Farbs' Captain Forever "has this really fantastic strong juicy affordance that you can construct your ships. Part of the magic of Captain Forever is the ship construction assembly mechanic that really pays off" and "there's a lot of room for player creativity." However, it's not all rosy -- immediacy is harmed because "if you're to pick a ship at random, that ship would probably be kind of crappy... towards the beginning of the game [when you have to learn]... There's a lot of failure that happens, which tends to escalate into more failure." The game's core dogfighting "tends to be a little repetitive... There's actually less depth there than you'd expect." 
 Smith thinks that "making that experimentation phase more forgiving would be great..." as "the game keeps pushing you through the phases... Making it more clear" to the player would be a helpful improvement. The game could also offer "more reason that [different approaches to] ship construction would make more of a difference." Environmental objects, like asteroids you could hide in and snipe, special weapons, resources in environment, and big and interesting events, like a giant swarm of very small ships, are Smith's suggestions. He also spoke about Derek Yu's Spelunky, which he described as his favorite indie game. "It's an extremely immediate game... [with] simple controls inherited from the platforming gameplay style." However, "really where Spelunky shines is the depth."
 There are lots of different activities in the levels, says Smith. You might see "a sacrificial altar, and think it must do something, and you will be right." The randomly-generated "Rogue magic" -- a reference to the randomly-generated RPG -- "has been applied to a 2D platformer," says Smith. "You don't get to memorize the level and you have to master the systems." 
 However, the game has "fussy, primitive controls that don't do much to help the player to drive his guy around." There is also "tons of player failure, especially at first for a new player." In Smith's view, roguelikes fail "when you know what you need to do, kind of, but you can't quite get through." Work on controls and difficulty and pacing would make Spelunky much improved, says Smith.