At last week's Austin Game Developers Conference, veteran designer Chris Bateman spoke in the Writing Track on improving the player experience through story, citing games from Metroid Prime through Discworld Noir. Though International Hobo's Bateman used his own-designed Discworld Noir for much of his talk, a Terry Pratchett-adapted adventure title that is somewhat obscure to North American audiences, his advice on improving player guidance through writing was clear and salient. Looking at disparate, popular games with a variety of different systems for steering the player to their conclusions, Bateman offered a helpful overview of the different story structures and, in some slightly design-oriented moments, techniques that can serve to create a path for players through your games. "There are no rules of game writing. If we're lucky there never will be. Instead, I suggest you see what works and copy those things that work well. Experiment and find what works. And try your ideas on other people. If you just try on yourself, you can easily mislead yourself and write yourself into a corner," was Bateman's advice. "[It's a] common problem -- confusion, resulting from incomplete or unexpected information. Which leads to anxiety, which if it's never resolved leads to anger. We obviously expect that the game writer is going to have a narrative that they want to carry across to the player of the game -- but there are other roles that are equally important." Bateman notes that these include expressing goals to the player and tutoring the player in the game's mechanics. He also said, "Perhaps most difficult [is] anticipating the problems a player will face. You have to do the best you can. What we like to do at our company is examine models of play." The Temperament Theory At this point Bateman introduced the 'Temperament Theory', which governs how people play games. According to Bateman, "Everyone shows the four types I'm going to show here to differing degrees. The strategic player is really quite happy to work through the problems the game puts in front of them. The logistical player wants to know what to do and go off to do it. The tactical player wants to do what needs to be done -- most FPSes rely on tactical play. Lastly, diplomatic play -- 'I'll make everything as it should be.'" He noted diplomatic play is not common in games as yet uncommon from a game design perspective. According to Bateman, research indicates that one-tenth of players fall into the first category, half into the second, a quarter into the third and one-seventh into the final diplomatic category. "Half the players would like to be told what to do, and 10% would like to work it out. This is an inherent contradiction. With a lot of work it is possible to provide both -- you can provide a path through the game anybody can follow, and a lot of options on the side that people can work out. If you are making a game for the mass market, you need to bear in mind this need to tell the player what to do, but if you're making a niche game... you probably have more of a luxury to allow people to work it out." In adventure games, according to the writer, "Clues are provided instead of directions. You're dealing with situations whereby you have puzzles to solve instead of goals to complete [the narrative]." This is basic information, but Bateman tied it into two important concepts. "In a game like Discworld Noir, the clues follow a trail of breadcrumbs... creating a path the player follows through to the end. The other thing you need is a funneling mechanism, something that means that if the player falls off the path, they can get back to it. There's a path to the game -- which is breadcrumbing, and something to get you back on the path, which is funneling." From Turok 2: "There's this idea of a 'golden path'. There are a series of health powerups -- these lead the player through the levels." According to Bateman, one weakness is the fact that the player could collect the times that comprise this path and then fall off it; then it would be gone. Another is that if the player did not collect the "there's nothing to lead you back onto the path if you fall off -- there's no funneling mechanism. Someone who prefers logistical play would give up in frustration." Conversely, in Metroid Prime, "you have a [breadcrumbing] system whereby you earn weapons to unlock doors. While there's a breadcrumbing system, there's also a funneling system... a message will come through where you get a signal which marks a target on the map. WoW or any other MMO leans toward logistical play -- complete the task set. Some might be more problematic, more puzzle-oriented, but you're dealing with a list of tasks. You have quests and a quest journal. Here you have your breadcrumbing and your funnelling. The game writer or the game designer or the team as a whole," must implement these systems, according to Bateman. "There are all sorts of techniques you can use for breadcrumbling. You can use hints or clues, just nudge the player. As a lot of FPSes do you can build your game in a tunnel, where there's nothing to do but go forward -- which is fine for a style of games. It's more than sufficient for [FPS fans] for them to play through it. You can of course use dialogue for breadcrumbing -- and you can combine the clues and the dialogue. Or the more expensive version of dialogue, cutscenes. You can be a little more subtle, use signs and markers. If someone's looking for a monster that knocks down fences, you can have a series of fences knocked down. You can use natural choke-points. Even in an open world you can have natural choke points." An example is a mountain pass. "You have an opportunity to have events occur in this point." The Spine of the Game? Bateman continued: "I call this 'The Spine of the Game'. These are the events the player must progress through to finish the game -- it's rare that a player must play through all the events to finish it. But to get to the end the player must play through all of these events. Each part of this spine requires a complete trail of breadcrumbs. Most modern games have one spine -- there's one fixed major path through the game. But it's perfectly possible to have multiple spines. And these create specific problems for game writers." "A linear spine is a very efficient way to make games. It's cost-effective" and, he suggested, much easier to write for. "If you're completely insane, you can approach a branching structure. If you hate your company why don't you suggest a game with a branching structure? Most players are going to play through the game once or twice. It's too much work for what it is." The need for a player to make arbitrary decisions is "just nonsense. It's not entertaining," Bateman claimed. "A game like Deus Ex," has parallel paths, Bateman put forth, "where the events combine and return to a central path. But it also has side routes. You can draw equal or different attention to the different routes. You've got this option here to point the player in a direction but have little intriguing things off to the side -- which gives the player more of a sense of agency. Again, it's slightly illusory because you're going to combine them back at the end. This is a reasonable compromise." In the threaded structure, which Bateman used for Discworld Noir, "there are these little separate vignettes that seem unrelated" and "the player can come to these at different times." In the game, a player that missed a plot point in the first chapter would have it forced on him by an NPC in the second chapter. "The players often feel like they have more agency than they do because they have more of a chance to investigate things. The important thing was that there were some nice ideas here but it didn't work. I was the only one on the game project who understood the game structure, so if I had left the project everyone would have been screwed." "A sort of dynamic object oriented approach -- imagine what's going on here -- you structure your game more like a TV show." Bateman posited, as an in-his-opinion untested alternative. Essentially, episodes and scenes -- which can link together in different ways. "It's a lot less wasteful than branching and it's a lot more tenable. But publishers don't want to pay for it, is the bottom line -- and the number of players who are going to truly appreciate it are in the minority. But the vocal hardcore sell the games to everyone else." In Bateman's view, if a developer could harness this form in a way that a mainstream audience could understand, it would work. "It's possible that you could use a [Mario 64-like non-linear, location-based] platform game system to hang this structure on." Exploring Narrative Techniques "You want to save your cutscenes for things that can only be delivered in a cutscene -- where it would be just a nightmare delivered in play," is Bateman's feeling. "Another good tip is to graph the game spine. If it's linear that's kind of a waste of time, but if it's got any element that comes out of a linear section, it's worth having a diagram to show how these events connect." He continued: "You want someone in the programming team and the art team to understand this" as well as the people who work with it more directly. "Group related parts of the narrative into threads -- if you can break the story down to acts or threads that makes a complete section of the narrative, it makes breadcrumbing much easier. You just have to connect the threads together to complete the narrative." Bateman then came back to funneling. "You can't control [players] and they will do the most imaginably terrible thing to break your game." A journal or notebook is a "nice, cost-effective way" of doing offering funneling. "Another approach is to mark the map. It's not enormously narrative, but it's pretty hard to go wrong with." An active compass, which points the player, is a permutation of the marked map, it was suggested. He also recommended trying guiding characters, and events that find the player, no matter the path he or she takes. He gave an example -- The Legend of Zelda: Ocarina of Time's Navi the fairy -- who annoyingly tells you the next plot point while you're trying to accomplish (and enjoy) something else. He observed that Nintendo abandoned this with Wind Waker, for the most part, but many games still use this irritating tactic. Another less annoying example is Castlevania: Aria of Sorrow's Mina, who hangs out next to the game's store, which you'll visit often and can ask for help. "Often the player knows what you'll tell them in the hint, but they're stuck on something else. It's difficult to anticipate," Bateman observed. Of course, he pointed out, the Grand Theft Auto series uses hotspots on the map and the compass, which is "hybrid breadcrumbing and funneling -- it's efficient and clear, but there's room for improvement from a narrative perspective." When asked if more breadcrumbing reduces the need for funneling, Bateman agreed. "I think that's the case. There's less of a need for funneling support" if the path is easy to follow.