Off The Script: The Problem With Storytelling In Open-World Games

Niko Bellic quietly contemplates the futility of war, of the violent life he has chosen. He mourns the friends stolen from him by his enemies, the tragic losses he has accrued through violence and crime. In the end, his experiences have taught him only this: that the American dream is an illusion, impossible for anyone to achieve.

The next day, he steps out onto his balcony and fires indiscriminately at the streets below, trying to see how many cops he can get to follow him. He spends five minutes carefully throwing grenades, trying to get the angle just right to put one under a car. Then, he takes the ruined cars, shoves them into a big pile, and ramps his motorcycle over it.

Niko Bellic's a complicated guy. Which is to say, he's the protagonist of an open-world game.

"Open-world" games came to prominence with the adventures of one of Niko's predecessors in 2001's Grand Theft Auto III. Obviously, games (especially RPGs) have been presenting players with large worlds to explore for years, but GTA III was one of the first to allow us to interact with and explore a large 3D environment in real-time*.

I love open-world games. Love the freedom they allow me, love the ability to let my id run rampant, love the thrill of exploring beautiful, interesting worlds. I'll play them happily for hours. There's just one problem, and, for a nerd with my particular interests, it's a big one. Because the vast majority of open-world games are terrible at telling a story.

It comes down to a simple question: How do you effectively tell a controlled narrative when your game's entire philosophy is built around letting the player do what they want?

If you're Rockstar, and most of their ilk, the answer is Missions. Tightly scripted narrative missions, where your amazing game world becomes nothing more than the backdrop for a shooter (or driver or brawler) level. Where you go out of your way to curtail player freedom so that he or she can't "break" the flow of your narrative.

In case you couldn't tell, I don't care for scripted missions. They downplay the genre's strengths (freedom, exploration, player choice) and emphasize weaknesses (see: any Grand Theft Auto game that forces you to shoot for extended periods). Even when missions are well-designed or allow the player to do cool things, they emphasize what CAN'T be done during unscripted gameplay. Worse than that, they lead to scenarios like the one outlined above, putting the player's portrayal of the protagonist directly at odds with the designer's. In a novel, everything a character does is part of who they are - in Grand Theft Auto IV, the character's actions only "count" when the designers dictate them.

In essence, missions divide open-world games in two - there's the wild anarchy of the sandbox portions, where the player is given complete control of who their protagonist is and what he or she** does. And then there are the missions, where the designers tell the story THEY have chosen to tell, through cutscenes, required player actions, and scripted events. The protagonists of these two games share a skin and (usually) the same controls... but that's it. Instead of using their open worlds to tell a story, these games deliver the story as a separate gameplay experience.


Generally, when developers try to combat these problems, they do so by curtailing the freedom of the sandbox portions. This sounds bad, but it works surprisingly well at times. Rockstar Vancouver's Bully, which takes place at a boarding school, limits the amount of violence and destruction protagonist Jimmy Hopkins can mete out (both by punishing certain actions, and flatly disallowing others). By cutting off actions like violence against young children, Rockstar a) quiets moral watchdogs, b) establishes fundamental, playable aspects of Jimmy's character, and c) sets a tone for the game - one significantly lighter than that of their flagship series. Essentially, Rockstar uses Bully's limitations on freedom to enhance the storytelling by cutting off avenues the player could take it down if given more choice.


This can be taken too far, of course - there's a point of limitation where the question becomes "Why did we build an open world in the first place?" Case in point - Team Bondi's fascinating-but-flawed L.A. Noire. The game features a beautiful, vibrant recreation of 1940s Los Angeles - and precious little to do in it. The player's entire time in the sandbox portions of the game is spent driving to locations related to cases, or answering calls (which generally lead to ill-advised shooter segments). The player is controlling a by-the-book L.A. detective - one so by the book, it leaves no room for freedom or fun. The game is all mission, no sandbox, an open-world in name only.


Developers have taken the balance to the other extreme, too. One of my favorite open-world games, Crackdown, developed by Realtime Worlds, forgoes missions entirely. Instead, the player is given a set of objectives (read: people) to accomplish (read: kill). Optional objectives are available to make these goals easier to achieve, and the game strongly suggests an order to follow to make things easier, but most of the game's choices are firmly in the player's hands. There is no scripting (and, unfortuantely, precious little story). The game feels, in many ways, like a trial run (albeit an extremely fun one) for a more interesting game, one where the player's choice of what areas of the world to explore and engage with informs the story***.


Lastly, there are those open-world games that embrace the destructive anarchy of player freedom, that tune their narratives to work synergistically with the chaos and mayhem the average player unleashes when given the chance. First and foremost among these are the Saints Row games, whose plots and missions strive to be crazier and more over-the-top than the stunts and nonsense that the players get up to when left to their own devices. Unlike Niko Bellic, there is no disconnect between the behavior of Saints Row: The Third's protagonist, "The Boss," in scripted missions or open sandbox play - he (or she, or it) is always an energetically sociopathic mass-murderer.


Obviously, this approach limits the kinds of stories that can be told - but it HAS been used to great effect to tell interesting stories about bad people. Far Cry 2, for instance, uses player freedom to force you into the shoes of a potentially amoral mercenary in war-torn Africa. If your character is a heartless killer, it's because YOU chose to play him that way. By allowing the player to be a monster, it more effectively throws the consequences of monstrous behavior in the player's face in a way that would be significantly less interesting in a more scripted game.

Video games allow the player's choices to affect the course of the story in ways unprecedented by other formats. Open-world games in particular embrace freedom. I look forward to seeing games utilize that freedom in the way they tell their narrative, instead of confining it to the sandbox.




*I'm glossing over a lot of games here, because GTA III was the first game to pull this off successfully on a console - opening the genre to the vast majority of players. Bethesda, for instance, published Daggerfall, which contains a huuuuuuge (and mostly featureless) open-world in 1996, and the very weird Terminator (described, entertainingly as always, by PC Gamer columnist Richard Corbett here) in 1990.



**I wracked my memory for examples of female sandbox game protagonists, but the pickings were pretty slim. The only ones that come to mind are (optionally) The Boss in the second and third Saints Row games, and Jade from Beyond Good and Evil****. If I'm leaving anyone out, shoot me an e-mail or leave me a comment.



***The S.T.A.L.K.E.R. games generally also fall under this category*****, which made me wonder... Why are so many open-world games set in third-person? If I had to guess, I'd assume it has to do with the genre's console roots (at least as it's thought of today), and the discomfort in previous console generations with first-person views. But there also might be the fact that sandbox games emphasize exploration, and it's easier, I think, to engage with environments if you can see the figure you're moving through them. Just a random thought.



****EDIT: Twitter chum @patchworkearth, who posts excellent game criticism at the site for his soon-to-be-published, certain-to-be-awesome web comic Project: Ballad, pointed out quite rightly that Beyond Good and Evil hews much more closely to the Zelda model than the GTA one. Which begs the question: is a game that features a large, open, explorable world, with the player funneled into meticulously designed dungeons, a "sandbox" game? Or does the term refer more specifically (at least, in the methods I've been using in this article) to the urban exploration games that descend from Grand Theft Auto III? (Bonus reading: The slightly-surprising Wikipedia page for "Grand Theft Auto clone").



*****Okay, maybe I'm stretching a little bit here to include S.T.A.L.K.E.R., since they break the game's world up into individual zones. But I think they still count, since they're non-linear real-time 3D games based on exploring areas. Which.. I'm just going to keep expanding this genre until it includes Skyrim. And Super Mario Galaxy. And everything, ever. Then, and only then, will I be free.

But, seriously, the goal of this essay was to look at how you handle telling a story in a genre built around non-linearity. If I strayed from dead center of that topic to look at how other games handled the problems, or how they reflect on the sandbox genre... I can live with that.