Aristotle was not a Game Designer
This week gamesindustry.biz published an interesting article by legendary game designer Warren Spector, in which he talks about universal rules for game design. Even though the text doesn’t contain completely new findings or revolutionary statements, he addresses one problem that I find is worth some more attention. In his last point Spector writes about the issue of “time” in narrative games:
Obviously games can tell stories. So, let’s start from the premise that all that Aristotle stuff applies to us. Agreed? Okay, now let’s talk about the one narrative problem Aristotle didn’t talk about – the one we have to solve that other media don’t. I call it the Act 2 Problem.
First off: I don’t think that Aristotle took games into consideration when he wrote about the structure of ancient Greek plays. Doesn’t it make a huge difference, whether the audience watches a play or movie or reads a book passively or if they actively take part in the story? This, for me, is still one of the biggest problems when it comes to story in games. We try to impose established story structures on games without properly adapting them to our medium.
We’re fine setting up a story (Act 1). And we’re pretty good at ending one (Act 3). We do denouement well enough. Our beginnings and endings tend to be fairly linear and brief.
Are we really good at this? And if so, why? Because we rely on techniques of another medium, namely movies, to present a stunning cutscene? From my head I can remember exactly one really good beginning and one really good ending that were not just cutscenes: The beginning of The Last of Us and the ending of Red Dead Redemption. Of course there are some more beginnings and endings that are quite ok, but to me it feels like cheating when most games don’t even try to bring some interactivity to these particular parts. So in my humble opinion we as an industry still have a lot to improve to create intros and endings that embrace the key strengths of our medium.
But the main problem, and that’s where I totally agree with Spector, is the part in between:
But Act 2? The part of the story where, having established the hero’s problem and gotten him up a tree you throw narrative rocks at the poor schmo? That part, we’re not so good at. And we have trouble with that for one simple reason, I think:
For some reason I have never been able to understand why players expect games to fill up 15 to 100 hours of their lives. No other medium is like that. Even a short game is the equivalent – in commitment of time on the part of the user – with the average television season. Think about that – a single game is roughly equivalent to an entire season of television.
This could be one of the reasons why so many players never see the end of a game. We lose their interest in this never-ending second act. Personally, I hate it to not finish a game that I started. But I get so bored by games like Assassin’s Creed – that stretch their narrative over dozens of hours without letting the plot make relevant progress – that I did not finish any of the series’ last installments.
While I fully agree with Spector, I think it’s kind of funny that he doesn’t see an answer for the problem, even though he mentions one briefly. He compares the scope of games with seasons in television. And this comparison hits the nail on the head! Due to their expected length games are much closer to TV series than to the average 90 minutes Hollywood movie or a typical stage play. Which brings us back to the three act structure and Aristotle: It simply isn’t made for stories of this length! Television has found much better structures for this purpose over the last decades. Daily soaps and prime time serials like Lost, Breaking Bad and many others are using them with great success. So, what’s the difference to the three act structure?
Modern television serials use some kind of tiered narrative structure. Episodes have self-contained story arcs, that bring a regular element of closure , while also providing series-long arcs and multi-episode arcs that join several episodes together.
This model makes the act two problem completely irrelevant as it basically offers an act one, act two and act three in every single episode. Instead of building up tension for that one climax after 40 hours of gameplay, this structure would offer a climax for nearly every play session. This needs much more work on the narrative side (that’s why television series usually have a large team of writers), but it would fit the time schedule of the average player so much better. Telltale already proved that this narrative structure can work for games pretty well, even though they didn’t dare to go beyond mini-series of at most five episodes.
Just as disclaimer: I don’t think that all games should be delivered episodically. They don’t have to be published in episodes to take advantage of a serial story structure. Breaking Bad, Lost and all the others are also extremely entertaining when buying all seasons as a bundle and watching them one after the other.
So, all I’m saying is: if we want to learn from established story structures, we should learn from the ones that fit our medium best. Games deserve more than just three acts! Nobody (at least nobody I know) plays through 50+ hours of content in one or two marathon sessions. Players consume games in small chunks anyway, so it would make a lot of sense to establish narrative structures that support this behavior. And by the way: Levels, missions or quests are already structures that go into that direction. It just often feels like they are randomly scattered throughout the game world and not linked to each other or the overarching plot in any way. But methods for linking story arcs together, for providing closure and climax points regularly are already out there – just look at them a bit closer the next time you enjoy your favorite television serial.