Question: What are the key areas in which video games need to be improved in order to tell powerful stories more effectively? 

I break “story” in video games into three distinct but overlapping elements: The “game narrative,” which includes the plot, the world, and the characters; the “story-telling,” which covers all the techniques used to convey the narrative; and the “player story,” which is the experience as described by one player to another and is wholly unique to games. Anything that could improve upon all three of these elements simultaneously is most likely to have the biggest impact on “story” in games.

Among the most obvious areas that meet this criteria are AI and animation, both of which allow us to create more engaging, surprising, responsive and relatable characters. Relatable characters make a plot more than just a series of events; I start to care about the stakes involved because I care about the characters. Responsive and engaging characters give the worlds we build a sense of verisimilitude, no matter how bizarre the setting; and a well-rendered and well-realized character can be an excellent story-teller as well. As we push further into AI and away from scripted encounters, we can give characters quirks and unpredictable – but understandable – behaviors that will allow them to surprise players and create more individualized player stories, while also making them more “human” and thus even more relatable.

When it comes to AI and animation, we like to dream big, envisioning Westworld-like interactions and Turing-level conversations with NPCs. But, there are thousands of smaller steps we can take now that would all improve “story.” We might improve scheduling or task systems so that NPCs have more flexible routines that can evolve over time or react more effectively to players. We could create more powerful macro-systems that allow an antagonist to relentlessly pursue an end goal, but change tactics based on player choices and actions. Even something as simple as giving characters different animation overlays based on the situation – the shopkeeper is relaxed and casual when addressing a Paladin, but keeps his arms crossed and jaw set when talking to the Thief – would improve all three elements of “story.”

Machine learning is another area that could dramatically improve story in games. There are AI applications in this area, such NPCs developing new behaviors or tactics over time based not just on interactions with one player but with all players. However, more immediately, we should be able to use machine learning to create more interesting assets more efficiently. This could allow us to produce more detailed and immersive worlds for players to explore; or experiment with a wider range of art styles to create more memorable settings. If we truly unleash machine learning for asset creation, we might end up with terrains that are wildly alien, yet still believable because they adhere to an internal logic. It’s conceivable that machine learning could also allow a game to make sweeping changes to environments and characters in very little time, which would in turn allow player choices to impact game worlds in more meaningful ways – transform a desert into a forest with the wave of a wand, or allow the cat people to rebuild a village you’ve already cleared of goblins. One day, machine learning might also make it possible for anyone to play a game in his or her native language without publishers spending huge sums on localization.

Finally, we need to embrace the idea that the players are co-authors of the experience, and innovate on this across all three aspects of “story.” Hundreds of games allow players to be co-authors in the player story by supporting multiple play-styles, branching level design, parametric loot and spawning, an array of weapons, and other features. But, beyond branching dialogue and binary choices at key moments, it is rare that players are actually co-authoring the narrative in any meaningful way; while games that allow players to co-create the world and characters or shape the story-telling techniques are even more elusive. It is my hope is that we one day see games that are essentially story sandboxes where the developers have provided the backbone, rules, and tools (including core gameplay) that players can then explore to craft compelling narratives, all of which are told in ways that are most interesting and engaging to them personally. If we can one day achieve this, then every game story and every player story will be truly unique.

Don't miss Haden's talk in the Design Track on Tuesday 9 July -  Everything Tells a Story
Find out more...