Interactive Drama
From CS260SP09
Readings
"Narrative Intelligence" Michael Mateas and Phoebe Sengers, American Association for Artificial Intelligence Technical Report, 1999.
"Interactive drama on computer beyond linear narrative" Nicolas Szilas, from AAAI Fall Symposium on Narrative Intelligence, 1999.
"Structuring content in the Facade interactive drama architecture" Michael Mateas and Andrew Stern, from Proceedings of the Conference on Artificial Intelligence and Interactive Digital Entertainment, 2005.
Supplementary Readings
"An Oz-centric review of interactive drama and believable agents" Michael Mateas, Lecture Notes in Computer Science, Springer Publishing, 1999.
Discussions
Please post your critiques/commments on the required readings below. To do that, first login by using your user name and password, then click the "edit" tab on the top part of this page (between the "discussion" page and the "history" page), New to wikis? Read the Wiki editing guide. . Hint - Please put a whole line == ~~~~ == (literally) at the beginning of your submitted critique, so the wiki system will index, sign and date your submission automatically.
KetrinaYim 21:54, 7 April 2009 (PDT)
Facade and other explorations of computerized interactive drama could open up new venues in game design. Currently, narrative and interactivity seem to exist in a tradeoff type of relationship. Games with a high degree of narrative, such as point-and-click adventures and the Metal Gear Solid series, don't offer a whole lot in terms of interactivity. Either the game consists of minimal player involvement (just clicking), or the interactivity is broken up by non-interactive interludes. Indeed, the Metal Gear Solid games consist of so many narrative cutscenes that it may feel more like watching a very long movie than playing a game. On the other hand, games with a high degree of interactivity, like the Grand Theft Auto series, have narratives that play only a minor role. So much freedom is given to the player that it's possible for them to ignore or miss the story. Regardless of where a game falls in the spectrum, it is often the case that the player will feel that he or she has little control over the story itself. The story establishes the frame for the player's actions, but in the end he or she is just playing along. The interactive drama presented in the papers could offer a new gaming experience--one where the player can have a significant impact on the characters and narrative and feel immersed in the story, instead of being a spectator with some degree of control over a protagonist.
Anuj 00:59, 8 April 2009 (PDT)
With regards to what Ketrina said above, I think what is even more interesting is the popularity of games based on various such patterns. While games like GTA have a loosely knit storyline and allow the user to play to his own whim, other really successful games like Myst have a story that bounds the behavior patterns of the player. Moreover, in games like Myst care is taken that there are unique endings to the game depending on the player behavior through the game. Also, games like Max Payne blend drama, narrative and interactivity together to create a successful game. I think future efforts in this direction should probably involve creation of decision trees in game based on user behavior, which means that the flow of the game should be loosely defined to begin with, but it should adapt and expand based on user behavior.
Himanshu Sharma 06:39, 8 April 2009 (PDT)
Ketrina and Anuj have already covered most of my views. Nicolas' study discussed the model on which a computer system of interactive drama could be built. Facade architecture gives more idea on how to actually go about creating the interactive drama. Though I am still a bit skeptical about the granularity with which story could be allowed to to be interactive from technological point of view but these papers seem to give a sense of possibility to this idea, which on my first thoughts seemed impossible.
Brian Tran 09:23, 8 April 2009 (PDT)
I agree with Anuj that there are games today that enjoy both interaction and narrative. Many games have FMVs that are the same quality as their playable sequences even if the FMVs could feature more impressive graphics because the game developers want to blend the transition between story and gameplay. They want the user to stay within the context of the game and not feel like they're watching a game. Metal Gear Solid, Halo, Legend of Zelda, and latter versions of the Resident Evil series employ this. Going off with Anuj said about decision trees, a lot of the most popular games today feature decision trees where the way that NPCs interact with you and your end mission is based on the decisions that you make. Jade Empire and Elder Scrools feature this.
Kenrick Kin 09:54, 8 April 2009 (PDT)
I'm curious as to how successful something like Facade would be as entertainment. I think currently we have games that sit closely on opposite ends of Facade. As others have mentioned we have RPGs that contain multiple endings and nonlinear storytelling, where the user can influence the eventual outcome of the hero. There are also online communities such as Second Life where you can assume a role and interact with other virtual characters who are controlled by real people. And of course there are MMORPGS, where the story is scripted, but your interactions with other players are not. Facade tries to walk this line between storytelling and user interference or interaction to generate a unique story. Will people accept this as a game or find emotional resonance with the story or just merely try to find ways to break it? I went to a talk by Jordan Mechner, the creator of Prince of Persia, and his thoughts on story is that story is largely unimportant and it's just a wrapper to service gameplay. I'm sure not everyone agrees with this, but for Facade, I'm wondering how one might break down the importance of story and gameplay. Here's the trailer for Facade: http://www.youtube.com/watch?v=GmuLV9eMTkg
David (Tavi) Nathanson 10:56, 8 April 2009 (PDT)
When reading about the use of narrative as a basis for HCI, and that in oral storytelling "information such as names and lists are embedded within the storyline such that the audience experiences this information as events unfolding in time," it occurred to me that this concept of events unfolding in time could be applied to debugging. When debugging an application, the execution of that application unfolds in time. Typical debuggers that I've used have lists of variables that change based what part of the program is executing, but I am imagining a timeline view of the variables that illustrates which ones are created and which ones are destroyed as the execution progresses.
Priyanka Reddy 14:44, 8 April 2009 (PDT)
I found Facade to be really interesting - I've never seen such an application before. One thing that I found to be most interesting was how the characters had quirks built into them that would allow the program to get around any technical difficulties - in this case, not knowing how to respond to something the user says. Making the characters self-absorbed makes the characters much more interesting and entertaining and doesn't take away from the game.
However, one concern I have about Facade is that it seems to be focused most on the story and trying to understand more about the characters personalities and history. It seems a lot like a TV show but the user has to draw all the information out of the characters. This might not be as appealing to those who watch TV as they might prefer to not do any work to get the story. On the other hand, this might also not be appealing to gamers who want to know what their progress is through the game rather than focusing on character development.
Nicholas Kong 07:50, 9 April 2009 (PDT)
Very late on this; CHI is so exhausting! When reading the Façade paper, I was reminded quite strongly about the (seemingly) uniquely Japanese game genre of the "visual novel". These are essentially digital versions of the Choose Your Own Adventure books: a number of set storylines which the player navigates through via choices made in-game. Façade seems to be attempting to enable something much more natural, though: "interactive drama" vs. "literary drama".
I was thinking about arenas other than game design in which these dramaturgical principles could be applied. The motivation for this research was that storytelling is almost a human compulsion, so any future humanistic intelligent agent will need to have storytelling mastery. In this sense, any area in which an intelligent agent might be useful would benefit from storytelling acumen; perhaps persuasive technologies could be enhanced, for example.
Simon Tan 12:14, 15 April 2009 (PDT)
I've actually tried the Facade game before, and I did find it to be a very unique experience compared to most computer games. It's very clever how close they built the system as a metaphor of a real drama, including the concept of 'beats' and 'beat mix-ins'. The authors admit that a lot of the cost of making this game was in the architecture, and it would be very interesting to see what would happen if they created many more beats for the system.
The ability to interject input whenever you wanted definitely contributed to the realism of the conversation. However, the character animations were kind of jerky and the natural language processing component seemed slow. One interesting aspect of the game was how the characters moved around - they would walk towards you, away from you, towards each other, or even storm angrily out of the room. This reminded me of proxemics, the metrics that determine what makes for 'personal' versus 'public' distance and how people interact with each other at these distances. (A tidbit from CHI 2009.)
