A Return to Form

In the following post, I discuss what changes have occurred since my last update in December, the progress I’ve made on the different aspects of my thesis, and a reflection of some of the readings I’ve been doing as of late. 

Introduction

It has finally come time to dust off the old journal and begin putting pen to paper once again to document my journey (well, digitally speaking of course). So then what has been happening since I left for winter break last December you might be asking? Well I’ve taken the past month and a half to do some serious reflection on what I’ve been studying and where I should go next. Through these long hours of pondering while staring at my ceiling, I’ve managed to finally pin down where my studies will eventually lead to (hopefully).

So what's new? Well first and for most it would probably be best to begin with what my thesis is going to be. In the end, I’ve decided that my aim will be to:

Explores how the concept of NPC believability can enhance story generation systems in narrative-driven games to allow for more immersive and engaging narratives. To accomplish this, I am developing a modified version of Short and Evans (2014) system based on the concept of NPC believability which I am calling the StoryGen Engine for the time being. To make this work accessible to designers, I will also create a tool to interact with the StoryGen Engine, and documentation providing guidance on using the tool and the new capabilities it allows for.

Below, you can see a visual breakdown of my thesis topic into the components I will be working on: 

Figure 1: A visualization of my research and 

how the different components relate

Why is this my research topic you ask? Well, my hypotheses are that:

H1: Through using a story generation system based on social practices to determine NPC actions, believability of NPCs will be stronger then when using GOAP or behaviour trees.

H2: Through adding personality modifiers to Versu, the believability of NPCs will be increased

H3: Through increasing NPC believability, Immersion and Game Enjoyment will be increased

Therefore, I have decided upon looking into integrating personality into Versu to be my area of research.

Of course, it should be noted that I still intend to continue my work on breaking down the concept of believability into quantifiable aspects. However, after discussing with my committee, it has been decided that my believability study will become a separate line of inquiry not to be focused on for my thesis.

So yeah, that’s been the largest change to my area of study, so then without further ado lets get into some of the updates with the different aspects of my research.

The Story Gen Engine

So most notably, I’ve spent a good chunk of my winter break recreating Versu based on the documentation I could find online. Using that system, I then made a very, very basic prototype to serve as an example of what this system can do. Additionally, I now can use this prototype as a sorta testing grounds for me to do simple iterations of different ways of writing using this system. If you would like to give this a go, you can try it here.

In addition, my advisor has been having me for the past month perform an exercise in visualizing my findings and my system into diagrams that I can present to others. As such, I’ve now obtained a large amount of visuals for breaking down my research. However, so that this page doesn’t become a long, extended essay on how my story gen engine works, I’ve instead created a separate post you can read if you're interested. [Link to other post, TBW]

In addition though, I’ve also now begun writing my thesis first draft. Through doing so, I hope to better increase my organization of all the concepts I’m learning about and integrating. Also, it will be a good exercise in communicating my ideas in scholarly writing. For this next week I will be

1. Applying the template / format used at my university

2. Writing the first draft of the StoryGen Engine description

Project: Theatre

Okay, next up is Project: Theatre, the newest installment into the Project: Conversation anthology! Although, anthology is definitely a strong way of putting it, more like an attempt to create what Project: Conversation from last semester was meant to be (Oh poor RoboCat, you shall not be missed). This time around however, we are incorporating several of the lessons learned from last time including:

1. Enjoyability: First and foremost, we learned that if the game is not fun, it makes people not want to play it. And if people don’t want to play it, it's really hard to tell if the system is actually useful to game designers. So thus, this time we began with prototyping three different versions of the game to determine which was the most fun while still being related to our thesis. You can actually try these versions here.

   
   

2. Integration of System: So previously, the gossip system I was working on essentially became a side feature of our full game which meant that most people tended to notice it. Additionally, it barely related to what actually was happening so thus, when noticed, often just led to more confusion. Therefore, this time around the StoryGen Engine will be center stage with it serving as the brain of our night phase in our game.

   
   

3. Plot Understandability: Another key change made was that RoboCat landed on the side of being very metaphorical and futuristic which kinda made the plot harder to follow. Often, when people would be playing the game, they’d just get lost during the intro sequence. Therefore, when both Rae and I’s social systems would be encountered, most players were already lost and thus had trouble noticing and / or appreciating the systems. This also resulted in the systems being hard to analyze since most feedback had nothing to do with either of them. Thus this time we decided to GREATLY simplify the story into a more grounded tale of the underdogs trying to fight back against a corrupt system. Or more specifically, a rag tag group of theatre kids trying to get their production to outshine the big name group on campus.

   
   

4. Gameplay Loop: One of the issues we also ran into last time was the core gameplay loop honestly not being that engaging. The biggest reason for this in my opinion was that the gameplay and the story simply were misaligned and lacked a solid direction. The gameplay seemed to want to range anywhere from a social simulation game to a mystery novel while the plot didn’t fare much better. Thus, this time we decided to discuss the gameplay loop early on in the process and are basing it off the game Dispatch so that we aren’t trying to reinvent the wheel.

   
   

So yeah, A NUMBER of changes were made this time around to hopefully allow this game to be received better. We also did a number of structural changes such as now using Linear to track tasks and a large amount of time being spent on pre-production. Finally, we’ve also decided this time to structure the game around a Vertical Slice with a timeline of a semester in hopes of allowing for better evaluation of how well the game is working and the ability to playtest sooner rather than later.

Speaking of pre-production, that has been the main area of work this past week between Rae and I. More specifically, we have been spending a lot of time with a white board trying to map out how our project works and what the core gameplay loop is. The reason for this being that we wanted to make sure that we were conceptualizing the game the same way. Additionally, the hope is that these diagrams will serve as useful tools for explaining the game to others. Once again however, to not let this post go too long, I’ve created a separate post describing the overall flow of Project: Theatre which can be found here! [Link to Post TBW]

Looking into the future, the main questions we still have left to address are:

1. What actions will be able to be taken during the night phase?

2. How will we balance the variability with the story against maintaining a cohesive tale?

3. How many events should players be experiencing during a given night?

4. What can help make this type of gameplay engaging?

   
   

So then what's next? Well, seeing as the game still needs to be made, the next steps for this week will simply be to continue working on the expanded prototype of the game. Hopefully, if all goes according to plan, by the end of this week we should have a “playable” version of the game that we can then begin to iterate on. After doing so, we should then be able to start looking into some of the questions I mentioned above.

So yeah, that’s what has been going on regarding the status of the project. Overall, already I am enjoying this more than last semester’s exploration and I am excited to be able to start sharing the game soon!

Believability Study

While not being the focus of my thesis, I do still find it worth some time to discuss the progress being made regarding this study. I would say the most notable shift that I’ve been exploring from last semester has been a recategorization of the aspects of believability. More specifically, I’ve been trying to redefine believability as a set of primary and secondary components rather than just a list of components. It's not perfect yet, but I do think it helps better conceptualize the concept I am exploring. Below you can see the current version of the diagram.

Figure 2: Believability Diagram

In addition, one of my committee members has been having me read more theories in communication in order to have me exposed to new ways of viewing believability. Specifically, she has had me look at skill and user-avatar bonds (Lynch et al. 2022), the uncanny valley (Gray and Wegner, 2012), objectification (Gray et al. 2011), and warmth / competence (Fiske et al. 2007). Additionally, I’ve also done some reading on how Lee and Heeter (2015) had defined believability across five dimensions based on mental schemas. Furthermore, I spent some time reading up on emotion, cohesion, and consistency with believable agents (Orton, 2003).

This next week, I shall be revisiting my original descriptions of believability to better define how I am viewing these concepts. Additionally, I shall be looking into a paper on media character impression formation (Sanders, 2010).

Conclusion

Phew, sorry for the rollercoaster of updates. Who knew missing a month of posting could get one so far behind. However, now that the semester has started back up again, I shall be resuming my weekly updates on the state of my thesis (hopefully next time with more room to go more into depth with the specific updates). So without further ado, until next time. 

Logging out.

Citations

Evans, R., & Short, E. (2014). Versu—A Simulationist Storytelling System. IEEE Transactions on Computational Intelligence and AI in Games, 6(2), 113–130. https://doi.org/10.1109/TCIAIG.2013.2287297

Fiske, S. T., Cuddy, A. J. C., & Glick, P. (2007). Universal dimensions of social cognition: Warmth and competence. Trends in Cognitive Sciences, 11(2), 77–83. https://doi.org/10.1016/j.tics.2006.11.005

Gray, K., Knobe, J., Sheskin, M., Bloom, P., & Barrett, L. F. (2011). More than a body: Mind perception and the nature of objectification. Journal of Personality and Social Psychology, 101(6), 1207–1220. https://doi.org/10.1037/a0025883

Gray, K., & Wegner, D. M. (2012). Feeling robots and human zombies: Mind perception and the uncanny valley. Cognition, 125(1), 125–130. https://doi.org/10.1016/j.cognition.2012.06.007

Lee, M. S., & Heeter, C. (2015). Cognitive Intervention and Reconciliation—NPC Believability in Single-Player RPGs. International Journal of Role-Playing, (5), 47–65. https://doi.org/10.33063/ijrp.vi5.236

Lynch, T., Matthews, N. L., Gilbert, M., Jones, S., & Freiberger, N. (2022). Explicating How Skill Determines the Qualities of User-Avatar Bonds. Frontiers in Psychology, 13, 713678. https://doi.org/10.3389/fpsyg.2022.713678

Ortony, A. (2003). On Making Believable Emotional Agents Believable. In R. Trappl, P. Petta, & S. Payr (Eds.), Emotions in Humans and Artifacts (pp. 189–212). The MIT Press. https://doi.org/10.7551/mitpress/2705.003.0007

Sanders, M. S. (2010). Making a Good (Bad) Impression: Examining the Cognitive Processes of Disposition Theory to Form a Synthesized Model of Media Character Impression Formation. Communication Theory, 20(2), 147–168. https://doi.org/10.1111/j.1468-2885.2010.01358.x