Process Document Conrad Bassett-Bouchard Emily Porat Jayanth Prathipati Kate Carey Riva Fouzdar Sina Siddiqi
Sections 1 Who Did What 2 Storyboards 3 Wireframes + IxD 4 Character Art 5 Environments 6 Animation 7 Dialogue 8 Playtesting 9 Development 10 What Worked 11 What Didn t Work 12 Future Improvements
Opening scene from Wolfe Goes to School. Dark, yet playful.
Who Did What
Team Wolf is: Conrad Bassett-Bouchard (Design lead) Emily Porat (Research Lead) Jayanth Prathipati (Development lead) Kate Carey (Design + research lead) Riva Fouzdar (Research Lead) Sina Siddiqi (Development lead) Animation, brand, game design, interaction design, process documentation, wireframes Character development, dialogue, playtesting Project manager, Twitch integration, Unity programming Animation, brand, concept documentation, environment, game design, narrative, playtesting, storyboards Character development, dialogue Concept creator, game producer, Unity programming
Storyboards
Storyboards were a crucial part of defining the game as a team. We used the storyboards to further refine the narrative and to split up asset creation between art, interaction, and development. After deciding on the structure of the game, we needed to better understand the sequence of events in order to divide up the work. We storyboarded all the cutscenes and the first round of dialogue to visualize how the game would look. These storyboards were early definitions the sequence of the cutscenes, game interactions, and visual style of the game.
Wireframes / Interaction Design
Video game design in six weeks is no small task. Audience participation games (APGs) rely on Twitch chat to handle both audience chat and interactions with the interface. Twitch streamers complain that APG interactions clog up the chat channel. Although time constraints led us to keep the input of votes in the chat, we did decide to focus on displaying information on the screen whenever possible, since a major part of the game is to inspire audience discussion in the Twitch chat. The game narrative takes place onscreen throughout the game, while the class attendance sheet is prominently displayed to show which students are still around.
We also wanted to make the game feel interactive, even if only the streamer was able to control the actual game interface. One way of accomplishing this task was to have vote tallies change in real time. Although each audience member only gets one vote, they are allowed to revote and overwrite their original choice.
Finally, we wanted to make Twitch commands simple by focusing on the recognition over recall usability heuristic. It made sense to leverage human intuition by associating the students faces with real names at every opportunity so that audience members would identify more strongly with the characters. This way, when audience members needed reminders of past answers, it would be natural to refer to students by name, as opposed to a job title or a player number.
Character Art
Despite its dark storyline, Wolf Goes to School is designed to be a playful game. The character art was created keeping in mind that dichotomous nature. The characters are in a very grave situations, and have professional identities and quirky personalities, despite clearly being schoolchildren. Alix, the gender-neutral mad scientist child.
Environments
The environments for Wolf Goes to School set the tone for the whole game. We understood early on that the concept of a wolf eating school children would have to be playful, rather than gruesome. The visual style of the environments was inspired by children s storybooks, with childish colors and playful textures. Using the storyboards, we quickly understood what environments needed to be created: a classroom, a playground with tetherball court, and a forest path. As development went on, we had to add variations on each environment in order to reflect the narrative and address technical concerns. For example, the playground
needed to be shown at night while the kids tie up the student who was voted as the wolf, and we needed another version in the daytime with the pool of blood for the scene the morning after.
Dialogue
Writing the dialogue for this project was truly an iterative process. In order to cut down on the volume of dialogue we needed to write, we initially attempted to set the wolf to be the same character in every playthrough of the game However, this made the game only playable once through - once the audience figured out the wolf was the same kid each game, there would be no point in playing again. Our next set of dialogue created clues that pointed to one specific character every round, regardless of the wolf s identity. This allowed us to create dialogue that purposely implicated one of the kids, influencing who was chosen each round. During playtesting, however, we found the audience often misinter- preted the clues, causing them to vote for the wrong character (skewing the remaining rounds). Finally, we decided on a way to create dialogue that has the kids equally throwing blame on each other, without fixing the wolf or who dies each round. We did this by assigning each character a set of dialogue that is either neutral, implicates him/ herself, or implicates a primary character they are blaming. If that character is already gone, we assigned each character two additional characters to blame. By organizing the dialogue this way, we were able to randomly throw accusations at various characters, essentially confusing the audience and increasing the overall effectiveness of the dialogue. It was an elegant solution, given our time constraints.
Kid doctor Greg s Round 1 dialogue.
Animation
Our designers had never animated before, so it was a big challenge going into the project. We used Adobe After Effects for computer keyframe animations. These animations are seen throughout the game, as well as during the cut scenes. Some triumphs included animating an effective opening scene, as well as animating a moving mouth with appropriately-moving teeth. With more time, we d like to go back and draw silhouettes as vector graphics instead of.png images in order to make more convincing walking animations. A sprite sheet of mouth animations for Bruno, the hipster artist child.
Playtesting
We were able to hold two playtesting sessions. The first session took place after creating about one third of the dialogue, and we read out the character responses aloud rather than having a digital game. There were four players total, and a few things became apparent immediately. First, the dialogue greatly contributed to playtester enjoyment. The players enjoyed the humorous bits of clever dialogue and had no problem conversing. At this point, however, we were planning on structuring the dialogue in a tree, with hundreds of different options for each decision. After this playtest, we realized that we had to downsize the dialogue creation deliverables.
We held a second playtesting session after completing all of the dialogue. We ran the game through a Twitch channel and copy-pasted responses from the dialogue document, simulating what actual play would be like (minus the graphics). One takeaway from this session was that people seemed hesitant to contribute at first, but chatted more as the game progressed and enjoyed the game, overall. Additionally, players did not like the response randomness - they wanted more pointed dialogue hinting at the wolf s identity.
Development
Coming into the project, the team developers were both newbies to Unity. Software development on this project was really challenging and rewarding. Our main challenges on this project were learning Unity, integrating our designs and dialog into a logical structure, hooking the game up to Twitch and getting all of the right input from there. We had a lot of moving parts on this project, from the Twitch chat integration to all of the various cutscenes animated in After Effects and spritesheets created in Illustrator that needed to be integrated into various Unity GameObjects.
Specifically, we had to learn how to integrate Twitch Chat and integrate their OAuth functionality. Finally, we had hundreds of lines of dialog that we converted over into JSON and then parsed so that our Game Logic could properly use it and figure out when to say what line of dialog. We also a multitude of different scenes with custom drawn designs and timers to create a great story telling experience for our crowds using timers and Unity CoRoutines.
We found that integrating with the Twitch API was generally easy, but using Unity Paradigms and mental models was quite hard. The idea of scripts being attached to Objects was very antithetical to OOP paradigms that were taught to us at school, so it was a bit of a challenge to adapt and learn. There are various intricacies to Unity that were hard to understand and we wished we had done more learning earlier to get past these roadblocks when developing Wolf Goes To School.
What Worked
Before this project, five of our six team members had never made a video game. So the odds were certainly not stacked in our favor, despite attacking the assignment with fervor. To make this game work, we had a lot of moving parts which needed to all come together in just six weeks. From writing effective, meaningful dialogue, to focusing on interaction design, creating thoughtful characters and environments, all while learning Unity and Twitch integration strategies, every member of the team was needed to make this project a success. In the end, despite scaling the project down as necessary, we ended up with both a playable video game, and a much stronger foundation surrounding the principles of video game design.
What Didn t Work
Grand ideas must always be molded into viable realities, especially when you re on the clock. In this vein, one major element of the game we weren t quite able to complete in time was the streamer s integration into the game. Originally, we intended for the streamer to play the wolf, thereby having to act on camera for the audience. In lacking this element of the game, we ve produced more of a Twitch Plays game, where the wolf is an NPC. One other hardship was the dialogue development. Although we successfully navigated problems of having way too many pieces of dialogue to write, playtesters still found that the dialogue was too random. Good for a playable game, but could be iterated upon in the future.
Future Improvements
Building a playable game was in itself a gratifying achievement for team Wolf. But that s not to say there isn t lots of room for improvement! As previously mentioned, our first step for developers would be to create the streamer side of the game, to bring theater and streamer-audience interactions to the spotlight. Also, we d like to integrate either a chatbot command that tells players what a character has said in past rounds, or ideally, create a screen that displays (at appropriate times) what characters have said in the past. In terms of dialogue, we d like to tighten up the overall body of text to give hints to audience members as to who their target is. Although the dialogue as is accomplishes our goals for the project, it has the potential to be a masterful puzzle of hints, red herrings, and comedy. Finally, our designers would first love to tighten up some animations, as we had to make some sacrifices to the animation of our anthropomorphized silhouettes. Additionally, we d like to create a couple of different screens to display information that currently clutters the Twitch chat somewhat, as well as to improve the overall connection the audience has to the game. For now, however, we present Wolf Goes to School as the first videogame five of our six team members have ever made. It was a challenging and rewarding experience, and we re proud of our final product.
The cast of Wolf Goes to School.
Thanks for playing! - Team Wolf Goes to School