Game Design Document
Betrayed Trust
• Management: Jakov Kadić, Sandra Sačarić
• Design: Jana Jambrešić, Luka Pošta, Dora Garafolić, Dorijan Kos
• Writing: Sonja Kolarić, Ivan Simić
• Backend and AI: Frano Šimić, Karlo Rosenthal, Mateo Zović, Karmelo Mrvica
We are developing a fantasy novel type game with AI elements in its core. AI will be used not as a gimmick but instead as a clever platform for our dialog and gameplay to reach hights not yet seen in a student project. The AI will be reactive and we'll get to decide to become our enemy or stay as our friend depending on the treatment they receive from the player.
• Genre: Fantasy, novel, strategy
• Target audience:
– Age: 15-30
– Gender: Male/Female
• Monetization: No monetization, student project
• Game engine: Monogatari
• Platforms & system requirements:
– Web game
The idea is to introduce the player into a fantasy setting, one where they are a part of the guild. They are to determine who they can trust and who they should stick to, as one of the main characters is bound to betray them.
The project introduces an innovative visual novel that leverages advanced cognitive agent technologies to offer players a unique experience of an interactive fantasy adventure. The main unique functionality is presenting the player with opportunities to talk to a cognitive agent in real time - cognitive agent being a trained bot with given visuals and voice samples. The difference between a regular visual novel and this one would lie within the way that the user is supposed to interact with the character; usually, the person is given fixed, hard-coded dialog options and choices, whereas in this case the player can decide for themselves what to ask, and in which manner to pose the question.
The project will deliver a game with several key features, including:
- Dynamic dialogues with cognitive agents that respond based on player interaction.
- A rich narrative with multiple possible endings, depending on the player's choices.
- A complete visual design that includes characters, backgrounds, and interface.
- Integration of multimedia elements such as video animations and voice content for higher level of immersion.
The main game mechanics include a branching story system that allows for different paths and endings, and a dialogue system with cognitive agents that allows players to freely ask questions and receive tailored responses.
Fantasy, guilds, and monsters are things that are at the forefront of our game. That setting will only serve as a foundation for our intriguing characters that you will meet throughout your journey.
By accepting the mission, the player gets to meet two teammates: the mysterious assassin, Bizarr and the fearless warrior, Rioth. As the trio embarks on their journey, you discover hints of treachery and betrayal. The player must then avoid potentially life ending paths to unmask the traitor among them.
Rioth and Bizarr, the two main characters that were given visuals via AI generation. The team would have hand-drawn them and brought their designs to a higher level, however due to time limitations that was not possible.
On the left you can see Bizarr, while on the right we have a preview of Rioth. Rioth wasn't initially thought out to look that way, in fact she was supposed to hold a much more fierce and dragon-like design, but due to limitations of our technology we couldn't achieve what we had in mind.
Our user interface had used a resource pack but it was a challenge implementing it and bringing it to a level of perfection we sought after. Within our design team, one of our colleague almost single-handedly fixed all the little nick-nacks regarding our UI and, aside from the base user interface, she did everything regarding the main menus, dialog boxes, and so on.
We've made animations that support the characters you previously saw. A preview of this feature is available on our trailer, where you can also see our valuable colleague (who had made the video all by himself) showcasing the speech recognition mechanic: Link!
In this section of the GDD we will go over our project budget, a timeline graph meant to visualize multiple phases, their respective deadlines, as well as the tools needed to complete each feature and reach our goals.
The project has no financial budget as it is a student project but utilizes various tools for development and content customization.
Risk Management
The main risks included technical challenges with the implementation of cognitive agents and time constraints for completing all phases. The team had discussed strategies to mitigate these risks, including plans for alternative implementation methods where necessary. For instance, we struggled to implement the cognitive agents dynamically via javascript, so we ended up doing it using HTML.
Research and planning phase
- Sketching UI, characters, story, background and setting up backend
- Planned end date: 31/03/2024
Design
- AI training, UI close to final version, story finished, video and sound content finished
- Planned end date: 04/05/2024
Development and testing phase
- Test compatibility of all existing elements and connect them in a sensible manner
- Planned end date: 17/06/2024
Finalization
- Making sure it works like intended and finalizing documentation
- Planned end date: 20/06/2024
Game engine: Monogatari
Art: Glibatree (chatgpt), Getimg.ai
Audio and video: 11labs (audio), wondershare filmora (video), heygen (video), convertio (audio and video), vmake (video), canva (video)
Writing: arcweave
Skolaztika, Link to assets pack
Project plan: Projektni plan
Visual novel - a type of interactive fiction game that primarily focuses on narrative through text-based storytelling
Game engine - a software framework or platform used for the development and creation of video games
Cognitive agent - an artificial intelligence (AI) entity designed to simulate human-like cognitive processes