Creating Immersive NPC’s in Roblox

Given a design brief, my team and I were tasked with creating a feature in a game engine that would allow for the creation of more immersive NPC’s. In a month’s time, after careful research, we created a Character-Centered Narrative Planner in Roblox.

Team Members: Danielle Dixon, Andres Brandt-Becker, Iyla Laugel, Carina Wang

The Problem
If you’re a gamer, you can probably remember a few characters that stood out in your experiences. They also probably have very different things about them that made you remember them.

Immersion is quite relative, and not all game engines are made equal.  Some people may define immersion as having a great story, while others may say it’s due to high-fidelity graphics. Regardless, we know that there’s something in the formula for creating games that leaves a lasting impression in people’s minds.

Small game development studios may particularly struggle with creating immersive games due to a lesser number of employees or funds. As such, it's very common for game designers to fulfill multiple roles that would normally warrant entire departments of people. With the variety of different softwares and methodologies that go into creating the components of a game, collaboration can quickly turn sour.

For those that wish to create games with great stories in Roblox, how can we promote skillful, agile planning?
The Solution
We created a Character-Centered Narrative Planner tool in Roblox where the character rigs that were already in your game world have planned details and story parts attached to them (seen only for developers, not in-game).

The tool allows creators to keep track of the name of the character, it’s narrative role in the game, any attributes (depending on the genre of game), its progression throughout the game, and a brief description of the story surrounding it.

Each character node has the ability to have a ‘Storyboard’ attached to it that allows for a longer description as well the ability for other developers to attach images of reference, comments, and links to external sites.  

In Summary...

WHAT IT IS
- A preview and reference of the big picture
- An agile non-linear story branch testing tool

WHAT IT ISN’T
-A live coding/programming tool
-A project management tool

Why this approach?
Do you remember the last time you played any kind of game? You may remember playing for five minutes when, next thing you know, you looked at a clock and five hours have passed!

What you entered was called a ‘flow state’, and it allows you to hanker down and grind things out. We discovered that Roblox was lacking a dedicated planning tool for developers that would let them seamlessly transition between action items. This often led to disjointed experiences amongst a team that would delay processes and at worst resulted in a poor game play experience.

Game design is a highly agile process, and things such as timelines, software, audio design are always subject to change. Knowing this, it can be difficult for those in charge of narration, and, say, programming, sound design, and character rigging to keep track of the happenings in-game. Roles in a small game development studio involve a lot of cross-collaboration, arguably more so than in large studios. This is due in part to one person oftentimes needing to wear many hats in the game making process to save funds.  

Our tool provides a central, digital location for those in any role to collaborate as the story develops. This will decrease the chance of a break in flow state by eliminating areas of confusion and preventing the need for information to be stored across different software.

After all, games aren’t created in a vacuum!

Research Highlights

After playing through some experiences with Roblox, we wondered about its popularity despite the lower fidelity graphics and simple gameplay. To get to the bottom of this, we spoke with level designers, narrative designers, concept artists and psychologists to discover what the ‘nugget’ was that tied everything together.

This is how we discovered the Zone of Proximal Development (ZPD), or the ‘flow’ state, hiding right under our noses. Players achieve this state when they lose track of time and become highly focused. For things in game that we can control, this flow can be broken by things such as placing terrifying dragons in a peaceful dress-up game or even putting Japanese motifs without warning in a narrative about vikings.

Non-Playable Characters or NPCs can have large effects on how players receive the information that is thrown at them in a gaming experience. NPC’s have the ability to engage players throughout the story, and can be points of relief even if the player doesn’t care for the overarching story. Where Roblox’s platform excels is in the low barrier to entry for both playing and creating as well as the large sense of community it creates.

Many of the top games on the platform across genres involve large player-to-player and/or player-versus-player interactions. Younger players enjoy the sense of repetition in these shorter and easier to play experiences while adults may come to appreciate things such as  narrative complexity or engaging characters

Our task then became to create a tool for developers that would marry these experiences for mass player bases for studios that want games with a great story.

A Persona of Steve Wilson, a narrative designer and animator at a small game dev studio. He has difficulty in balancing roles on his team and has frustrations surrounding collaboration
Considerations
We interviewed 5 people, asked 2 people questions asynchronously, and received 4 survey results. Here’s some powerful considerations we gathered:
Different players have different views on NPCs. For example, some people don’t wish to engage in dialogue but still need to be immersed. In Everquest, NPCs highlight immersive experiences without dialogue. Alternatively, some games highlight the balance between what the narrative wants the player to do vs. what the player wants to do, like in Baldur’s Gate 4. Getting NPCs to behave in realistic, unpredictable ways to make the game’s challenges dynamic is what allows them to shine.
Will
Indie Narrative Designer and Academic Program Director
Concept art is one of the first things to happen in the development process as the story is very liable to change. One of the biggest challenges with character creation and dialogue is in globalization. For example, UI has to take into account how much screen space is taken up in different languages. Issues can also arise with certain symbols and translations if they are offensive or don’t directly translate into another language. So how do we know how players feel about this? For indie games, lots of playtesting is required to determine if players find a character intriguing. Typically you want to lean into archetypes based on what appeals to your audience.
Gabby
Concept Artist, 3D Modeler and Game Design Professor
When designing for a platform, there are essentially two users in mind: the end-user and the game developer. Your job is to meet both people’s needs an deal with conflict. Remedying these issues involves lots of clean up and numerous iterations of features of the product/arrangement to keep things fresh and exciting. You also want to consider if the problems that arise come from research. Depending on how much you want to business-ify the matter at hand, the solution may depend on specific metrics or goals of the business. At the end of the day, timing is everything. Figuring out when to offer help and how to interpret when your user groups need help is crucial.
Matt
Lead Product Designer

Usability Testing

Our biggest brainstorming sessions revolved around the question, “How much tooling is required to record a story?”

There was so much that went into creating a game that it seemed like just providing a digital pen and paper may not suffice. And, sure enough, through unmoderated remote usability testing, we found that narrative designers felt very isolated in the creation process. It was difficult to convey creative ideas for non-text based games and ensure that ideas were correctly communicated to artists. We also found that they weren’t given enough text related tooling to show connection or show emphasis.

This is where we had our lightbulb moment that led to some of our final iterations. We included ideas from the text-based platform Twine and the collaboration tool Miro to allow more flexibility as the game evolves. There was much more success with these additions as artists now had the ability to attach images of concept art to text nodes and programmers could leave comments and link to code snippets.

Once getting the main ideas down, we conducted A/B testing to remedy some layout concerns with ensuring an appropriate amount of characters and information was displayed without overwhelming team members. This allowed us to add categories to character nodes to help ‘break up’ text and determine a more creative way of displaying the full cast.

We would like to continue reviewing the visual design of the project to make more careful color decisions and help certain elements further stand out.