Dialog systems for unreleased game (2D)
(Stopped in production)
My role: System & feature design, UX
Credits (art): Vittorio Truffarelli, Nils Camin, Christian Nauck, Marcus Mestrov, Calvin Verhoolen, Tamas Molnar
Summary
The challenge was to create a simple, low-resource narrative system for game-to-player communication that would be easily handled by the writers.
Note: The game was stopped before we were able to polish further. So, some things may seem little rough. I'll sum this at the end of the post.
Description
The game consisted of detailed, high quality 3D animals in 3D space. While it was possible to narrate through scripting dialogs in 3D area, it raised a lot of potential issues and didn't fit all the requirements we had. The decision was made early on to add a 2D dialog system as a narrative tool of communication focused at the player.
Separating the autonomic animal behaviour in 3D from player focused narrative felt like a good way to go.
Though as result the feature seems quite simple, there was a handful of challenges solved to make it seem so.
For design it was coming up with a system that would both fit existing story scripting tool AND be easily used. Then, with UI/UX, we needed to come up with rules for the actions flow (such as transitions between parts of the story) that would apply at any given moment
The dialogs were controlled from Excel, and had a bunch of story needs: display characters, hide them, change expression, change position (L/R), choose speech or thought, and similar. The system we used so far only enabled pair commands for pair actions (aka if you don't tell character to go away, they will stay on the screen, if you don't end the camera lock, camera will stay locked), so we needed to come up with a set of fallbacks.
Here is a visualisation of the different layers: set up (what is in Excel config), screen (what player sees) and background (what orders actually happen).
Since the project got stopped before we were able to polish anything, I want to shortly sum up.
One of the big discussions was whether to call the cinematic mode (camera mask as seen above, also indicating temporary lack of control over the game) automatically or not. An option was to call the mask with the character or even the story text piece itself -- in the end we went for consistency, and opted in for calling the mode at beginning of every story session (that could be in 2D and 3D alike), to make it particularly clear in the Excel file what is a story sequence and what is a random dialog. (Leaving aside that at the moment in production it wasn't possible to look for an Excel-free setup. Also, everybody loves Excel.)
The thing I'm quite happy with is how we dealt with the order of operations for any potential parts of the narrative. Setting up a flow that gets handled automatically (update character, then update expressions, then update speech type, then update text) meant that any story piece would have same pace and no additional oversight when writing it would be required.
What was becoming clear already was that the presence of the background had its issues. Our vision focused on animals living their lives whether player observes or not, and was causing disconnections. Case one being someone talking to player about an animal being angry, while this particular animal jumps happily in the background; another more generic case then would be animals doing their thing in the background takes away focus from the main narrative. Both these would require a separate set of rules of what can happen while a 2D dialog is happening that would possibly interfere with the vision, and most likely we would look at what kind of background or overlay we can use that fits the narrative needs.
