Hey everyone, welcome to another Stonehearth Desktop Tuesday! Sometimes, I see you guys asking: what goes into making a feature in the game? Why do features take so long? How do you know if a feature will be good? For this week and the next few that follow, I’d like to do a deep dive on this process, taking you through each stage of feature development, touching design, engineering, art, and QA. Since A22 is about conversations, and that feature is currently in progress, let’s use that as our starting point!
Conversations – Design Phase
Recap from the video:
- Stonehearth has had hearthling conversations for a couple of alphas already, but right now they’re idles that mildly contribute to your sense of the hearthlings as people, and that also serve as a tech demo showing two entities with AI interacting outside the combat system.
- In Alpha 22, our first step was to define the goals of conversations: how will they make the game better? We decided that conversations should further accomplish our gameplay mission of inviting you to engage with hearthlings as individuals. This means we want the conversation system to perform the following three functions:
- Commemorate things that have happened in the game. If your hearthlings have recently encountered goblins, the game should change to reflect this, for example, by having your hearthlings talk about it.
- Help you think of your hearthlings as individuals, for example, by giving them opinions about stuff that has happened to them; liking the cornbread they’ve just eaten, or not liking the goblins they just encountered. Over time, hearthling experiences would cause them to accrue different likes and dislikes from each other.
- Create new systems that can be tied to the larger game: for example, by integrating conversations into the traits and thoughts systems.
- So where do we start? Once upon a time, a single engineer, probably Tom, would have started writing code and animations at the same time, and integrated them. Though efficient, it means that a system like this would have had just one animation, and would probably not integrate with any other systems in the game. Now that we have specialists in design, art and engineering, each person can make the system far richer than any could alone, but we have to do a lot of coordination to make this happen.
- Designer Richard starts the process with a design doc that dives deep into each system that supports the three main objectives of the feature.
- For example, to make sure that conversations celebrate stuff that’s happened in town, here he describes in writing, that each hearthling will have a subject matter table–that hearthlings know of a set of possible subjects. Subjects are added to the table as hearthlings encounter them. Hearthlings speak about subjects from their permanent set and recent set.
- To make sure that hearthlings develop differences and that they feel like individuals, he specifies that subjects are also associated with opinions, so hearthlings don’t just know about things, they have emotions about them, which they then express via animations, and can spread to other hearthlings they talk to.
- The design doc also defines conversation structure, so conversations look natural. Our old conversations were statically pre-authored: there’d be a conversation in which a hearthling talked about their job, and another hearthling applauded. New ones could be created in json by stringing together specific animations. However, if hearthlings can have an infinite number of conversation topics, the conversations should fit that, and not be immediately recognizable, so the new conversations dynamic: they’re assembled out of nodes: a greeting, subject, reaction, conclusion. Each section has a number of possible animations associated with it, and they’re combined in realtime, so you can see any number of conversations about any number of subjects.
- Other design docs written by Richard define how conversations affect the game: for example, we have traits associated w/ conversations, like gregarious/loner. Conversations can also affect your mood: if hearthlings disagree on a topic, they may get a sad thought that they had a fight with a friend. Or if they agreed with another heathling, they may have a happy thought that they had a good conversation with a friend.
- The design doc ends with suggestions about how to depict subjects, mood, animations
- At the end of our last alpha, while we were fixing bugs on traits, Richard sent the design doc out to the team. Everyone gave him feedback, he gave a presentation about his improvements, and then passed the doc to Linda, Allie and Malley, for implementation.
- By this point in the process we’ve gotten quite far, but aren’t quite done yet–we have animations and mutable conversations, but no subjects yet.
- Next week, lets see how Engineer Linda took Richard’s doc and figured out how to turn into into systems and services in the game.
We hope this has given you some insight into what it’s like to be a game designer on a small team! Let us know what you think, and if there’s any other questions you have about conversations, game design in general or about design in Stonehearth in particular.
Other Announcements
This week’s stream is cancelled due to scheduling issues. We return next week on 6/8 with a stream from Chris Klochek about his progress with the cantankerous building service!
