Hello everyone, 

My game is the result of a combination of techniques that I've always wanted to try. The end result is kinda lackluster, but I've used the jam mostly as a learning experience.

First things first, the map is completely generated through PCG using Perlin Noise and the object and NPC placement is also based on that.

Secondly, the NPCs have 2 different kinds of Decision Making Layers: 

1) The Top Layer uses an agent modelling tool "FAtiMA-Toolkit" to define its Macro Actions: Speak, Pray, Wander, Fight..

2)The Bottom layer uses Behaviour Trees to complete the actions in a Micro level: For example, in order to Speak with the Player, the NPC must first reach them. 

Finally, the main objective of this game was to try to see how an open-ended-input game could work. 

The Player can "say" anything to the NPCs and would react accordingly.

My idea was to include both the agent modelling tool and GPT-3 to see if it could work .

Unfortunately I did not have time to complete this part and, honestly, I would've run out of budget from my GPT-3 playground account. 

At the moment the player writes a sentence as an answer to the NPCs but on the back-end I use a Sentence Similarity function to map it into an already available dialogue.

That's it, I didn't have much time to work on this unfortunately, but I loved the challenge and next time I'll be "readier"

Credits:

- Me: worked on this project for ~2 days

The art is all made by the talented Cainos: https://cainos.itch.io/


Download

Download
WebGL.zip 11 MB

Leave a comment

Log in with itch.io to leave a comment.