Week 5: Feedback and Decisions

Last week was our Quarters Open House and we had few people from Electronic Arts come in and play with our prototype. We also finalized on the various platforms and tools we will be using for our project from now on. Another major decision was finalizing our character design and voice that would go with it. This post is going to cover the process that was involved in making these decisions and also why we chose to use and do these things.

Playtest feedback and stats

We received a lot of reactions and feedback that we will use to design our experience. Here is a summary of what most people said they liked and did not like:

And some observations and suggestions to go with the feedback:

We also collected samples of responses from the people who playtested our project and hope that this would help us build a better Language Understanding Model (LUM) for our character’s speech.

Character Design

We started developing different designs that our character could be like. Our client was very interested in knowing the kinds of design we would come up with. Based on our decisions from the past few weeks, we went with a humanoid-animal character. Here are our top picks:

After a couple of discussions and tech considerations, we decided we would go ahead with the humanoid Hippo as our AR character. The nature of the Hippo – huge and slow would set up expectations from the player who would want it to act similarly.

Voice Options

Just like the character design, we wanted to iterate on some voices and sounds too. We use the Amazon Polly service and the voices that come with it. Here are our samples that we generated with the above characters in mind:

After a couple of suggestions from our client and faculty, we will also be iterating on the voice of the Hippo to make it sound more pleasing and natural to fit our character.

Project Architecture

Since Quarters, we have finalized some crucial parts of our project. Here is a Summary of our tools and platforms:

To have a more clearer idea of planning for the next few weeks, we decided to come up with a project architecture diagram which would modularize our workflow.

App States Machine: Consists of modules that constitute the different stages of the game the agent would be facing.

Agent: Different components of the agent that allows it to converse and interact with the game and the user.

AR Objects Manager: The AR component of the application that includes the tracking of physical game pieces.

Also notice that few sections have been marked with a ‘ * ‘. These are the core modules that are required to be working for us to lay down the basic pipeline. Therefore, these are the tasks that we currently have for Halves at Week 8.

Plan for Halves – Week 8

Since we received quite a lot of feedback from our playtesting both on features and functionality, we decided to split our tasks into two different categories – Priority and nice to have. For halves, our team will focus on completing the prioritized tasks done first before moving on to add other features that would make the experience more polished and exciting.