Concept and Research

The first project in Design 6400, an open design project based graduate studio class, was a group project with Yuxin Yan and Raven Serenity Glover. As a group we decided we were interested in keeping an open mind as our project could be anything from animation, an interactive experience, or game development. Yuxin, Raven, and I wrote out our interests and goals on a whiteboard during class. I pitched several of my existing animation ideas and areas of interest. During this phase we all gravitated towards a game with a gameplay mechanic that had something to do with half and half as that had the most overlap to all our thesis research.

Buster Bros is an older game where you shoot spheres that split in half and get progressively smaller.

I then found a game on steam called SPLIT. There was no user commentary or playthroughs that I could find so I started playing on my own and made a playthrough video of myself playing the first 19 levels. The game has a unique mechanic that splits the character and environment. There is also glass that can be transparent or solid depending on the split state.

After we shared our ideas and findings as a group we began narrowing down our plan for the concept presentation the following day. We added to the MIRO board and then tried to come up with ideas for our favorite concepts. Our favorite idea was a game where our past self would do tasks to help our future self.

We also constructed a rough visual style board. Hollow Knight was our favorite reference.

Prototyping and Level Design

My main task for this project was level design while Raven took on the Unity scripting and Yuxin took on the lead for art. I started with paper prototype before moving onto white boxing. The main goal for this stage was to progress through quickly enough for ample play testing for our next phase.

I created an initial paper prototype and we filmed the intended gameplay for a time based puzzle solve. The idea was that a magic past self would come out and perform an action then that action would loop until the player recorded a new action. As a group we talked about level design and how to introduce the player to the gameplay. The core mechanic is a stop watch that allows the looping behavior. This requires the player to have the ability to start, stop, and reset the stopwatch as they time themselves through actions. I also went home to play a bit of Braid to review how they used the time mechanic. World 5 seemed to have the most similar mechanic but the way the player was introduced was extremely minimal. We decided this puzzle would serve as our tutorial level in order to introduce the players to the time mechanic.

We met again to refine our story and game mechanic. The basic idea we came up with was that our character has a tree that they love and they are gathering items to take care of the tree. And then the levels are based on seasons. So in the beginning when it is spring the character is searching for a watering can to take care of their tree. Summer would be fertilizer. Fall would be a rake reward item. And then in winter their tree suddenly dies. A grief level involving a wind mechanic to push back. Raven is still working through our basic mechanics so right now the narrative ideas are loose as we are not ready to implement art assets or level design just yet.

Our feedback for the narrative was that we should connect everything more to the half and half initial idea. So Yuxin and I transitioned the narrative to a mage/wizard character that has the ability to magically control time. Raven started working through the 3D prototype and Yuxin and I found reference images and design assets and elements. As the story evolved we decided to set the first level in a cave. Here are some of the references images we collected:

As I started designing the level I ran into limitations with the character movement based on our 2D platform style. I talked with Yuxin about designing floating platforms that we could add to allow the character to safely reset up top if they fell down and became stuck. I also discussed with Raven if our character would be able to walk in front of objects or not and different workarounds to solve that. For our first level I created a few drawings to plan out three puzzles that can be solved using the time mechanic. The first puzzle is the same as the tutorial. The second intended puzzle is a lever that releases falling rocks that make a spiked floor. This is a reach goal for us but Raven was excited to explore ways to have the rocks fall down as that relates to their research tangentially.  The third puzzle is a double button but your character has to be on the one closer to exit.

I then built a whitebox in Unity at Raven’s request in order to provide a simple testing path for our game design layout. The green is the push puzzle. The blue is the lever and button puzzles. The light blue is the objects impacted by the blue switches.

Art, Animation, and Layout Design

Yuxin’s art style was very cute and playful so our game design and narrative formed organically with her artwork. I jumped in to help Yuxin with a few assets. I started out with a magic staff that could match Yuxin’s style and tie the character to our magical environment.

While working through the art assets I created a production schedule and asset list to help Yuxin and I track our status and distribute tasks evenly.

With a few assets created I put together a style sheet and narrative guide for our game to collect our references, assets, and story into one document.

The next stage of the project was the graphics layout into our test scene. There are clearly gaps and assets we still need to make but I was able to recycle some assets, at least temporarily, to give an idea of things. I also extended the height of Yuxin’s BG in order to fill the top and bottom of the screen.

After the first half of the layout I went back to draw some stairs and train tracks. Then I went in to fill out the rest of the level with assets. There are still a number of assets that needed cleaned up but for now the scene is starting to look cohesive.

I then created new assets to fill in some of the layout gaps and asset in player feedback. The stairs would allow players to naturally reset if they fell and get a better handle on the jump timing and skills. The railroad tracks were important to help indicate to players that the mine card object could be pushed and pulled. The levers were an essential game puzzle element that needed included. I had to setup the layers in a way that the asset could be animated.

I learned the very basics of the rigging system in Unity in order to animate the lever by following this tutorial

And here is the lever animating. It needs to be setup so that it only moves when E is pressed on the blue trigger behind it but for now I left it animating so my team mates could easily see the update. After this point we started using the Plastic SCM system to update our project with a check in and check out system and leave notes as we progressed through our development.

Yuxin had setup our 2D character rig. I was able to animate the lever push, lever pull, use magic, and death animation fairly quickly after Yuxin showed me the Unity animation setup. I then wrote out the narrative script so we can add in the text dialogue boxes to help explain the game. My next task was to save out all the dialogue boxes so they can be implemented into our game. Below is a clip of the animations I added.

Play Testing, UX, & UI

I took on a significant portion of the play testing. Because of our recorder system for the past present mechanic I knew we would likely have some bugs. I discovered one bug while hitting reset while actively holding something to push/pull would allow you to fly through the whole game level.

I also did a rough layout for text overlay for the tutorial and instructions level and then a graphics layout for the tutorial stage.

When our game was in a playable state I gathered a number of play testers with a wide range of gaming ability to test our gameplay.

Kyoung:

After completing the playtest she suggested that we have a more clear reset than just hitting the key. I suggested that we use the ‘Use Magic’ animation to address the reset issue. If you hit R and it starts and animation that teleports to the top then that might be more clear of the consequence of the action.  She also said a ladder climb would be helpful for players that fall down. We saw during Kyoung’s playthrough that our timing was too fast for the recording limit. So we adjusted from 10 seconds to 12 seconds.

Juan:

Immediately observed that the ‘Press space to continue’ text was far too small. Juan approached the tutorial puzzle in a way I didn’t anticipate. He used two pushes to get the cart to one side before pulling it to the end. It worked and was a valid play style. Juan also said seeing feedback on the R reset would be good. Juan also pointed out the column art asset was inconsistent from the rest and suggested extending the asset down. Juan also appreciated the art style and that we had animation incorporated.

Mila:

Mila also found holding the R button so long frustrating. Mila said checkpoints were great and she appreciated them. The swinging door that held up the rocks kept blocking her jump path on puzzle 2. She requested that the executable be a kitty Icon. Loved the art style.

Emma:

One of the fastest of the tutorial puzzle solves and figuring out the time loop mechanic. She also figured out the levers quickly. She pointed out the rocks giggling for time loop was happening.  Emma said she would have preferred each ability be introduced one at a time in the tutorial. Loved the art style.

Noor:

Suggested that bulleted instructions at the beginning would be easier to remember. Noor loved the art and said it was very cute and suggested we remove the smiley face at the end as it was not in line with the rest of the game.

Sebastian:

Also approached the tutorial puzzle with the same two step record method that Juan used. Sebastian also suggested the pillar be extended. He said overall the puzzles were similar to ones he had encountered and were solid for getting a good bit of game out of the mechanics.

I compiled all the feedback notes I was able to capture and then created an additional to do list in our schedule.

Playtest Feedback To do List:

After our first play test I placed our reset locations into our scene and then began working on adding in our cutscene at the beginning of our game.

Because Yuxin’s lovely cutscene drawings were the size of the window and I needed to add our backstory text I had to do some graphic design to get a layout to show all the information.

I couldn’t quite figure out how to exit without a button rather than on finishing the loop but Raven was able to help me get the destroy object parts of the script integrated.

Always trying to be organized I updated our to do list with new player feedback and changes we wanted to add to our game before our deadline.

For our last push we made sure everything was clean and clear for our final delivery. I did a thorough playtest for a few hours while streaming for Raven and Yuxin. We found several unexpected errors like getting pushed through the elevator if you get caught on the edge, the timer effecting the character and objects differently if the player hit escape, and then a minor bug where the character could be stuck in the falling position during animation loop. We were able to fix all the major bugs we uncovered. A few minor ones exist in the game but they don’t break the gameplay.

For the elevator fall bug we made a kill box so if you fall through the floor it is as if you are crushed by the elevator and sends you to reset before you fall through the world.

The Esc bug had two issues that were fixed over the last few days. One you would get locked in place and another where the timers would offset, which is discussed in the video above. The video below shows the movement lock.

We made a number of minor tweaks to graphics and other things to add final refinements then tested over and over to see if we could break the game. I created a final build and we uploaded our remaining elements to the Y Drive.