For the second half of the semester, we develop two demos for the CAVE in ETC. Eric and I worked on one of them called Puzzle Mania. Puzzle Mania is inspired from an old boardgame, Ubongo. It is a two players cooperation game in the 2D CAVE.


Based on the former research, we make a multiplayer game because we believe the shared space in CAVE is unique and can actually enhance the interaction between players. Also, we need to make a 2D game. These are the constraints from the CAVE. Taking our team into consideration, we just have 2 people and 6 weeks. Eric is a programmer, and I need to take care design, art, sound, and part of programming. It will be heavy for us. So we decided to go to a minimized and abstract art style. At last, since the game will be used for people whoever comes to ETC for a tour, our target audience will be from kids to adults and to entertain both players and audience.

To sum up, we have the following constraints:

  1. Interactions between multiplayer
  2. 2D game
  3. Minimized art
  4. Fun to play and fun to watch

One Concept and Two Ideas

The puzzle concept is proposed in the first half. The puzzle concept matches the minimized 2D art style and is easy to be a two players game. The core part of the concept is the pattern – map mechanics, which means using patterns to fill a map.

Core Gameplay
Core Gameplay

Based on this core concept, we propose two potential ideas for the actual game: a competition game or a cooperation game. In competition one, the patterns are public, but players have their own maps to finish. So the main conflicts, or say, interactions are mainly focused on the moments of grabbing patterns. Players fight for finish their own maps quicker than the opponent. In the cooperation one, the map is public, but the patterns are private and set on the side walls. So players work together to finish the map by combining patterns they have.

Competition Gameplay
Competition Gameplay
Cooperation Gameplay
Cooperation Gameplay

Tech Prototype

In the beginning, we start with the target to make a workable prototype and test the basic of these two gameplay. First, we define potential behaviors when playing:

Basic Behaviors
Basic Behaviors

First Playtest

The first playtest have two main focuses: test the control, and make a decision on competition/cooperation gameplay. We do both survey and observation for the playtest. Survey generally asks about the ratings and preference for different gameplay. Observations help understand players behavior more.

Statistics from survey:

Gameplay Preference and Average Rating
Gameplay Preference and Average Rating

Points from observation:

  1. (Cooperation) Players discuss and instruct companion sometimes.
  2. (Competition) Players seldom talk.
  3. Players use flip & rotate to solve the puzzle.
  4. (Competition) Teenagers are nervous to lose in the competition.
  5. Friends are more easily to be entertained by these prototypes.
  6. (Cooperation) Audience propose some solutions sometimes.

The Decision and Mechanics Design

Players’ interactions and audience’s engagements are the important things when making this game. So we decide to go to the cooperation one because it appears more potentials on facilitate participations from both players and audience. Then we propose the whole system for the game based on the patterns – map mechanic.

Second Playtest

Soon we have our second playtest to see if the prototype is easy to understand and fun when we finish the gameplay. We find it varies to different players. When players are good at solving this kind puzzles, they are really engaging and feel comfortable to communicate with the companion. However, when players are not good at this, they always try different patterns by themselves instead of talking with the companion. The reason is that the process of solving a puzzle is a mental computation on which the communication is based. In this situation, players will feel frustrated. Another reason for the lack of communication is that patterns can be used by the player always appear on the side wall which close to the player. So the system actually teaches players to focus on their own patterns instead of noticing the companion. It conflicts with our initial idea to facilitate communication.

So we decide to add a new feature that patterns belong to different players will appear on both side walls. Players can only pick up the patterns of their own colors. So an interesting interaction will be calling your companion to pick up the right pattern. On the other hand, we also decrease the number of patterns appear at the same time. Providing fewer options make players easier to stay on the same page and start to discuss the solution. As a trade-off, we change the original randomized-generated pattern system into manually set patterns-map puzzles because we need to make sure with a limited number of patterns the map is still solvable.

Flow chart

Flowchart for each level
Flowchart for each level

Art Style

With the limitation of our skill set, we choose to make abstract art style. We use simple textured blocks and related cursor icon. The backgrounds are all low polygons. Since we are all programmers. We use a lot of computer-generated art, such as particle system, and different kind of script-controlled effects. (Breath, Blink, Shining, Dust, Firework, Highlight, etc.)

Art Samples
Art Samples

Interactive FXs

We use a lot of visual and audio effects to deliver the game information with players, including DMX lighting. Here is the index for all effects:

Interaction Table
Interaction Table

Third Playtest (Soft Opening)

In the soft opening, puzzle mania is presented to all the faculty in ETC. At this moment, we receive comments, critics, and product score.  Before soft opening, we make a simple tutorial and integrate it with the title screen. However, people still have troubles on understanding how to control. So for the last week, we pay all our attention on making a new better tutorial.


Tutorial Flowchart
Tutorial Flowchart

For the tutorial, we design the flow to make sure teach basic rules and control on the surface. We repeat control practice twice to make sure players get used to it. Moreover, we teach players the basic mental strategy of solving the puzzle during the tutorial which is critical to understand and play the game.

What went right?

  1. Paper-based prototype. It helps a lot on both conceptualizing the rough idea and adjust specific levels as well.
  2. Realize team constraints. Without a professional artist, we choose to make a abstract game and using a lot of script-based art effects.
  3. Well-designed tutorial. When designing and iterating on the tutorial. It is interesting a simple change of the process or the sequence of the contents in the tutorials impact players’ behaviors. So in our final tutorial, it not only teaches them the basic control, but also it teaches the mental process and basic strategy they should use in the game.

What went wrong?

  1. Wear too many hats. It is so messy when I need to create arts and find sound effects while the playtest reveals bugs to be fixed and mechanics to be delivered better.
  2. Environment and genre do not match. CAVE is a very open space. Puzzles gives players too much pressure especially when there are audiences.

Future Development

One of the biggest problem for the current version is lack of variations and surprises. Generally, in the tour mode, only the difficulty is increased between levels. In the infinite modes, variations are based on randomized contents. So to make the whole game more interesting, the first thing we should work on is to add some new mechanics. Here is the list of some new mechanics:

  1. Flashlight effects based on cursors.
  2. Cooperative picking up and manipulating.
  3. Obstacles on the map and patterns.
  4. More complex scoring system which requires better management.
  5. Customized patterns.


Qiaochu (Mac) Li, Eric Tsai.


Leave a Reply

Your email address will not be published. Required fields are marked *