top of page


k3G0CV (1).jpg

3D dungeon crawler with interactive streaming | platform : PC | engine: Unity | language : C# | role: design, coding| team size: 3

Dungeon Democracy is a 3D dungeon crawler created for an interactive streaming and spectating experience, using Mixer streaming platform. Stream your game and let your viewers decide your fate. With the multiple options, spectators has the choice to choose player and boss abilities periodically. They can help and harm the player. It's all up to the streamer, how popular he can become!

My contributions:
  • Created overall game design (co-op)

  • Planned and executed game play-testing (co-op)

  • Helped with debugging several issues along the way (co-op)

  • Created UI

    • Coded logic using C# and Unity​

    • Designed UI layout

  • Created game items​

    • Built item using primitives​

    • Created visual effects using animations and scripting

    • Created interactivity features for interactable object


Team members: Terry Beyak & Christopher Huynh

We planned to create a dungeon-crawler type game that was influenced by spectators on a screen. The player, or streamer, would navigate through the dungeon and defeat back to back bosses. While fighting bosses, the spectators could vote on certain events to happen in the game. This included possible events such as reducing player health, changing player abilities, or change boss behaviors.


Our gameplay was inspired by Diablo and Path of Exile for their “point-and-click” targeting and combat system. Most of our gameplay decisions and even UI decisions were based upon these games. To get the dungeon crawler feel, we pulled some inspirations from Enter the Gungeon and Risk of Rain 2. These are rogue-like games that are focused toward providing an endless “how-far-can-I-go” experience.


The player is placed in an arena with a single boss. The player is able to use the mouse to move around the battlefield. The player can choose to use a close range auto attack to damage the boss as well as an arrow projectile. The player also has the ability to quickly dash to a targeted area and use a shield to block one instance of damage.


The player must avoid the boss’ three attack moves. The boss can inflict a high damaging ground smash at close range. When the player is out of range, the boss can choose one of two attack options. The first is for the boss to track the player and attempt a long range attack move. The other option is to do a spin move that does more damage to the player the longer they are within radius.


Using Mixer as our streaming platform, the spectators are able to vote on specific voting options to affect how the player must interact in the level. Our current choices implemented were changing the arrow types to bouncy or sticky, as well as damaging or healing the player.


We planned to have a lot more bosses, levels, and spells for the final functionality, but because this was a prototype we had to scope down to only one primary boss as well as three working abilities. We would like to add a more variety of voting options such as ones that may place obstacles in the level. We would also like to add a real time voting count to show the player which vote is about to win.


Game Spectator Modality

The spectators are acting as co-strategist for the player of the game, enacting vote outcomes that they believe will be either beneficial to, or alter the streamer’s strategy in an interesting way. This could fall into the Spectacle modality, where the player is momentarily confounded by the vote outcome, but resolves it in a profound way, or fails to, and dies despite his best efforts.


We suspect there would also be some amusement and entertainment found in the personality of the streamer’s vocal lobbying for certain choices given the strategy he or she is aiming for.


Play testing

Our first play test was mainly through using only a pitch as our core gameplay was not fully implemented yet. We talked to Curt and Pat and received the following feedback.


  • Issues with Scope: Our original plan to have multiple skills, multiple bosses, and multiple different voting options was a concern. We ended up re-scoping down to only provide 1 boss and 3 abilities for the second play test.

  • Technology: We pitched the idea to have Twitch integration but it seemed to be a big concern if we would be fighting against the Twitch API. Through the help of Pat, we decided to choose Microsoft Mixer as our streaming platform. Thankfully, they have an easy to understand API to help implement our voting functionality

  • More Interesting Voting Choices: One feedback from Curt was to make sure our voting options were not purely black and white. Our voting options had to be strategy altering for the player.


Our second play test was a live demo of our prototype.  We received a lot of feedback regarding the UI. Overall, it seemed like our gameplay was considered solid for the play test. The following are some notes from the play test.


  • The legibility of our UI was great but needed some tweaks:

    • It was easy to tell the cool downs on abilities

    • Easy to see what the vote options were and which vote won but needed more visibility when showing them.

    • Needed more legibility to correspond what each ability corresponded with which button

    • The health bar in the bottom right corner was considered unusual.

    • The boss health was first thought to be the player health.

  • Some abilities were confusing.

    • Chris Z was not sure how our shield ability (W) worked. This was definitely due to our visual effect for the shield having good accordance.

    • He was also not sure if our dash (E) did any damage.

    • He was also not sure how the vote affected the arrow functionality.

  • Interesting feedback:

    • Chris Z preferred to auto attack rather than use abilities because he felt like the auto attack was more useful and did more damage. This may be a need for us to change how our other abilities work or balancing damage numbers.

    • Because this game is focused for a streamer, it was recommended to not include any UI in the bottom right corner, as that is where streamers usually but their capture of themselves.

    • Also it would be nice to enable streamers the option to customize where the UI was on the screen.


Dungeonocracy was my chosen project to run for my final game pitch for the university capstone project. The game ran in competition with other 26 project and progressed through the first round of elimination. While it did not make it to the final 5, I am very happy to show the 10 minute presentation pitch that brought the project through the first round.

bottom of page