The Eclipse

Unity Engine

Dungeon Crawler Jam 2023

Developed over 6 days

The Eclipse was my submission for Dungeon Crawler Jam 2023. It is a first-person Dungeon Crawler that resolves combat with trick-taking gameplay. I was responsible for the full design of the project. Third party resources were used for unit models, music, fonts, and animations. This was my first time designing/coding, card based combat, a modifiable deck, and also enemy AI. To invoke a retro aesthetic for the game, I also experimented with a shader to pixelate the camera.

Core Gameplay Loop

The original scope of the game called for character classes/abilities, action/modifier cards, and environmental interactions. The scope needed to be reduced to fit within the time-limit of the jam, so the core focus became combat.

  1. Grid-based first-person level exploration

  2. Player engages enemies or interacts with healing orbs

  3. Battle enemies with trick-taking gameplay and loot new cards

Combat Design: Trick-Taking Card Gameplay

  • The trick-taking combat is an original design and where much of my focus ended up in this project. Combat decisions are meant to feel tactical and flexible, but most trick-taking games involve 3+ player counts to accomplish this. My design requires two cards to be played by each party to make 1v1 combat also have interesting card play opportunities. Requiring 2 cards each allows for bluffing and informed decisions that are normally only available for higher player counts.

  • The jam theme of Duality is a core mechanic in combat due to the reversible win condition. It can be highest wins under the Sun or lowest wins under the Moon. Both players and enemies will have access to cards that will change the win condition when played intro a trick. This opens up satisfying combos and interesting deckbuilding choices.

  • Unlike many trick-taking games, the goal is to not take the trick. Instead the loser will take damage per card involved in the trick. This makes hit points a tactical resource for players as well as a lose condition. 

  • The Parry mechanic encourages players to make high-risk, high reward gameplay decisions since it increases the damage of the next trick. It also provides an alternate choice to using health points as a resource. Parrying a trick you can't win to push damage towards trick you might win instead is exciting.

Grid Movement and Instructive Level Design

  • To fulfill the jam requirement of step movement, I designed a node based movement system. This also allowed me to easily fulfill the requirements of 90 degree turns and for all "grid squares" to be equal to the size of the walls.

  • The nodes are designed to be located with world position coordinates. Therefore we can use world position coordinates and relative directions to check for surrounding nodes.

  • All levels began with a basic blockout that would then be refined to offer various routes to the boss of each level. 

  • Players are rewarded with cards and healing items through exploration.

  • At the start of the game, the player will start with low hit points and after a few steps they will come across their first healing orb. They cannot progress until they interact with the orb. This sidestepped the need for a tutorial while teaching a core gameplay mechanic.

Enemy AI

  • The goal was to design interesting rather than "hard" AI.

  • The original AI prototype would simply select cards at random. This basic unpredictability was functional for the scope of the jam, but I felt it was important for combat to feel more meaningful.

  • To improve the AI, I had it check logic during different states each turn. This would allow it to check who was playing first and what suit it should play.

  • Each turn the AI would check if it's hand contained a wild and there was a small chance it would play a wild instead of playing in suit. This gave a splash of inconsistency that would keep players on their toes. Players couldn't assume the enemy no longer had a suit thanks to this change in AI logic.

  • Overall, the enemy AI would usually play in suit each turn, but there was a small chance it would not follow suit or play a wild. This isn't the most elegant solution, but it was my way to create a sense of trickery or foolishness by the AI. To give it personality rather than be a math problem the player can solve. 

Post-Jam Evaluation

What went well?

  • The game feels complete, unique, and incorporates the theme of Duality into gameplay.

  • The trick-taking combat is fun, thoughtful, and has plenty of room for growth.

  • The deck-building allows for unique playthroughs and compelling player progression.

  • Sufficient play-testing caught a lot of game-breaking bugs.

What went wrong?

  • Some players felt the gameplay or movement was a bit slow.

  • The UI was very practical and didn't enhance the game's aesthetics. 

  • Mostly consistent drag and drop of cards in combat UI, but it could be unreliable sometimes. 

  • Camera bug that could sometimes place the enemy out of view during combat.

  • Suit typo in game jam build - lesson to not rely too heavily on autocorrect.

What would scaling up this project look like?

  • I really enjoyed designing the trick-taking gameplay, deck building, and dungeon crawler mechanics. This is something that I want to continue working on and do a full release of in the future. 

  • I would like to add more RPG mechanics, exploration events/interactions, and expand the deck building options.

  • For a full release I would need to address the issues found in the jam build and improve the game using player feedback.