top of page
Search

Baa~Howl~~

Updated: Mar 29



Global Game Jam Page: Go To Page

Download Game Now: Download Now

GitLab Repository: Project Source Code


Game Overview

Baa~Howl~~ is a top-down stealth game created for Global Game Jam 2022 under the theme "Duality". In this game, the player takes on the role of a wolf disguised as a sheep, blending in with the flock while secretly hunting them. The objective is to eat 20 sheep before the huntsman is aware of it. If the player is caught, game over.


How to Play

  1. Move with "W, A, S, D".

  2. Eating sheep with "E".

  3. Avoiding the huntsman.


My Role

Gameplay Programmer


Development Purpose

I participated in Global Game Jam 2022 as a programmer in Pussycat's group, where we developed Baa~Howl~~ under the theme "Duality". Our goal was to create a top-down stealth game that explores the dual nature of disguise and deception.


Figure 1: These are the in-game graphics and the sheep
Figure 1: These are the in-game graphics and the sheep

Core Mechanic that I created


Sheep AI

This project features sheep that players must consume. The sheep have a grouping system called "Flock", which calculates cohesion based on three factors: Cohesion, which calculates vcentre to ensure each sheep tries to move towards the group and does not stray. Separation, which uses vavoid to check distance. If the sheep are closer than 1 meter, the sheep will push itself apart to prevent their models from overlapping. Finally, Alignment, which uses gSpeed ​​to adjust its speed to match its surroundings, ensuring the flock moves at the same pace. There are also sub-systems like Boundary Control, which checks the distance from the map center. If a sheep goes outside the boundary, the turning variable will become True, forcing it to turn back towards the center immediately.


Figure 2: Sheep Flocking AI logic utilizing cohesion, separation, and alignment rules alongside boundary control to maintain group behavior within the map.
Figure 2: Sheep Flocking AI logic utilizing cohesion, separation, and alignment rules alongside boundary control to maintain group behavior within the map.
Figure 3: An example of how Flock works in Sheep AI.
Figure 3: An example of how Flock works in Sheep AI.

Huntsman AI

This project also features an AI that players must avoid: the huntsman. This AI constantly monitors the player, and if a player is caught by it, the game ends immediately. A State Machine system uses IEnumerator and Coroutine to generate random behaviors, such as idling, walking, stopping, and randomly turning left or right.


Figure 4: Huntsman AI in game.
Figure 4: Huntsman AI in game.

Furthermore, the AI ​​has a Vision System to detect players using Efficient Scanning. Instead of checking every frame in an update, it scans every 0.2 seconds, significantly reducing CPU load while maintaining player responsiveness. Even if a player is within the huntsman's line of sight, a wall will obstruct the huntsman's view, a benefit of the Raycast system used with the huntsman. Once the huntsman sees a player, the system automatically sends them back to the respawn point to restart the game.


Figure 5: Efficient vision system utilizing periodic scanning and raycasting to detect players while accounting for environmental obstructions and triggering respawn logic.
Figure 5: Efficient vision system utilizing periodic scanning and raycasting to detect players while accounting for environmental obstructions and triggering respawn logic.

Tools Used:

  • Unity 2021.2+

  • GitLab

  • SmartGit


Global Game Jam Page: Go To Page

Download Game Now: Download Now

GitLab Repository: Project Source Code



 
 
 

Comments

bottom of page