Scrapper Bots was created by Milo Fisher, Jalen Pastor, Alex Basinski, Elijah Rossman, Alexander Shaham, Pierce Sullivan and was made for the Game Design Studio class at UC Santa Cruz. In this game, the player controls a team of 3 robots and must defeat all opposing robots. Each robot has its own stats and unique set of abilities. The core gameplay is to strategize with these abilities to overcome being outnumbered by the enemies. The build is just a prototype for a larger game concept, and only features one scenario with limited content and polish.

On this project, I was the lead programmer. I was in charge of implementing the core gameplay mechanics and defining the main helper functions used later in the ability effects. Of these features the most important would be the hexagonal grid based functions for getting tiles at a position, getting adjacent tiles, getting players at a position, BFS on tiles, and A* pathfinding on tiles. We used these functions in most of the game features for calculating movement, attack targets, and for enemy AI calculations.

For gameplay features, I made the initiative system, the terrain effects, the movement system, the abilities for most of the playable robots, the AI and abilities for the enemy robots, and the camera panning and focusing. The initiative system orders all the units to take turns based on their initiative value, and players may hold their action once per turn to move themselves to the end of the initiative queue. After the queue is expended, it is then refilled with the current initiative values since they can change during the game. The different tiles each have unique properties, and I implemented each of them. Some, like the wall and scrap pile, are not traversable by units, while others like water, electrified air, and mud change a unit's speed or initiative stats. The movement system essentially calls a BFS on the tiles starting from the unit's tile for a length of the move range, and marks all checked tiles as moveable. The abilities in the game have a large variety of effects, so for the most part implementing them was just combining our helper functions to create unique movement/attack effects. For the AI, there are two types of enemies. The first is a melee unit so the behavior is quite simple: it pathfinds to the nearest target and attempt to get within melee range so it can attack the target. The second enemy is a ranged attacker, so I gave it slightly more advance AI. The unit checks all tiles it can move to and ranks them based on a number of factors: is there a line of sight to a target (in range) and are there enemies nearby. It then picks the highest rated tile according to those parameters to move to. Finally for the camera panning and focusing on targets I defined a function to pan at a constant speed without using coroutine calls. This way the function can be overwritten or called multiple times and it would work without issue.