Downloadable to follow
This is both a game and demo. I created a model for a football game inspired by tactical chalkboards. The player can direct the team’s movements and passing in real-time, as well as pausing the action to give commands. The game and techniques are ideal to be run and used on relatively low-powered hardware.
The number of attackers and defender vary with level or scenario, while the objective can either be to complete a number of passes, to score a goal or to get an attacker into the area with the ball. The defenders are either given set actions to perform or use a team AI algorithm I devised. Rather than using some of the standard AI approaches, my technique calculates defensive movement based on the relative positioning of on-pitch agents. It allows for the quantitative variation of:
Aggressiveness towards the ball
The tendency of defenders to mark attackers
The team’s cohesiveness
Positional rigidity within the set formation
This means different defensive tendencies can be created and modelled, either for the defending team as a whole or to create variation between individual defenders.
The algorithm was further changed and refined to eliminate its main weakness (as illustrated in the video). The video shows some of the features of the demo and the game - the code contains many more examples and variations.