The AI of Power Pong - Learning over time

Published February 27, 2017
Advertisement

Welcome back!

In this entry I will be discussing the AI of Power Pong. A decision was made early on as to how the AI would work. Would it cheat? Would it be an instant professional? No to both. Instead, I opted to draft a more adaptable AI approach.

The AI has a few rules to consider when decision making

- Where is the ball heading?

- How much do I need to move?

- Do I need to use my skills?

- How good is my opponent?

With these rules in mind, the AI will begin (depending on the difficulty set), rather jittery, as it does not know how far to move in order to successfully deflect the ball. Furthermore it never cheats, it does not know ahead of time where the ball will land. It will only attempt to keep up with where the ball is going and make decisions based on that.

If the AI is losing too much, it will slowly start to adjust. It will start to alter its movements over time, becoming more and more accurate at deflecting the ball. The AI is not perfect, and there will always be a chance for it to miss, now matter how much it's mastered the game.

If its opponent is losing, however, there is a small adjustment that affects its learning rate, as to not exponentially increase the difficulty into the AI's favour.

Hopefully this new approach brings a more exciting gameplay experience, allowing for a more realistic AI to battle against.

The next post will discuss the mechanics of the game in more detail!

Next Entry Game Mechanics
0 likes 0 comments

Comments

Nobody has left a comment. You can be the first!
You must log in to join the conversation.
Don't have a GameDev.net account? Sign up!
Advertisement
Advertisement