You are green. The AI trained with a character that just went up and down. This is my second AI That is doing well. If you want to see more like this please like and share.
I used an artificial Neural Network The numbers under the Difficulties show how many iterations he trained When I tested V1 I found out that instead of learning to hit the ball it learned to follow who it was playing against because it trained with someone who just went up and down. So he learned to copy his opponent instead of play. So I fixed it by not allowing him to see his opponent. Check out how it went :) He is good V1 : https://scratch.mit.edu/projects/991035893/
