If I reward the AI for healing their pokemon (HP goes up in general) it may learn to use pokemon centers which enables both healing as a behavior and potentially sets the pokemon center as a further along home allowing for less backtracking upon whole party fainting / blacking out
If I reward the AI for healing their pokemon (HP goes up in general) it may learn to use pokemon centers which enables both healing as a behavior and potentially sets the pokemon center as a further along home allowing for less backtracking upon whole party fainting / blacking out