I suggest that you watch the lectures again, because I certainly do not use the probabilities when using maximin, maximax, or minimax regret criteria!!!! It would be complete nonsense.
Probabilities are only relevant when calculating expected values.