Original Post
Here is a quick idea I just had while thinking about fuzzy logic and finite state machines. Stepped State Machines: -Instead of have states like Attacking they have state values such as Attacking7 -The is a upper bound to the states value -The lower bound is zero -The upper bound of states can be changed -Actions have thresholds at witch hey occur -The thresholds are between 0 and 1 -Threshold checks are done as follows if( ( StateValue / StateUpperBound) >= Threshold) take action In summary -It is kind of like a FSM with a variable number of states. -Actions will occur under similar circumstances with a different number of steps. -Changing the number of steps basically changes the precision with witch the actions will be taken I think it could be useful for AI especial RTS games. Any thoughts or comments please.