Psy/Orf 322     Human – Machine Interactions          Spring 2004


Comments on Prof. Kornhauser’s Homework #1  CMN.


#1   Many of you said that the decision processes had to occur “at the same time”.  Not so… While mathematically, nothing occurs “at the same time”, the “window of opportunity” defined as “the time difference between the inception of each of the individual’s decision process that lead to each of their individual action” is finite and “quite” large; thus, the occurrence of this phenomenon is not rare.  Many of you failed to realize that as long as the decision for the second person to move Left started while the person implementing their decision to move Right, was during this “window of opportunity” the instability will tend to occur. This time span may be as long as 240 msec (the nominal time that it takes to complete a sequence of perception (100 msec, you are coming right at me), cognition (70 msec, I’d better move Left) and motion (70 msec, I’m moving Left)).   


If we reset time to be zero when the first person (A) begins the final sensing process that leads to their decision to move Left.  If now the second person (B) starts their decision process to move Right within up to the next 240 msec , but certainly within the next 100 msec, then they are doomed to continue on a collision course.  While B is committed to a movement to the Right, they will sense that they are on a collision course and will implement a move to the Left which will take 240 msec to implement and in the mean time their original motion  to the Right is being implemented.  A finally seeing that B has moved Right now begins the process of moving Right, but unfortunately, this comes after B has begun its implementation to move Left.   The instability continues and, in fact, can continue for several cycles until one decided that this is ridiculous and pressed against the wall.  (By the time each person begins to implement a motion, that motion is based on “old (up to 240 msec) information” that is contrary to the motion that should be taken, so the process to immediately correct the decision begins, but because of the delay in the implementation, that decision is contrary to the motion that should be take, so the process to immediately correct …   (and so on, the instability!!)



#2  Goalies know to cover most of the goal, but can’t cover it all.  They also like to believe that they can react fast enough to make saves, if the shot heads toward an unprotected area.  However, there simply isn’t enough time to react to the shot.  The time for perception, decision and action is simply too time consuming, even in the best of situations.  Every shooter knows that. 


For the first two shots, the positioned herself to cover as much of the goal as possible and relies on reaction to make the save.  The shooter shot to the opening and the goalie didn’t have time to react.    So… realizing that she will not have time to react, the goalie decided to again cover as best she can, be cognizant of the openings that she is leaving and deciding to initiate the move to cover that opening, before waiting to react to the shot but late enough to not give the shooter an opportunity to change her shot is response.  The bottom line is that the goalie “suckered” her to shoot at an opening,  guessed properly which opening she felt was most inviting and made the save!  She had no choice.  ( by the way, the goalie was also prepared to react to the shot if the shooter was so incompetent to shoot early enough that it would have given her time to make the save.  The shooter was NOT that incompetent!)