>(2) The statistical analysis above implicitly ASSUMES that any one game between >the two teams is completely INDEPENDENT of another with the same probability of >p=0.3. Clearly if one team gets blown out on the first night, it will have >some effect on how they will play the following night, not only psychologically >but also the Coach will most likely change the game plan. Thus, the 3 game ^^^^^^^^^^^^^^^^^^^^ >series can not be modeled as three independent random events. More likely, if a team gets upset my a close score because they played flat, bad bounce/ref, etc. (the scenario I thought was originally proposed), the coach gives his team a post-game 'analysis' (aka: chews their butts out) and they play a more motivated game in the next two. I think the western 2 game series regular season format gives some insight into 2/3rds of a best of 3 game format, if someone wanted data to 'crunch'. -John Hughes [log in to unmask] [So who do we "Eastern" hockey fans now root for? ...Michigan!]