Flip a coin 9 times. which of the following estimators is most likely considering your data? p is probability of heads. p = 25%, p = 50%, p = 75% that is, fine the probability of seeing your data under each of these 3 possible values of p. which of the 3 had the highest probability? now try p = # heads / 9 and see how that compares. now let's try a bayasian version. assume the priori distribution on p p( p = 25% ) = 20% p( p = 50% ) = 60% p( p = 75% ) = 20% now what is your answer? note this is basically taking the assumption that you believe the true value of p is three times as likely to be 50% than either 25% or 75% before you even started taking data.