Note that one can prove that Maximum Likelihood is
asymptotically the ``best algorithm,'' i.e., in limit
(N total events, there is NO method which will
give a smaller standard deviation).
Also, Maximum Likelihood method will give one the ``right
answer,'' i.e., Bias = Correct Estimate---Maximum Likelihood
Estimate (like ) as .
(N total events, there is NO method which will
give a smaller standard deviation).
(like 
) as 
.