When Should You Use and When Maximum Likelihood?

• Which is best? Likelihood method is ``best,'' but method is computationally easier as .

• Typically, as number of events in each bin must be large

  

• method will give an estimate with a larger standard deviation, but the difference will not be large if is small, so probabilities do not vary across bin. So, can choose correctly

  Small total number of events---no good method.

  Large number of events--Maximum Likelihood works and is best.

  Very large number of events---Maximum Likelihood and works, but will give a good answer with less computational complexity.