Maximum likelihood estimators have nice properties.
- Large sample normality
- Asymptotically consistent, i.g. converge to true parameter as n increases
- Asymptotically efficient, i.g. have large sample standard errors no larger than other estimation methods
[reference](Stat 835 - KUMC lecture note)