1 paper
Michael Hardy
We give a visually appealing counterexample to the proposition that unbiased estimators are better than biased estimators.