Figure 2

Left: the dot corresponds to a, namely the probability to be estimated; \(a_{1}^{\min}\) and \(a_{1}^{\max}\) define the confidence interval whose width is \(2\epsilon _{1}\). Right: the same confidence interval represented in terms of angles. Note that the “true value” (represented by either a and \(\theta _{a}\)) falls in a generic position within the confidence interval. To avoid clutter, here we are not representing the central value of the confidence interval