The skewed ê_CHP distribution as an example of the effect of asymmetry in a sheared ellipticity distribution. Plotted are the estimation biases Δe for 10⁵ simulation runs, shown as a density in grayscale. Over-plotted are the convex hulls at approximately 38.3%, 68.3%, 86.6%, and 95.4% coverage. The mean of the distribution is shown as a white plus. Top: estimation biases Δe_CHP for samples with an underlying shear of g = −0.21 + 0.21i (left) and g = 0.3 (right). Note that these estimator distributions are effectively mean-biased, because they are skewed, but still centered around Δe = 0, as indicated by the CHP estimation of the distributions. Bottom: Δe_CHP for the same samples, after iteratively de-shearing the samples until the final CHP estimate vanishes. These iterations remove asymptotic mean-bias and increase efficiency.