If posterior means are used to estimate nonlinear functionals of an ensemble of parameters, biased estimates of those functionals typically result. Posterior means are optimal under sum-of-squared-error loss (SSEL); introducing weighting and at the same time discarding SSEL optimality may lead...