I show how simple modifications of existing two-point statistics can reap large rewards. In the first half of the talk I discuss an optimal strategy for extracting information from the redshift-space distortion (RSD) signal using projected angular statistics. This is a first step towards performing an optimal joint RSD X weak lensing analysis. I then present k-cut and x-cut cosmic shear. These statistics allow us to extract more information from the weak lensing signal while remaining robust to small scale modelling uncertainties at virtually no additional computational cost.