The overarching goal is to understand the mechanisms of planet formation and their respective predictions about planets' observable properties. One approach is through population synthesis, which allows an exploration of the statistical distribution (e.g. of masses and luminosities) expected from a formation paradigm. In this context, I am working with Christoph Mordasini on improving the calculation of the planets' cooling using state-of-the-art atmospheres, which also lets magnitudes in different filters be predicted. This will be particularly useful for the SPHERE project of the ESA, in which the MPIA is strongly involved and which should yield a number of detections sufficent for statistically meaningful comparisons with the predictions of complete formation models.