Methods. We compute Bayesian evidences and Bayes Factors for a set of variations of the classical radial models by King (1962), Elson et al. (1987) and Lauer et al. (1995). The variations incorporate different degrees of model freedom and complexity, amongst which we include biaxial (elliptical) symmetry, and luminosity segregation. As a by-product of the model comparison, we obtain posterior distributions and maximum a posteriori estimates for each set of model parameters. Results. We find that the model comparison results depend on the spatial extent of the region used for the analysis. For a circle of 11.5 parsecs around the cluster centre (the most homogeneous and complete region), we find no compelling reason to abandon Kings model, although the Generalised King model, introduced in this work, has slightly better fitting properties. Furthermore, we find strong evidence against radially symmetric models when compared to the elliptic extensions. Finally, we find that including mass segregation in the form of luminosity segregation in the J band, is strongly supported in all our models. Conclusions. We have put the question of the projected spatial distribution of the Pleiades cluster on a solid probabilistic framework, and inferred its properties using the most exhaustive and least contaminated list of Pleiades candidate members available to date. Our results suggest however that this sample may still lack about 20% of the expected number of cluster members. Therefore, this study should be revised when the completeness and homogeneity of the data can be extended beyond the 11.5 parsecs limit. Such study will allow a more precise determination of the Pleiades spatial distribution, its tidal radius, ellipticity, number of objects and total mass.
Download