Magnetic reconnection is the primary driver of magnetospheric activity by coupling the magnetosphere to the interplanetary medium with an efficiency that depends critically on its location. Several models have been proposed for the location of reconnection, but none are consistently supported by global simulations and in-situ measurements have been too scarce to fully address the problem from a global and parametric standpoint. In this work, we investigate how the spatial distributions of physical quantities known to be important in the magnetic process might constrain the location of global X-lines at the magnetopause. We use in-situ measurements from four missions (Cluster, Doublestar, THEMIS, MMS), automatically selected using statistical learning, to reconstruct the global distribution of the magnetic shear angle, current density, and asymmetric reconnection rate at the dayside magnetopause. The comparison of the magnetic shear maps from in-situ measurements with those obtained with magnetic field models reveals important spatial discrepancies for a certain range of IMF cone angles (12.5°±2.5°≤|Ө_co|≤45°±5°), but also a difference in the behavior of the lines maximizing this quantity with respect to the IMF clock angle. The parametric study of the effect of the IMF and the dipole tilt orientation shows that the IMF cone angle creates strong asymmetries in the distribution of the above-mentioned quantities and changes their dependence on the IMF clock and the dipole tilt angles. Finally, we show that the X-line constructed by maximizing a given quantity gives local orientations of magnetic reconnection that are inconsistent with the predictions suggested by local simulation studies.