Optimizing the basis of B → K*l+l- observables in the full kinematic range

We discuss the impact of recent LHCb data on the B -> K* mu+mu- decay for New Physics searches. We use an optimised set of observables with a limited sensitivity to hadronic inputs (form factors), leading to a significant deviation from the Standard Model for the semileptonic operator O9. We ascertain the robustness of our results with respect to models of power corrections, charm-loop effects, factorisation approaches, possible effects from resonance tails at low-recoil, choice of observables and binnings. We explain some contradictory results obtained for the chirality-flipped operator O9prime by different analyses by focusing on the sensitivity of observables at low recoil afflicted by the psi(4160) and possibly other resonances for B+-> K+ mu+mu- and illustrating the need for a finely binned analysis at large recoil (rather than a wide [1-6] bin). Finally, we present updated results including experimental correlations in B -> K* mu+ mu- observables, confirming the deviation seen with respect to the SM hypothesis (4.2 sigma for large-recoil data, 3.5 sigma for large- and low-recoil data).