The current wealth of charmless B decay data may suggest the presence of final state rescattering. In a factorized amplitude approach, better fits are found by incorporating two SU(3) rescattering phase differences, giving δ ~ 65° and σ ~ 90°–100°. Fitting with unitarity phase ϕ3 as a fit parameter gives ϕ3 ~ 96°, the CP asymmetries Aπ π, Sπ π agree better with BaBar, and the σ phase is slightly lower. Keeping ϕ3=60° fixed in fit gives Sπ π ~ -0.9, which agrees better with Belle. With the sizable δ, σ rescattering phases as fitted, many direct CP asymmetries flip sign, and B0→π0π0, K-K+ rates are of order 10-6, which can be tested soon.
On the role of the final-state interactions in rare B decays
