We consider multipole amplitudes for low-energy pion photoproduction, constructed with minimal model dependence, at single energies. A previous fit making minimal use of Watson's theorem has been reexamined in light of more recent measurements. Problems associated with the choice of a unique solution are discussed. Comparisons with more recent fits to the full resonance region are made. Explanations are suggested for the discrepancies and more precise measurements of the recoil polarization, $P$, are motivated.