Bi-layer splitting in overdoped high $T_c$ cuprates

Recent angle-resolved photoemission data for overdoped Bi2212 are explained. Of the peak-dip-hump structure, the peak corresponds the $\vec{q}=0$ component of a hole condensate which appears at $T_c$. The fluctuating part of this same condensate produces the hump. The bilayer splitting is large enough to produce a bonding hole and an electron antibonding quasiparticle Fermi surface. Smaller bilayer splittings observed in some experiments reflect the interaction of the peak structure with quasiparticle states near, but not at, the Fermi surface.