Posterior correlation plot demonstrating an injection-recovery exercise with a synthetic inhomogeneous phase curve signal injected into the real Kepler light curve of KIC 8424992. The blue lines indicate the values chosen for each of the injected phase curve parameters. The most challenging inference in this analysis is determining which fraction of the phase curve signal comes from reflection or thermal emission, since higher geometric albedos and hotter daysides both increase the planetary flux. This produces the anti-correlation between C₁₁ and A_g, since a cooler dayside can produce a similar phase curve to a more reflective atmosphere. The model produces accurate inferences despite this degeneracy by fitting for complementary parameters that define the shape of the phase curve at all phases, such as x₁, x₂, and g.