Panel a): the reduced χ² as function of blended light fraction k. The horizontal dashed line shows the 2,3 and 5σ rejection criteria. Our lightcurve model directly indicates an early type main sequence stellar host, with a 2σ upper limit for blended light of k ~ 20%. The low stellar density implies a large secondary radius, rejecting the planet hypothesis. Panel b): best fitting EB models with blended light fraction k =  [0.2,0.5,0.90,0.95], clearly showing that solutions with low k are favoured. Note that an orbit with an eccentricity of e = 0.5, orientated in the right way, could increase the estimated stellar density to that of the Sun, and decrease R₂ to 2 R_Jup. This ambiguity can be easily removed by taking a single spectrum of the star, resolving its spectral type.