We reconsider the predictions of inflation for the spectral index of scalar (energy density) fluctuations (ns) and the tensor/scalar ratio (r) using a discrete, model-independent measure of the degree of fine-tuning required to obtain a given combination of (ns, r ). We find that, except for cases with numerous unnecessary degrees of fine-tuning, ns is less than 0.98, measurably different from exact Harrison-Zel'dovich. Furthermore, if ns >or= 0.95, in accord with current measurements, the tensor/scalar ratio satisfies r >or= 10(-2), a range that should be detectable in proposed cosmic microwave background polarization experiments and direct gravitational wave searches.