In this comment, I take a critical look at the idea that Masicampo & LaLande provide support for a prevalence of p-values just below .05. Although I do not doubt the presence of p-hacking in the literature, I do not think M&L provide convincing support for the idea that it can be observed in a large heterogenous set of studies, nor that it should yield a large number of p-values just below .05 (i.e., in the 0.045-0.05 range). I provide a better model of the p-value distribution in the literature, and show how it leads to an (regrettably) often invisible increase in p-values across .00 to .05.