In 2003 Werngren and Hoffner reported the earliest quantitative mutability study comparing Beijing and non-Beijing strains of Mycobacterium tuberculosis. Their null findings appeared to be at odds with the then-popular hypothesis favoring characterization of the Beijing genotype by mutability. Three recent attempts to reexamine the experimental data have resulted in three successively smaller p-values in the literature, each supposedly buttressing a non-null conclusion. In addition to identifying errors responsible for the three misleading p-values, we focus on salutary lessons that will facilitate future research on microbial mutability.