The human antibody repertoire has an exceptionally large capacity to recognize new or changing antigens through combinatorial and junctional diversity established at the time of V(D)J recombination and through somatic hypermutation. Influenza viruses exhibit a relentless capacity to escape the human antibody response by altering the amino acids of their surface proteins in hypervariable domains that exhibit a high level of structural plasticity. Both parties in this high-stakes game of shape shifting drive structural evolution of their functional proteins (the B cell receptor/antibody on one side and the viral hemagglutinin and neuraminidase proteins on the other) using error-prone polymerase systems. It is likely that most of the genetic mutations that occur in these systems are deleterious, resulting in the failure of the B cell or virus with mutations to propagate in the immune repertoire or viral quasispecies. A subset of mutations is tolerated in functional surface proteins that enter the B cell or virus progeny pool. In both cases, selection occurs in the population of mutated and unmutated species. In cases where the functional avidity of the B cell receptor is increased significantly, that clone may be selected for preferential expansion. In contrast, an influenza virus that "escapes" the inhibitory effect of secreted antibodies may represent a high proportion of the progeny virus in that host. The recent paper by O'Donnell et al. [C. D. O'Donnell et al., mBio 3(3):e00120-12, 2012] identifies a mechanism for antibody resistance that does not require escape from binding but rather achieves a greater efficiency in replication.