A Cell study led by Gabriel D. Victora at The Rockefeller University tracked thousands of B cells across more than 100 germinal centers in mice and found the process of affinity improvement is far more selective than previously thought. The work suggests germinal center evolution operates through repeated, nearly random rounds of mutation and selection rather than rare, high-growth “bursts” from the best-performing clones. The findings, published as “Replaying germinal center evolution on a quantified affinity landscape,” reshape assumptions about how immune systems maintain weaker clones and how consistently they arrive at high-affinity solutions—insights that could guide vaccine design for rapidly mutating pathogens.