Evolution can happen by pure luck, no survival advantage required
Imagine four bacteria surviving a shrinking drop of water, drawn at random from a colony split evenly between two gene versions. The odds that those four still carry an even split are only 6 in 16. That quiet reshuffling by chance, called genetic drift, changes populations even when neither version helps anyone survive.
Genetic drift is a shift in how common a gene variant, or allele, is in a population caused purely by chance. A jar of 20 marbles, half red and half blue, makes the idea concrete. Build the next generation by drawing a marble at random, dropping a new one of the same colour into a fresh jar, and returning the original, twenty times over. The new jar will rarely hold exactly ten of each. Repeat for many generations and the balance wanders, and if one round happens to pick no red at all, red is gone for good and blue is fixed.
Because chance can delete an allele but never restore one, drift steadily pushes populations toward uniformity. It stops only when a variant hits 100 percent, after which change needs a new mutation or migrants bringing genes in. The chance that a neutral allele eventually takes over equals its current frequency: one carried by 75 percent of a population wins three times in four. Fixation comes far faster in small groups, and the effective population size, which accounts for inbreeding and bottleneck stages, is what matters. A sudden crash to a handful of survivors, a population bottleneck, is drift at its most dramatic.
Biologists model the process mathematically. The Wright–Fisher model, named for Sewall Wright and Ronald Fisher, assumes each generation replaces the last entirely, like annual plants. The Moran model lets generations overlap, with one birth and one death per step; its equations are easier to solve, and drift runs twice as fast in it, though both give similar pictures.
How much drift matters was fiercely argued in the mid-20th century. Fisher, who united natural selection with Mendelian genetics, thought drift played only a minor role, and that view held for decades. In 1968 Motoo Kimura revived the argument with his neutral theory, proposing that most genetic changes spreading through populations come from drift acting on mutations that make no difference. Drift is one of four evolutionary forces on allele frequencies, alongside migration, mutation and selection.
Source: Genetic drift
