Genetic Drift, explained.
Genetic drift is random change in allele frequencies caused by finite sampling, even when variants have equal fitness.
Why it happens
Each new generation samples N independent allele copies with A probability equal to the previous frequency. Start at 50%; change the population size or seed to compare paths. Zero and one are absorbing because mutation and migration are absent.
Genetic drift is random change in allele frequencies caused by finite sampling, even when variants have equal fitness.
Read the result
Read the tile count and the path together. A jump without a fitness difference is sampling variation, and absorbing endpoints matter more than one dramatic intermediate step.
A worked example
A small founder population
Keep thirty allele copies and sample several generations from the initial fifty-percent frequency, noting the realised counts.
Reset and increase to two hundred copies. Repeat sampling with equal fitness still imposed on both alternatives.
Compare proportional fluctuations, recognising that a particular path is not proof that every larger population changes less in every generation.
OPTIONAL DEEPER DETAILGo deeper: inside the model
Inside this model
Each new generation samples N independent allele copies with A probability equal to the previous frequency. Start at 50%; change the population size or seed to compare paths. Zero and one are absorbing because mutation and migration are absent.
Where this idea is useful
A practical use
A founder population may carry very different variant frequencies from the source population through sampling alone.
A common misconception
“If an allele becomes common, it must improve survival.”
Neutral sampling can change allele frequencies without a fitness advantage. Distinguishing drift from selection requires more evidence than a rising line.
What this explanation leaves out
- This is a neutral haploid Wright–Fisher-style model with fixed copy count, independent reproduction and no selection. A short path does not establish the fixation probability or time in a real population.
Why does the generation button stop after fixation?
Without mutation or migration, sampling from all A copies always produces A copies. Lost alternatives cannot reappear under this model.
What would convince you that an observed frequency change exceeds plausible sampling variation?
Associated thinkers
Associations marked provisional are awaiting source review.
Further reading
Explore the original research or the teaching reference behind this experiment.