Compare offspring counts with an inheritance model

An expected count need not be observed

For eight independent Aa × Aa offspring, expected genotype counts are 8/4 = 2 AA, 8/2 = 4 Aa and 8/4 = 2 aa. Expectation describes an average across repeated samples under the model. It does not assign particular genotypes to eight scheduled positions.

The simulator draws an allele from each parent, combines the pair and records the result. A seed makes the pseudorandom sequence reproducible. The displayed offspring are synthetic; they are not experimental observations.

Follow one sample as it grows

Choose a seed before inspecting outcomes. Increase the count from 8 to 12. The first eight remain the same; four more are added. These nested samples are not independent replications. A new seed creates another programmed sequence, but choosing only seeds that agree with the expected ratio biases the evidence.

Under independent draws, larger samples tend to have smaller proportional fluctuations. That does not guarantee that each extension of one particular sample moves every fraction closer to its expectation. Inspect the actual counts rather than treating convergence as a monotonic animation.

Quantify one unusual result

The probability that one offspring is aa is 1/4. For four independent offspring all to be aa, multiply four factors: (1/4)⁴ = 1/256. Such an outcome is unusual but possible under equal segregation.

An unexpected result is a reason to investigate, not automatic proof that inheritance works differently. Plan the sample size and comparison in advance; retain all results. For real data, examine counting errors, survival differences, parental genotype identification and whether the model’s assumptions apply.

In your reflection separate a probability claim, an observed sample fraction and a proposed explanation. Replaying one seed checks reproducibility of the software; it does not create new biological evidence.