In 1913 an undergraduate, Alfred Sturtevant, took Morgan's fruit-fly crossing data home and stayed up all night. His insight: the frequency with which two genes recombine is a measure of the distance between them — 1% recombinants ≈ 1 centimorgan — so a whole chromosome could be drawn as a linear map.
The trick works for near neighbours because crossovers between them are rare enough to be counted one at a time. Far apart, a second crossover can undo the first: the outer markers look parental even though two exchanges happened. Those hidden double crossovers are never counted, so the raw outer-marker frequency runs short — and it can never exceed ½, the point where genes on one chromosome look unlinked.
The cure is Sturtevant's own: add up short adjacent intervals instead of measuring the ends directly, or apply a mapping function such as Haldane's RF = ½(1−e^−2d). And when one crossover discourages another nearby — interference — doubles grow even rarer than chance, which is why nearby map distances are the ones you can trust.
The simulation stopped unexpectedly — the lesson continues without it. You can move on; nothing you did was wrong.