Simplified sequence-replication model -- week 7 pack
A simplified synthetic replication model. It has no selection, no population structure, and no biological claim attached to its symbols -- it exists to show how ancestor-agreement decays as a copying error rate rises, and where that decay stops being a reliable signal.

settings.json has the ancestor, alphabet, error rates and seeds -- everything
needed to reproduce every number in this pack from the method it also states.

results.csv is ancestor agreement (fraction of positions matching the
generation-0 ancestor) for every error rate, seed and reported generation.

plot.svg draws all nine runs; trend.txt is the same information as a table
plus a prose description, for the accessible route or a non-visual reader.

reveal/analysis.md names one conclusion this pack cannot justify, and states
what the model omits. Open it after you have written your own answer.
