I used simple patterns (without environmental factors) to predict run sizes, relying on the persistence of a 4-year cycling and location of the relevant peaks. Fraser runs have a minimal age structure, but the run size of 1.2s often provides a forecast of the 1.3 in the following year. Where age compositions were not useful, I relied on ad-hoc impressions of patterns suggesting the year-to-year shape of the relationship, often drawing from the “shapes” of previous years as quantitative examples.

My simple calculations were done on Excel spreadsheets, and are self-documenting.