How Weighting Decisions Get Made Behind the Scenes
Launch library · evergreen read
Every composite ranking ultimately rests on a set of weighting decisions, essentially how much each input factor counts relative to all the others, and those particular weights are rarely arrived at through any single obvious or fully automatic method at all. These choices are rarely arbitrary, but they are also rarely as scientifically settled as the finished ranking might suggest.
Compilers typically balance practical considerations, like which data is reliably and consistently available, against more philosophical ones, like which factors best represent the kind of success the chart is actually trying to capture in the first place. Data availability often quietly constrains an ideal scheme.
Because reasonable people can genuinely weight the same factors differently from one another, two well built rankings using identical raw data can still end up producing two genuinely different, equally defensible final orders in practice and outcome. Recognising that reasonable disagreement is possible is a healthy way to approach any ranking encountered.