Reading an Artometrics report from scratch — David Lee walks through observed data, derived metrics, editorial indices, and outside literature — the four evidence types behind every report. —
Written in byDavid Lee Editor, Atlas & Power
transcript
Interviewer: David, Artometrics publishes a shared methodology page. Why make readers learn vocabulary before they read a report?
David Lee: Because confusion erodes trust faster than a weak chart. When we say “observed data,” we mean a value taken directly from a public table or API. “Derived metrics” are calculations we performed — medians, shares, ranks. “Editorial indices” are interpretive frameworks when no official measure exists, and we label them as interpretation. “Context literature” is the scholarship or agency documentation that explains what the data can prove.
Interviewer: Give an example where the distinction matters.
David Lee: Take a sports dynasty report. Wins and championships are observed. A dynasty index that weights playoff performance across eras is derived — and we show the formula. Calling a team “the most glamorous franchise” would be editorial unless we define glamour with cited inputs. Readers should know which layer they are looking at without guessing.
Interviewer: What mistakes do data-curious beginners make?
David Lee: Treating precision as proof. A clean visualization does not make a non-representative sample representative. Another mistake is ignoring licensing — some datasets are public to read but not to republish wholesale. We cite sources in the report body and on charts so the inspection path is obvious.
Interviewer: Where should someone start on the site?
David Lee: Pick a desk you already care about — Culture if you follow anime or music, Atlas if you think in cities, History if you like long timelines. Read one report end to end, then open the methodology page. The goal is not to memorize jargon. It is to know where the claim lives in the evidence stack.