Public forecast documentation
Normalized forecast error explained
The audited display uses absolute prediction error divided by observed price. That percentage-normalized quantity is not raw-unit mean absolute error (MAE).
Historical audit snapshot created 2026-09-10 — not a live/current forecast
Definition, units, and reporting
normalized absolute error = |prediction − observed| / |observed|
This is dimensionless and is often displayed as a percentage after multiplying by 100. MAE instead averages absolute errors in the original price unit. A future dataset should publish both the row-level numerator and denominator, aggregation method, sample count n, and results by horizon.
The supplied audit did not include a dated aggregate normalized-error value, so none is fabricated here.
Data provenance and recommended public export
The statistics shown here come only from the forecast audit snapshot created 2026-09-10. Current values remain on the external live product. The local scheduled-worker forecast table is a distinct implementation and is not represented as that external production feed.
Recommended source: a versioned, downloadable JSON and CSV dataset at the live product, with immutable issued and resolved rows, UTC timestamps, source venue, sample count n, scoring formulas, model version, and a content hash. The Dataset JSON-LD on this page preserves the audit date rather than implying freshness.