Current definition logreg-v2. Charts stamped with it were computed by src/metrics/logreg.py, and every value plotted for a past date could have been computed on that date.
Because appreciation is rapid early and settles over time, log-log linearises the trend. Bands above and below mark historical over/under-extension.
The fit is expanding by default, and that is a deliberate departure from how everyone else draws this chart. The conventional presentation fits once over the whole history and plots that line back to the start — which means the 2013 fair value is drawn using 2026 data. That is exactly the lookahead the project's core rule forbids for normalisation, and it is the same mistake in a different place. Expanding fit answers "what did fair value look like at the time".
Pass expanding=False for the conventional view. It will look tidier and it is a lie about the past.
Newest first. Each entry says what changed and whether values that had already been published moved with it, followed by the sentence from the module that decided the verdict.
project() added for the projected-x chart shape. Purely additive: compute(), residual() and bands() are byte-for-byte the v1 definitions and every historical value is unchanged. Only the stamp moves.
Output was added. Nothing already published changed, so every chart drawn under the previous version still reads the same.
Purely additive: compute(), residual() and bands() are byte-for-byte the v1 definitions and every historical value is unchanged.
Only the stamp moves.
expanding OLS on (log t, log price), residual-sigma bands.
The first version of this metric. There is no earlier definition to restate.
44 charts name this metric in their readout line. 33 name it first, so a count taken from the leading stamp alone reports 33 rather than 44. A chart built on several definitions carries all of them, and being second in that list is not the same as being unused.