Sources & forecast method
The source behind every reset
We read the public Codex Resets feed and keep the original post links. We check the source every five minutes; the page shows the last successful check. A delayed refresh keeps the last good record visible.
An announced reset stays pending until a reset record arrives. Regular resets and banked resets are shown separately. Your individual account limit can differ from a global announcement.
The 48-hour forecast
We calculate how many completed historical intervals lasted at least as long as the current wait. Among those, we count how many ended within the next 48 hours. The percentage is shown only with five or more comparable intervals. Below 30% is Unlikely; 30–64% Possible; 65% or higher Likely. This is a historical baseline, not a calibrated guarantee.
A sourced editorial judgment can override the baseline until its stated expiry. Announced is reserved for an actual pending announcement. Expired judgments automatically fall back to history. We do not claim a historical prediction win without a timestamped forecast record.
Photos and map
Tibo weather and share cards are unofficial fan creations. Suit photo reference: Nicole Tang video; other photo references supplied by the founder. Map geometry: SVG Maps (CC BY 4.0).
Back to the tracker →