DJ Calendar Builds Tour Predictor From Past Show Data Using Simple Clustering

The DJ Calendar, a site tracking upcoming shows for 99 DJs, developed a tour prediction feature drawing on its archive of 629 past shows across 86 artists. The system uses PHP and a single SQL table rather than machine learning or third-party APIs. It groups shows into tours whenever consecutive dates fall within 45 days of each other, then estimates the next tour window from the average gap between clusters. Artists with fewer than three recorded shows are excluded, leaving 38 of the 99 on the roster without predictions. Predictions are displayed on artist pages along with a confidence level, such as an estimate of 55% for John Summit's next window.