The saved chart model divides four point five million streams by ninety for fifty thousand points and thirty-two million radio impressions by sixteen hundred for twenty thousand. Sales fifteen thousand add directly. Bars compare stream and radio base points at point zero one pixels per point. These are authored coefficients, not a verified current chart methodology, certified consumption or actual artist data. No chart provider or model API is called. At default target week three, sales remain fifteen thousand and stream growth factor is one. Radio factor point forty-five plus three times point eighteen is point ninety-nine, giving nineteen thousand eight hundred. Bars compare sales and radio at point zero two pixels per point. General aggregate would then be eighty-four thousand eight hundred. Target week times seven supplies twenty-one days-to-peak, a selected timeline rather than observed performance. Exact default streams, sales, radio, target week three and decay one trigger an override setting week-three total eighty-five thousand rather than the computed eighty-four thousand eight hundred. Bars compare at point zero zero five pixels per point. Later stream factor declines by point twelve per post-target week and floors at point twenty-five, while radio factor caps at one point two. Native complete stable JSON export and fixture reset preserve source assumptions and chart; actual rankings, chart rules and all other controls remain unverified.