Read the explanation
The default probability is anchored at point one three with fifteen percent capex growth, negative eight percent token revenue growth and energy index one hundred forty two. The capex slowdown preset changes those to negative fifteen percent, negative eighteen percent and one hundred sixty. Its three additive contributions are thirteen point five, three point five and two point seven percentage points, yielding thirty two point seven percent. Bars compare thirteen and thirty two point seven at five pixels per percentage point. These are linear authored coefficients, despite a comment describing nonlinear capex impact. Liquidity does not enter the probability formula. No actual market price, bubble forecast or historical evidence validates the assigned thirteen percent anchor. The capex preset differs from the thirteen percent contract anchor by nineteen point seven percentage points. Squaring point one nine seven yields point zero three eight eight zero nine. Bars use one thousand pixels per squared fraction unit versus the baseline zero. The interface calls this a Brier score, but it compares one probability with another fixed probability rather than a resolved binary outcome. It is therefore an anchor-distance square, not a proper forecast calibration score. Its peak quarter also comes from probability thresholds: above thirty five selects fourth quarter twenty twenty five, above twenty selects first quarter twenty twenty six, below eight selects second quarter twenty twenty seven. These quarter captions do not consult a current clock or event observations. The intended Monte Carlo panel generates sixty twelve-quarter paths. Each step pulls twenty percent toward the heuristic probability and adds uniform noise from negative point zero two to positive point zero two, before probability clamps. Bars count sixty paths and fifteen synthetic order book levels per side at five pixels per item. Volumes use configured liquidity divided by fifteen and level index, times a random multiplier between point eight and one point two. These are constructed drawings, not actual exchange orders, independent simulations of economic fundamentals or statistical evidence. Missing D3 throws at a top level selection before control and export listeners bind, so native capex edits leave thirteen percent displayed and no actual export occurs offline. Local layout and playback proof preserve that failure and source footers without public publication or exhaustive validation.