Philotica Forecast Studio, public beta
Two statistical forecasting pilots, a personal-forecast workspace, and an experimental link to the χ chaos-screening work — now open for testing.
Philotica Forecast Studio is now open for public testing at philotica-forecast-studio.philotica.workers.dev.
It brings together three things. P1 estimates the probability that US Lower 48 natural gas storage falls below a chosen threshold, from EIA data. P2 estimates the probability of typhoon approaches to Japan meeting a chosen monthly threshold, from JMA statistics. A third module, My forecast, is a workspace for turning your own question into a properly specified forecast – event, deadline, resolution rule, evidence and counterarguments – without asking the app to invent a probability for you.
Both statistical pilots also carry an optional χ module: an experimental alternative estimate connecting the application to the ongoing Screening Chaos / IEH / PNP research programme. It is exactly that – an operational test of whether patterns in a model's own past errors carry information a baseline is missing – not a claim that the theory is proven, and not a claim that the experimental number is better than the baseline. Some comparisons currently favour the baseline; that result is reported as-is.
The release is marked Experimental Beta on purpose. Both the application and the underlying research methods are still being worked on, and public testing – including the feedback form built into every page – is part of how that happens. Full background, method notes and the relevant citations are on the Tools page.