Three steps. You bring data you already have. It hands back a single verdict.
Graded on how far off you actually were, not a feel-good “% right” you can game.
You’re measured against the toughest dumb guess, never the easiest, so a model that just parrots last week can’t sneak a PASS.
Graded only on data it hadn’t seen yet. Numbers from data it already saw don’t count, and we won’t quote them.
We don’t sell you a model, a forecast, or an edge, only the read on whether yours holds up. The verdict is the product, and a FAIL is a real result, often the most useful one you’ll get.
Three of them, actually. We build the three laziest ways to “forecast” anything, then grade you against whichever one is hardest to beat, never the easy one:
Grading you against the toughest of the three is the whole trick: if your data has a strong weekly or seasonal pattern, a model that just parrots that rhythm won’t pass. It has to beat the rhythm, not repeat it. Full, dated method: Method v1.2 →
The free test runs on your device. Use it to check a model yourself, as often as you like; nothing leaves your browser. Pro is for when you need to take the verdict with you: export it as a shareable proof PDF, batch many forecasts at once, and get deeper analysis. Same independent method (graded only on data it hadn’t seen) same bar. See pricing →
Independent statistical analysis of data you provide, for your internal use. Not forecasts, trading signals, or financial, investment, or trading advice; not an audit, assurance, actuarial or expert opinion; guarantees no result. © McLaughlin Capital Investments LLC, not a registered investment adviser, broker-dealer, or CTA.