Every metric explains itself in plain language, links to the research behind it, and tells you honestly how much to trust it. Learn by using the tool — or explore the whole map of ideas in the Signal Atlas.
Every metric, Greek, and concept is a node. Every relationship is an edge — "derives from," "feeds," "predicts," "contradicts." Click a node to see its definition, its evidence grade, its neighbors, and a link to the live tool that computes it. The graph is the glossary: learn one idea by wandering to the ones it connects to.
Curated routes through the desks — each chains concepts in the order that actually builds understanding, ending in the live tool.
The most trustworthy thing a tool can teach you is what doesn't work. These are the popular ideas the evidence doesn't support — and knowing them protects your capital more than any signal.
Cheap far-OTM options are systematically overpriced — the lottery preference is compensated to whoever sells it to you, not you. Boyer & Vorkink (2014).
Real pinning near expiry comes from dealer hedging (a documented effect). The idea that price "gravitates to the max-writer-profit strike" is a different, weakly-supported claim. Ni, Pearson & Poteshman (2005).
Common look-ahead filters inflate options-strategy Sharpes from ~0.5 to above 5. A too-good backtest is a red flag, not a discovery. Duarte, Jones, Khorram, Mo & Wang (2025).
Unusual-options-activity alerts are widely sold as informed-money signals but are largely unvalidated as return predictors. We show the flow — and the caveat.
That's not modesty — it's the whole positioning. Read how the evidence grading works.