
Field guide
For when a number is not yet evidence.
Statistics Field Guide
Start here when a dashboard, model score, trend line, or anomaly looks persuasive and you need to know what it can actually support. The guide turns statistical methods into concrete questions: what changed, what broke, what leaked, what will be new in use, and what decision the evidence is allowed to move.
Editorial on-rampThe Question Before the NumberMost statistical failures begin before the calculation.- How to read diagnostics
- Foundations for unfamiliar data
- Time, cadence, and sampling pressure
- Model changing vectors, variance, states, regimes, and events
- Censoring, survival, and remaining-life decisions
- Data leakage before it becomes confidence
- Build and evaluate a model
- Why the model includes the evaluation protocol