HOW DO
WE KNOW?
Between what happened and what we believe happened lives an entire universe of uncertainty.
Come investigate it. No intellectual superiority. No guaranteed truth machine. Just better questions, better evidence and the occasional humbling surprise.
A good explanation survives serious alternatives. It does not merely sound convincing.
/ 001 — THE PRINCIPLES
Eight ways to be less wrong.
Each idea contains a small example, a competing explanation and a question you can use in actual life. Open any card to explore further.
/ 002 — THE EXPERIMENTS
Less preaching.
More playing with ideas.
Three small interactive models. The numbers and cases are explicitly invented for learning, not presented as research findings.
CLASSIFICATION / 12 THOUGHT EXPERIMENTS
What kind of claim is this?
Choose the best description (the case context matters)
The goal is to notice distinctions, not collect a badge.
CAUSALITY / FICTIONAL CITY
Two trends. One hidden cause.
Move the temperature and watch two measures rise together. In this toy world, neither variable causes the other; temperature drives both.
No arrow goes from sales to visits. Their observed association arises from a common cause in this model.
BAYES / FICTIONAL QUALITY CONTROL
When a positive result misleads.
Imagine a factory sensor that flags potentially defective items. Set the prevalence and accuracy rates. The calculated chance of a true defect among flagged items may surprise you.
/ 003 — THE PRIVATE WORKBENCH
Bring your
own question.
Use this five-part inquiry worksheet on an everyday claim, a study, a headline or a decision. Your writing stays in your browser unless you choose to copy or download it.
No accounts, submissions, trackers or server storage. “Save locally” uses this device’s browser storage only.
WE KNOW?
THE EVIDENCE AUDIT FIELD GUIDE
Protocols / Cases / WorksheetsCritical inquiry for ordinary life.
/ 004 — SOMETHING TO KEEP
A field guide for
the real world.
This isn’t a trophy PDF or a one-page advertisement. It is a substantial, printable companion to use when a confident claim deserves a second look.
Free to read and keep. Credited to Sahil Tayal · sahiltayal.com. PDF is a real file included with this website, not generated by a third-party service.
/ 005 — OPEN INQUIRY
What would
change your mind?
Found a weak argument here? Have a better experiment or an interesting disagreement? Serious inquiry gets better when people can talk back.