Force System 2 thinking. Track your forecast calibration over time.
Most bad decisions aren't made — they're slid into. This app adds deliberate friction to consequential decisions: before you commit, it makes you state a measurable claim, assign a probability, name your assumptions, and set a resolution date. Then it scores you against reality.
Over time, the analytics view answers the question almost nobody can answer about themselves: when you say you're 80% sure, how often are you actually right?
Forecast calibration is one of the most learnable and least practiced skills in decision-making. The research (Tetlock's superforecasting work, Kahneman's System 1/System 2 framing) is clear that people improve when they make explicit probabilistic claims and confront the outcomes. Almost no one has a tool that makes that loop a habit. This is that tool — deliberately simple, no account, no backend.
- New Entry — structure a decision as a measurable claim with a probability, confidence rationale, and resolution date
- My Decisions — open ledger of pending and resolved claims
- Analytics — calibration tracking: stated confidence vs. realized outcomes
- Single-page vanilla JavaScript app, no build step, data stored locally
Open index.html in a browser. That's it.
Vanilla HTML/CSS/JS, built with AI coding tools as part of a broader exploration of decision quality and judgment under uncertainty.