Calibration

Are your decisions actually getting better?

Most people never find out. Calibration measures your forecasts against what really happened — a Brier score, the same instrument used to grade professional forecasters.

Brier Score
0.19
LOWER IS BETTER · 0.00 PERFECT · 0.25 COIN FLIP
14
TOTAL
9
RESOLVED
71%
ACCURACY
The difference, shown — not told

No scorecard vs. an honest one.

The follow-up that never comes

"I was 80% sure that hire would work out. Was I right to be that confident?"

A normal AI assistant

Reflecting on your past predictions is a great habit. Consider keeping a decision journal: write down what you expect and why, then review it later. Over time you’ll get a sense of how calibrated your confidence is.

It’s wise to review your judgment — self-awareness is the first step to improving it.

◉ DAUDIT · Calibration
◈ Your actual number

Across 9 resolved decisions, your Brier score is 0.19 — and improving. When you said 80%, you were right 74% of the time.

⬡ Where it slips

Your 90% bucket is well-calibrated. Your 70% bucket isn’t — that’s where overconfidence lives for you.

◉ The honest answer

Yes, your decisions are getting better. Here is the number that proves it.

A normal assistant suggests you reflect. Daudit keeps the actual scorecard — and it can’t be argued with.

What it does

The honest answer, not the comfortable one.

Confidence feels like competence, but they are not the same thing, and only one of them can be measured. Calibration measures it. Each time you record what you expected and DAUDIT later captures what actually happened, your Brier score updates — a single number where lower is better, 0.25 is a coin flip, and the truth is wherever you actually land.

It also breaks down by confidence bucket: when you said you were 90% sure, were you right 90% of the time? For this persona — trained to be loss-averse and self-aware — the calibration score is the most quietly addictive thing DAUDIT offers, because it is the one number that cannot be argued with.

How it connects

It depends on closing the loop.

Calibration only works because the Ledger remembers to ask how each decision turned out — and it feeds the Genome’s read of your confidence patterns.

Find out if you’re actually improving.

AUDIT A DECISION →