IdeaNavigator AI Daily opportunity briefs

Verdict Calibration

Is the bar real, or just strict?

The validation rubric rarely returns a "Build" verdict. This page makes the threshold a measured choice, not a guess: the full score distribution, the exact criterion that blocks each near-miss, and a graph-grounded thesis confidence that credits ideas with real evidence behind them.

159reports scored
0Build verdicts
20near-misses (65–79)
12evidence-lifted theses

Score distribution

Rethink (<50)9
Research (50–64)65
Validate (65–79)85
Build (80+)0

A real bar, not a hype machine: most reports land in Validate, and Build is reserved for ≥80 with no weak criterion.

What blocks Build

Among near-misses, the criterion most often scoring below 7/10:

Feasibility67
Demand signal18

Closest near-misses

One criterion from Build.

Reports scoring Validate-or-better, ranked by how little stands between them and a Build verdict.

Thesis confidence

Where real evidence lifts the score.

A graph-grounded signal: the raw validation score plus bounded credit for declared complaint evidence, a trend seed, and cross-vertical keyword demand. Monotonic and capped at 100 — evidence can only help.