What this rubric has actually said
35 ideas have been underwritten and decided on this path. 5 carry a BUILD verdict — 14%. 12 were BLOCKED before a score was ever computed. A rubric that mostly says no is the claim; publishing the distribution is how you check it.
Every verdict this rubric has issued
| Verdict | Count | Share | Mean score | Range |
|---|---|---|---|---|
| WATCH | 10 | 29% | 35 | 15–48 |
| KILL | 6 | 17% | assessed, no composite | — |
| BLOCKED-NOMONEY | 5 | 14% | assessed, no composite | — |
| BUILD | 5 | 14% | 65 | 61–74 |
| BLOCKED-MANUAL | 3 | 9% | assessed, no composite | — |
| BLOCKED-LEGAL | 2 | 6% | assessed, no composite | — |
| BLOCKED-NOMOAT | 1 | 3% | assessed, no composite | — |
| BLOCKED-DARK | 1 | 3% | assessed, no composite | — |
| REWORK | 1 | 3% | 38 | 38–38 |
| VALIDATE | 1 | 3% | 59 | 59–59 |
KILL AND BLOCKED CARRY NO COMPOSITE SCORE, AND THAT IS NOT WHAT I FIRST WROTE HERE. The first version of this note said they were blocked at a gate before scoring — checked, and it is false. All six KILL and all eleven BLOCKED candidates carry a full set of dimension scores AND an arbitration ruling. They were fully assessed and then refused on a specific finding, which is a stronger claim than declining to look. What they lack is a COMPOSITE, because a single number averages away the one dimension that killed it — a business with no route to money is not a 40, it is a no.
The highest-scoring ideas, and what they scored
| Idea | Verdict | Score | Confidence |
|---|---|---|---|
| Pre-LOI Screening for SBA Business Buyers | BUILD | 74 | 76 |
| Unbilled Change-Order Watchdog for Subcontractor | BUILD | 67 | 55 |
| Calibrated Base-Rate Data for Business Buyers | BUILD | 62 | 73 |
| GA4 Symptom-to-Root-Cause Diagnostic | BUILD | 62 | 76 |
| Publicly Scored Daily Volatility Ranges for Trad | BUILD | 61 | 83 |
| Deal Intelligence for Used Heavy-Duty Diesel Tru | VALIDATE | 59 | 70 |
| On-Site Mobile Diesel Repair for West Texas Owne | WATCH | 48 | 22 |
| Supplier Recall Lookup for Injectable Kit Assemb | WATCH | 42 | 76 |
Each links to its full record — the charges against it, what the defence conceded, and the test frozen before the outcome was known.
Put an idea through it
| Do this | What it gives you |
|---|---|
| Analyze an idea → | Describe it in a sentence. The free industry record comes back instantly; an underwriting against this rubric is the metered step. |
| Read a full underwriting → | Every decided candidate, with the charges against it and what the defence conceded. |
| The ledger → | Predictions frozen before their outcomes were known, and how they have resolved. |
A distribution with no way to be judged by it is a statistic. This rubric is the thing being offered, so the page ends with the door to it.
What it cost to produce this record
564 agent stages, $71.68 all time, across 35 decided candidates.
That figure is published because a verdict distribution with no cost beside it invites the assumption that assessments are cheap and therefore casual. These are not: a full run is 22 to 26 stages and roughly forty minutes.