JSA's Monitor

The NCI Engineered Reality Score

A transparent, structured rubric for measuring signals of coordinated media manipulation — indicator by indicator, in the open.

What this is — and what it is not. The NCI scale measures indicators of coordinated manipulation. A high score means many manipulation techniques are present in how a story is being told across outlets. It does not, by itself, prove that a deliberate influence campaign exists, that any claim in the story is false, or that any specific actor is responsible. It is a structured prompt for skepticism, not a verdict.

Every story that appears across two or more outlets is scored against 20 indicators, each rated 1 (not present) to 5 (strongly present). The raw total (20–100) is normalized to a 0–100 scale and mapped to five bands.

About 14 of the 20 indicators are computed automatically from the live coverage itself — publication timing, phrase synchronization across outlets, and direction-neutral language lexicons. The rest require world knowledge (who benefits, what context is missing, whether the statistics are cherry-picked, historical parallels) and are scored on demand by a local AI model or by you directly. Any indicator can be overridden with your own judgment; the scale was designed for a human in the loop.

The five bands

0–20Low indication of coordinated manipulation
21–40Some persuasive or propaganda techniques present
41–60Significant manipulation indicators
61–80Strong likelihood of an organized influence campaign
81–100Extreme, highly coordinated engineered-reality environment

Most ordinary news scores low. Higher bands are reserved for coverage showing many, specific, citable manipulation techniques at once.

The 20 indicators

Each is scored 1–5. "Auto-scored" indicators are computed from the live coverage; "AI / manual" indicators need world knowledge and are scored on demand.

1
Suspicious timing auto-scored
Story bursts across many outlets nearly simultaneously, or lands at a strategically convenient moment.
2
Emotional manipulation auto-scored
Fear, anger, or disgust language doing the persuading instead of facts.
3
Uniform / scripted messaging auto-scored
Identical or near-identical phrasing across nominally independent outlets.
4
Important missing information AI / manual
Key context absent everywhere; includes moving goalposts and contradictory official messaging over time.
5
Overly simplistic narratives AI / manual
Complex events reduced to a single clean cause or villain.
6
Tribal / us-vs-them framing auto-scored
Language dividing audiences into in-group and out-group; includes public shaming and social coercion.
7
Authority overload auto-scored
Stacked appeals to experts/officials/agencies substituting for evidence.
8
Pressure for urgent action auto-scored
Act-now framing that punishes deliberation.
9
Novelty / unprecedented danger auto-scored
Relentless emphasis on newness or historic uniqueness of the threat.
10
Financial / political beneficiaries AI / manual
Identifiable parties who gain; includes conflicts of interest and pharmaceutical, corporate, or government payments.
11
Suppression of dissent auto-scored
Dissenters banned, punished, or delegitimized; includes censorship and algorithmic suppression.
12
False dilemmas auto-scored
Only two options presented when more exist.
13
Bandwagon / social proof auto-scored
"Everyone agrees" pressure; consensus asserted rather than demonstrated.
14
Repeated emotional messaging auto-scored
The same emotive terms recycled across outlets and days.
15
Cherry-picked statistics AI / manual
Selective numbers, denominators hidden, baselines omitted.
16
Logical fallacies AI / manual
Strawmen, ad hominem, appeal to consequences, motte-and-bailey.
17
Manufactured outrage auto-scored
Outrage framing ("slams", "fury", "backlash") disproportionate to events.
18
Framing & language manipulation auto-scored
Loaded labels and euphemisms steering conclusions before facts arrive.
19
Rapid behavior-change push auto-scored
Messaging engineered to alter public behavior quickly.
20
Historical propaganda parallels AI / manual
Techniques matching documented influence campaigns.

How the automatic scoring works

Synchronized phrasing is the core signal. When multiple outlets use the same distinctive multi-word phrase, that is either wire-service copy (normal, and labeled as such when a wire agency is among them) or coordinated messaging (flagged when no wire source is behind it). A phrase repeated across many outlets, or pushed for days, weighs heavily.

Timing looks at how tightly a story bursts across outlets. Emotional and framing language uses direction-neutral lexicons — the same words are flagged whoever uses them, so the score does not take a political side. Coverage asymmetry notes when independent outlets cover something mainstream ignores, or when wall-to-wall coverage has no independent corroboration.

Direction-neutral by design. The language lexicons flag emotive and delegitimizing terms regardless of target. A story is not scored higher for criticizing one side or another — only for how it is told.

Use it as a starting point

A high NCI score is an invitation to look closer, not a conclusion. Read the source columns side by side, check what the local press says, follow the unique claims to their origin, and decide for yourself. The tool's job is to surface the pattern quickly; the judgment stays yours.