Confidence Tiers: How We Label What We Know

Confidence Tiers: How We Label What We Know

CommandEleven Intelligence Desk // Tradecraft Desk, Tradecraft Primer. Part of the series How CommandEleven Intelligence Thinks.

Bottom Line Up Front

A confidence tier is a label on a claim that tells the reader how much weight it can bear. Every claim in a CommandEleven Intelligence dossier is meant to carry one: CONFIRMED, ASSESSED, REPORTED, ALLEGED or SPECULATIVE, with DATA DEFICIT used as a notation when the information does not exist or cannot be reached. The labels separate what we can document from what we judge and from what someone else has claimed.

The tier is not decoration. It is the reader’s way to check us. If a statement reads as certain, the label should show that it is. If it rests on one report or one accusation, the label should say so, and no graphic, headline or summary is allowed to sound more certain than the text it rests on.

Why Label Confidence at All

An assessment that does not show its confidence forces the reader to guess. Two sentences can sound the same and rest on very different ground: one on a document anyone can read, the other on a single unverified account. US intelligence standards deal with this directly. Intelligence Community Directive 203 asks analysts to separate underlying information from their own assumptions and judgments, to describe the quality and credibility of their sources, and to explain how uncertain their major judgments are and why.

We follow the same logic in public. The tiers keep three things apart that are easy to blur: what the evidence shows, what we conclude from it, and what others say.

The Tiers

Each label has a specific meaning, and the meaning stays the same in every dossier.

Ladder of CommandEleven Intelligence confidence tiers: CONFIRMED, ASSESSED, REPORTED, ALLEGED and SPECULATIVE, each with its meaning and whose voice it carries, plus the DATA DEFICIT notation

CONFIRMED

Supported by multiple independent sources or direct corroboration. A confirmed claim has been checked, typically against a primary document or against sources that do not depend on one another. It is stated as fact, with its source. Reporting from a major established news organization, such as CNN, the New York Times or the Wall Street Journal, is labeled CONFIRMED on that outlet’s credibility, not ours, and we name the outlet.

ASSESSED

A reasoned judgment based on available evidence, with remaining uncertainty. This is our own analysis. It is marked as judgment so that nobody mistakes it for a documented fact, and it is paired with what would change our mind.

REPORTED

A claim that rests on a single outside report that we have not corroborated, from a source without that standing. We say who reported it and we do not vouch for it. A reported claim can still matter, but the reader should know that it stands on one leg.

ALLEGED

A claim made by a party to a dispute or a legal case, such as prosecutors, a government or a group, about someone else’s conduct. We present it as that party’s claim and attribute it, and we do not treat an accusation as an established fact. Where a group denies an allegation, we note the denial.

SPECULATIVE

Plausible but thinly supported, offered for context only. A speculative claim is never folded into the key judgments and never carries weight in the conclusion.

DATA DEFICIT

DATA DEFICIT is not a confidence level. It is a notation that the information we would need does not exist or is not available. Saying so openly is more useful than filling the gap with a guess.

How a Rating Becomes a Tier

The tier is the end of a chain. Before a claim is promoted into analysis, the Vetting Matrix grades the source for reliability (A to F) and the information for credibility (1 to 6), and the claim must pass Triple-Node Verification. Single-source publication is prohibited, so a claim that rests on one report or one accusation does not stand alone. If the source is a major established news organization, we label the claim CONFIRMED on that outlet’s credibility and name it. Otherwise it carries the REPORTED or ALLEGED label, with its source named.

Put simply, the ranks grade what comes in, and the tier is the label a reader sees on the finished claim.

Tier and Likelihood Are Different Things

Two ideas are often run together. Confidence is how much weight the evidence and reasoning can bear. Likelihood is how probable an event is. A forecast can be highly likely and rest on thin evidence, or unlikely and rest on solid evidence. ICD 203 notes that confidence depends on the logic and the quantity and quality of the sources, and it tells analysts not to combine a stated confidence level and a degree of likelihood in the same sentence. We keep them apart for the same reason.

For likelihood, the standard reference is the scale in ICD 203, which attaches words such as “unlikely,” “likely” and “very likely” to probability ranges.

Scale pairing likelihood words from almost no chance to almost certain with the probability ranges defined in Intelligence Community Directive 203

Our dossiers use plain wording in the same spirit, such as “most likely” and “could,” and the dated calls we register on the Foresight Record carry no percentages. The point of a call is the date and the outcome, which we then score against what happened.

Rules We Hold Ourselves To

  • A single source without established credibility is never CONFIRMED. One report or one accusation from such a source is labeled REPORTED or ALLEGED, with its source named. A major established outlet is CONFIRMED on its own credibility, and we name it.
  • The label matches the text. A graphic, headline or summary is never more certain than the paragraph it illustrates.
  • Judgment is marked as judgment. Anything that is our analysis rather than a documented fact carries ASSESSED.
  • Speculation stays out of the conclusions. SPECULATIVE material is context only and never appears in the key judgments.
  • Tiers change when evidence changes. If events contradict an assessment, the tier changes and the correction is shown in a Revision Annex, not buried.
  • Gaps are stated. Where the information does not exist, we write DATA DEFICIT instead of guessing.

How to Read the Labels in a Dossier

Every dossier ends with a short legend, “How to Read Our Labels,” that defines each label. In the key judgments, a CONFIRMED claim comes with its source, an ASSESSED claim comes with its reasoning and what would change it, and a REPORTED or ALLEGED claim names who said it. For a worked example, see Al Qaeda Aviation Threat: What the FBI-DHS-TSA Warning Means for US Air Travel.

Frequently Asked Questions

What is a confidence tier?

A label on a claim that tells the reader how much weight it can bear. Ours are CONFIRMED, ASSESSED, REPORTED, ALLEGED and SPECULATIVE.

What is the difference between REPORTED and ALLEGED?

REPORTED is a claim that rests on a single outside report we have not corroborated, from a source without established standing. ALLEGED is a claim made by a party, such as a prosecutor or a government, about someone else’s conduct. Both are attributed to their source and neither is treated as established.

Is ASSESSED the same as probable?

No. ASSESSED marks a judgment that is ours, based on the evidence, with uncertainty remaining. It tells you the claim is analysis, not a documented fact. Likelihood is a separate question.

Why does a claim sometimes get no tier at all?

It should not. Where the information does not exist, we say so with the notation DATA DEFICIT, and where it is thin we use SPECULATIVE.

Do you give probabilities?

Not on dated calls. We use plain wording and register each call on the Foresight Record, where it is scored against what happened.

What should I read next?

Structured Analysis: Testing Our Own Thinking explains the techniques behind our ASSESSED judgments.

Sources

How to Read Our Labels

Every claim in a CommandEleven Intelligence dossier carries one of these.

CONFIRMED

Multiple independent sources, direct corroboration, or reporting from a major established outlet. Stated as fact, with its source.

ASSESSED

Our reasoned judgment from the available evidence. Uncertainty remains.

REPORTED

One outside report we have not corroborated, from a source without established standing. Attributed, not vouched for.

ALLEGED

A party's claim about someone else's conduct. Attributed, not treated as established.

SPECULATIVE

Plausible but thinly supported. Context only, and never weighs on a conclusion.

DATA DEFICIT

Notation: the information does not exist or is out of reach. Stated, not guessed.

Read the full primer: Confidence Tiers, How We Label What We Know

Related dossiers

More from CommandEleven Intelligence

Opens in a new tab
Vol. I No. 1 is out
The monthly magazine

Convergence

Our dossiers, drawn together each month into one edition.

Get the first edition →
Publishing mid-December
The annual report

2027 Business Threat Exposure

Ten threats to American business, ranked by exposure.

Get notified →
The scorecard

Foresight Record

Every call we have made, scored in public against what happened.

Open →
The audit

GCTA Audit

The April 2026 forecasts, checked section by section.

Open →
The panel

Survival Dispatch News

Our regular panel appearances, with the full conversation.

Open →
The weekly show

Soft Targets America

Weekly on Tuesdays, with co-host Jason Piechura.

Open →
For organizations
Our core service

Intelligence as a Service

Private intelligence, delivered to your team.

Engage us →