The Problem ACH Solves

The Threat to Your Thinking

Before you can understand what ACH does, you have to understand what it is designed to fix. The problem is not laziness. It is not carelessness. The problem is the normal operation of a human brain under conditions of ambiguous information, which is exactly the condition that every threat assessment, every vetting decision, and every MAG security evaluation operates under. Your brain is not malfunctioning when it produces bad analysis. It is doing exactly what evolution built it to do, and that is precisely the problem.

The core issue is this: human beings do not process information neutrally. We process it through a set of cognitive shortcuts called heuristics, which under normal social conditions save enormous time and produce good-enough results. The problem arises when those same shortcuts are applied to analytical problems, where good-enough is the threshold at which bad decisions get made and surprises occur. In intelligence work, surprise is the metric of failure.

Confirmation bias is the most operationally dangerous of these heuristics. Once you form an initial impression of a situation — a person seems trustworthy, a threat seems unlikely, a scenario seems implausible — your brain begins filtering incoming information to fit that impression. Evidence that confirms the initial impression gets processed and retained. Evidence that contradicts it gets minimized, reinterpreted, or discarded. This process is largely unconscious. You do not experience it as bias. You experience it as reasonableness. The person who disagrees with your assessment seems to you to be the one who is not thinking clearly.

The intelligence community has documented the consequences of confirmation bias repeatedly and at cost. The failure to anticipate the 1973 Yom Kippur War, the misreading of Iraqi weapons programs, numerous tactical surprise events — in each case, substantial contradictory evidence existed in the record. Analysts had access to it. It did not change the assessment because the assessment was already formed and the incoming evidence was being filtered through it rather than testing it.

The Other Biases That Compound the Problem

A Catalogue of Analytical Failure Modes

Confirmation bias is the dominant threat, but it operates alongside several others that compound its effects. A serious student of analytical tradecraft needs to be able to name these, recognize them in their own reasoning, and understand how ACH addresses each of them.

Anchoring is the tendency to place disproportionate weight on the first piece of information received about a situation. Once that anchor is set, subsequent information is evaluated relative to it rather than on its own terms. In a vetting context, a strong first impression from an early referral contact functions as an anchor. Everything the candidate does afterward gets measured against that impression rather than evaluated independently.

Availability bias causes analysts to overweight information that is recent, vivid, or emotionally salient. A threat that happened in a neighboring community last month is treated as more probable than a threat with a higher base rate that has not produced a recent local example. The dramatic incident feels more analytically important than the quiet pattern.

Mirror imaging is the assumption that adversaries, unknown actors, or ambiguous persons think, value, and behave the way you do. An analyst who would never attend a community event for the purpose of intelligence collection assumes that most people at such events are there for legitimate community reasons. That assumption is the operational opening that a skilled collector exploits.

Groupthink is the suppression of dissenting analysis within a team or organization in favor of consensus. When everyone at the table agrees on an assessment, it feels like the assessment is robust. It is often instead a sign that the dissenting interpretation has not been heard or has been socially discouraged. ACH addresses groupthink directly because its structured process forces every hypothesis, including the dissenting one, to be evaluated against the full evidence set.

What ACH Does Differently

The Method in Brief

ACH was developed by Richards J. Heuer Jr., a CIA analyst who spent decades studying how intelligence analysts actually think and where that thinking goes wrong. His foundational text, Psychology of Intelligence Analysis, is the source document behind this curriculum. Heuer built ACH on a single insight that inverts the normal analytical procedure: instead of asking which explanation has the most evidence supporting it, ask which explanation has the least evidence against it.

This inversion matters because of a basic logical asymmetry. Finding evidence that is consistent with a hypothesis does not prove it. Many hypotheses can be consistent with the same evidence. But finding evidence that is genuinely inconsistent with a hypothesis — evidence that cannot plausibly be true if the hypothesis is true — does something real. It weakens or eliminates that hypothesis. ACH is built around that asymmetry. It is a machine for finding and weighing contradictions rather than confirmations.

The method produces a matrix: hypotheses across the top, evidence down the side, and a structured assessment in each cell of whether that evidence is consistent, inconsistent, or not applicable to each hypothesis. The analyst reads across the rows, one evidence item at a time, assessing how each item fits every hypothesis simultaneously. The hypotheses that accumulate the fewest serious inconsistencies are the most credible. The hypotheses that are heavily contradicted are eliminated or downgraded. The conclusion that survives is the one that has been most aggressively tested, not the one that felt most plausible at the start.

The full eight-step process, the principles behind it, and a complete applied scenario follow in Lessons 1 through 3. This lesson’s job is only to establish the problem clearly enough that the solution makes sense when you reach it.

↑ Back to Intelligence Analysis