Analysis Competing Hypotheses

← Back to Analytical Techniques

Or: How to Prove Yourself Wrong in the Most Structured Way Possible

What is ACH? (Besides a Sound You Make When Realizing You've Been Wrong All Along)

Analysis of Competing Hypotheses (ACH) is a structured analytical technique that helps intelligence analysts avoid the cognitive trap of falling in love with their first hypothesis - a relationship that, like many first loves, is often doomed to heartbreak and embarrassment.

Developed by Richards Heuer at the CIA, ACH is essentially a methodical way of playing devil's advocate with yourself before someone else does it for you in front of your entire team.

The ACH Analyst's Motto

"I'm not just here to prove myself right; I'm here to prove myself wrong in the most elaborate way possible."

The ACH Process: A Step-by-Step Guide to Structured Self-Doubt

• Identify possible hypotheses - Come up with all the possible explanations, including the ones that make you think, "Well, that's ridiculous, but I guess I should include it."

• List evidence and arguments - Gather all the evidence, including that one piece you've been conveniently ignoring because it doesn't fit your favorite theory.

• Create a matrix - Build a spreadsheet that will haunt your dreams, with hypotheses across the top and evidence down the side.

• Analyze how evidence relates to each hypothesis - For each piece of evidence, ask: "Does this prove my hypothesis wrong?" (Spoiler alert: it often does.)

• Draw tentative conclusions - Figure out which hypothesis has survived your brutal assault of evidence. It's rarely the one you started with, much to your ego's dismay.

• Analyze sensitivity - Ask yourself: "What if that important piece of evidence turns out to be from a source who also believes the Earth is flat?"

• Report conclusions - Explain your findings in a way that doesn't reveal how attached you were to your original, now-disproven theory.

Why ACH Works (When You Actually Follow It)

ACH forces analysts to consider alternative explanations rather than just gathering evidence to support their initial hunch - a cognitive bias known as "confirmation bias" or, in technical terms, "only seeing what you want to see."

By focusing on disproving hypotheses rather than proving them, ACH helps analysts avoid the trap of becoming emotionally attached to their favorite theory - a relationship that should have ended several pieces of contradictory evidence ago.

ACH Reality Check

Remember: In the intelligence world, being wrong loudly and confidently isn't just embarrassing - it can lead to real-world consequences. ACH helps you be wrong in private first, so you can be right when it counts.

Common ACH Pitfalls (That You'll Probably Fall Into Anyway)

• The "I'll Just Do This In My Head" Trap - No, you won't. Your brain is lazy and will take shortcuts. Use the matrix.

• The "My Favorite Hypothesis Is Still Right" Delusion - If your original hypothesis survives intact after ACH, you probably did it wrong.

• The "I Don't Have Time For This" Excuse - You also don't have time to explain to your boss why you missed a critical alternative explanation.

• The "I've Added Too Many Hypotheses" Spiral - Yes, alien intervention is technically a hypothesis, but let's keep it reasonable.

ACH Resources

-

ACH Matrix Template (Excel Format, Tears Not Included)

-

Heuer's "Psychology of Intelligence Analysis" (The ACH Bible)

-

ACH Case Studies: When Being Methodically Wrong Saved the Day

Related Analytical Techniques

-

Structured Brainstorming: Organized Chaos for Better Ideas

-

Assumptions Check: Questioning Everything You Hold Dear

-

Indicators Development: Crystal Ball Not Required

What is Analysis of Competing Hypotheses?

Analysis of Competing Hypotheses (ACH) is one of the most widely taught structured analytical techniques in the intelligence community. Developed by Richards Heuer, ACH addresses a fundamental cognitive challenge: analysts naturally gravitate toward a single explanation and interpret subsequent evidence to support it. ACH forces the systematic consideration of multiple explanations before evidence is evaluated.

• The Core Insight: ACH inverts the natural analytical process. Instead of asking "what does the evidence tell me?", ACH asks "if this hypothesis were true, what evidence would I expect to see?" This reframing reduces confirmation bias.
• Not a Prediction Tool: ACH is not designed to predict the future. It is designed to evaluate which explanation best fits the available evidence, given the inherent uncertainties of intelligence analysis.
• When to Use ACH: ACH is most valuable for analytical problems where there are multiple plausible explanations, where the stakes of getting it wrong are high, and where the analyst has sufficient evidence to discriminate between hypotheses.

The ACH Process

ACH follows a structured eight-step process:

1. Identify the Possible Hypotheses: Generate a set of plausible explanations for the phenomenon under investigation. Include the null hypothesis - that the observed situation is the result of normal, innocent activity. Analysts often stop at two or three hypotheses; the technique works best when at least four or five are considered.
2. List the Evidence and Assumptions: Catalogue all relevant evidence, including the absence of expected evidence. Distinguish between confirmed facts, reported information, and analytical assumptions.
3. Build the Matrix: Create a matrix with hypotheses as columns and evidence as rows. For each cell, assess whether the evidence is consistent (+), inconsistent (-), or irrelevant (N/A) to the hypothesis.
4. Refine the Matrix: Re-examine each entry. Evidence that appears consistent with all hypotheses may be too general to be diagnostic. Evidence that is inconsistent with a hypothesis is analytically valuable and should be scrutinised for accuracy.
5. Draw Tentative Conclusions: The hypothesis with the fewest and least damaging inconsistencies is tentatively the strongest. However, also consider the hypothesis that would be most damaging if wrong - sensitivity to being wrong is an important analytical consideration.
6. Consider the Sensitivity of Conclusions: Test how robust the conclusion is to changes in key assumptions. If a single piece of evidence shifts the assessment, the conclusion is fragile and should be caveated accordingly.
7. Identify Diagnostic Evidence: Determine what new evidence would change the assessment. This focuses future collection efforts and gives consumers a basis for monitoring the situation.
8. Document the Reasoning: The ACH matrix itself serves as documentation of the analytical process, allowing reviewers and consumers to understand how the conclusion was reached.

Common ACH Mistakes

Even experienced analysts make predictable errors when applying ACH:

• Including the Hypothesis You Want to Be True: Analysts may unconsciously include a favoured hypothesis and then treat the matrix as confirmation. The technique works best when hypotheses are generated before evidence is reviewed.
• Neglecting the Null Hypothesis: The "nothing unusual is happening" hypothesis is often dismissed too quickly. Strategic surprises often occur because analysts rejected the null hypothesis based on ambiguous evidence.
• Evidence that Fits Everything: Generic evidence - "the country has a history of regional tensions" - is consistent with many hypotheses and is not diagnostically useful. ACH works best when evidence is specific and discriminating.
• Treating ACH Output as Truth: ACH is a thinking tool, not a truth machine. The matrix reflects the analyst's current understanding and is limited by the quality and completeness of available evidence.