PACHT | PLESSAS Analysis of Competing Hypotheses Toolkit
Plessas Experts Network · Research & Analysis

PLESSAS Analysis of Competing Hypotheses Toolkit

PACHT · A structured, private, analyst-directed intelligence analysis workspace

The Analyst Community

The Analyst's Annex

Support the development of analytical tools like PACHT. Share ideas, provide feedback, suggest improvements, and participate in future tool builds with Plessas Experts Network.

Visit the Analyst's Annex ↗
Important: PACHT is a methodology, not an intelligence analyst.

The quality of your analysis depends on the quality of the information you enter. Do not let AI think for you. PACHT does not verify evidence, generate conclusions, or replace independent research and judgment.

Read complete methodological disclaimer ↓

This tool is intended to guide users through the Analysis of Competing Hypotheses (ACH) methodology. Users are responsible for providing at least two mutually exclusive hypotheses and relevant evidence.

PACHT does not supply or validate intelligence information. It does not determine truth, assign probability, or guarantee analytical accuracy. A hypothesis with fewer recorded inconsistencies is not necessarily true.

Confirmation bias, incomplete evidence, unreliable reporting, hidden assumptions, and poorly framed hypotheses can affect results. Analysts should challenge each hypothesis and review all evidence independently.

This standalone tool does not run an AI model. Ratings and assessments are entered by the user. Any external AI-generated analysis must be reviewed critically and should never replace the analyst's own reasoning.

Responsibility for interpreting the matrix and acting upon its contents remains entirely with the user.

USER GUIDE · Click for Instructions ↓
  1. Frame the issue. Define the analytical problem without asking PACHT for a narrative answer.
  2. Develop hypotheses. Enter at least two mutually exclusive and testable hypotheses.
  3. Confirm exclusivity. Check that your hypotheses do not overlap or express the same explanation in different words.
  4. Collect evidence. Add observations, reporting, documents, or other independently gathered information. Include source and reliability notes.
  5. Construct the ACH matrix. Choose manual or guided assessment.
  6. Rate consistency. Use C for consistent, I for inconsistent, N for neutral, or leave unassessed.
  7. Review your findings. Focus on contradictions, diagnostic evidence, assumptions, source reliability, and gaps.
  8. Export. Save the project locally as JSON or export an Excel-readable workbook.

ACH encourages analysts to test competing explanations against the same evidence. Disconfirming evidence is particularly important because confirming observations may support multiple hypotheses.

Methodological reference: CIA — A Tradecraft Primer: Structured Analytic Techniques

01 / Define the Issue

Frame the analytical problem

Define the uncertainty that will be examined through competing hypotheses.

View PACHT operational protocol and ACH scope
ROLE Guide users through the Analysis of Competing Hypotheses methodology. SCOPE Do not generate independent intelligence judgments, answers, factual summaries, speculation, or narrative conclusions. HYPOTHESES Require at least two user-defined, mutually exclusive hypotheses. Warn about potential redundancy or overlap. The analyst must confirm logical exclusivity. EVIDENCE Use only the evidence supplied by the analyst. Do not generate, authenticate, or independently interpret evidence. MATRIX Construct a blank ACH matrix using user-provided hypotheses and evidence. Allow the analyst to choose manual or guided assessment. Do not assign consistency ratings automatically. GUIDED ASSESSMENT Present each evidence item in full. Display all hypotheses by H number. Invite the analyst to mark each pairing as consistent, inconsistent, neutral, or unassessed. Continue through all evidence items. ANALYTICAL RIGOR Encourage consideration of confirmation bias, missing evidence, assumptions, competing explanations, and source limitations. AI-GENERATED RATINGS Not available in this local edition. If ever implemented, use the disclaimer: "AI does not think or reason. The following ratings are not conclusions and may be inaccurate. They reflect only patterns present in the training data and may contain gaps, oversights, or biases. Please review all results critically." Any AI-generated matrix analysis must begin: "AI does not think or reason. The following analysis may not be credible and should be reviewed carefully by the user." All interpretations and conclusions remain the user's responsibility. PRIVACY No remote AI model, tracking service, online storage, or server-side processing is used in this edition.
PLESSAS EXPERTS NETWORK · TEST TOOL

This is an experimental analytical tool developed by Plessas Experts Network. Plessas Experts Network does not claim invention of Analysis of Competing Hypotheses. PACHT adapts an established intelligence analysis methodology into a structured, analyst-controlled workspace.

Local-first privacy: Your case information, hypotheses, evidence, and assessments are processed in your browser. They are not transmitted to a PACHT server. Saved projects and exported workbooks remain under your control.

Support tool development through the Analyst's Annex ↗

ACH methodological reference · CIA Tradecraft Primer ↗

PACHT · Standalone Edition 2.1 · Plessas Experts Network