FOCUS

Forecasting Counterfactuals in Uncontrolled Settings
FOCUS

Intelligence Value

FOCUS is enhancing analysis tradecraft by improving analysts’ ability to learn lessons from the past. FOCUS is helping analysts correctly answer the question, “how would history have been different if actor X had instead done action Y on date Z?”

Summary

Counterfactual forecasts -- statements about what would have happened if different circumstances had occurred -- are the foundation for lessons about what to do in the future. For example, analysts performing a postmortem review of an intelligence failure might conclude the Intelligence Community could have avoided the failure if it had employed different methods based on prior experience and best practices. Unfortunately, research across multiple disciplines has concluded that experience-based best practices are generally ineffective. There has been no systematic prior research on improving counterfactual reasoning or lessons learned. The FOCUS program aims to develop practical, accurate, evidence-based techniques for counterfactual forecasting.

Because we can’t re-run history in the real world, FOCUS methods are evaluated by having hundreds of people make predictions about outcomes in complex simulations.

Proposers' Day Briefings

Related Publications

To access FOCUS program-related publications, please visit Google Scholar.


Contact Information

Program Manager

Main Office


Program Manager

Dr. Steven Rieber

Contracting Office Address

Office of the Director of National Intelligence

Intelligence Advanced Research Projects Activity

Washington, DC 20511

Research Area(s)

Analytic tradecraft, Cognitive biases, Counterfactual reasoning, Evidence-based approaches to lessons-learned analyses

Broad Agency Announcement (BAA)

Link(s) to BAA

IARPA-BAA-17-08

Solicitation Status

CLOSED

Proposer's Day Date

October 19, 2017

BAA Release Date

March 29, 2018

BAA Question Period

March 29, 2018 — May 1, 2018

Proposal Due Date

Wednesday, 30 May 2018

Program Summary

Testing and Evaluation Partners

  • Johns Hopkins University Applied Physics Laboratory

Prime Performers

  • Charles River Analytics
  • Raytheon BBN Technologies
  • The University of Pennsylvania