Kavanaugh - Ford Senate Testimony Bayes' Rule Calculator

How can we reason about Judge Brett Kavanaugh's and Dr. Christine Blasey Ford's testimony before the U.S. Senate Judiciary Committee, to draw conclusions about whether Kavanaugh really did assault Ford?

Use the calculator below to apply Bayes' Rule.

The calculation is based on a simple but plausible mathematical model for how a thoughtful person of any political persuasion might reason about the evidence. The model is described here.

Adjust the green sliders to set assumptions about Ford's and Kavanaugh's belief states, credibility and motivations, their testimony, and other factors.

On the right is the posterior probability that Kavanaugh actually assaulted Ford, under these assumptions, given their testimony. At the bottom are other posterior probabilities taking their testimony both in isolation and in combination, and also considering what one would conclude had Judge Kavanaugh's testimony been calm rather than angry.

This question addresses a counterfactual. Of course in the end, Ford did testify. But in the chain of events it didn't have to turn out that way. Depending on Ford's motivations and convictions, and depending on the politics of the Senate Committee, it could have turned out that she or the Democrats did not manage to get her an opportunity to speak at the hearing. Proper Bayesian analysis requires that we consider all alternatives for the variables and events in question, even if we know after the fact which ones actually happened or not.

This question addresses a counterfactual. Kavanaugh had the option of testifying either calmly, or else with passionate outrage and anger. This question probes assumptions about whether Judge Kavanaugh would be more likely to testify in one manner or the other, supposing that he were either guilty or innocent.