A Bayesian Model of “Making a Murderer”

In our previous posts (here and here), we revisited two of our research papers–one on range voting; the other on the Turing Test–and created alternate legal universes in which jury trials were decided using a range voting procedure or some form of Alan Turing’s “imitation game.” In this post, we shall discuss our 2011 peer-reviewed paper “A Bayesian Model of the Litigation Game” published in the European Journal of Legal Studies. Instead of creating an alternate legal universe (like we did in our previous posts), our Bayesian litigation paper models the existing legal system as is, warts and all. Specifically, we developed a Bayesian model of criminal and civil litigation, a model that is relevant to the central question posed in “making a Murderer”: how confident are you in Steven Avery’s guilt? Our Bayesian model includes a scenario in which the outcome of a trial is purely random (like a coin toss) and in which the moving party is “risk-loving” (i.e. in which the prosecutor is only 60% confident the defendant is guilty). Unfortunately for Mr Avery, the surprising result about our Bayesian model is that even in this random, risk-loving scenario, the posterior probability that the defendant is, in fact, guilty is pretty high.

Painting by S. Uchii

This entry was posted in Current Affairs, Justice, Law. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s