Hey, what’s up? For our part, we’ve just finished watching season 1 of the amazing Netflix documentary series Making a Murderer, which shows beyond a reasonable doubt how one criminal suspect, Steven Avery, was framed (not once, but twice) by the Manitowoc County Sheriff’s Department. Also, for what it’s worth, several of our recent research papers are very relevant to the issues raised in the series. In this blog post, we will discuss our 2015 paper “Why don’t juries try ‘range voting’” published in volume 51 of the Criminal Law Bulletin. Briefly, instead of requiring jurors to vote all-or-nothing, i.e. “guilty” or “not guilty,” why not replace this binary tradition with a more nuanced range voting procedure. Specifically, why not let jurors score or rate the prosecution’s case on a scale of 0 to 10. Under our range voting proposal, the highest possible score the prosecution could receive would be a perfect 120, while the lowest possible score would be 0, and the defendant would be found guilty only if the sum of the juror’s individual scores exceed a certain threshold, say 100.
Now, let’s apply our alternative range voting procedure to the Avery murder trial depicted in “Making a Murderer.” In the series finale, we learn that two of the jurors in the Avery case were undecided at the start of deliberations; three jurors were ready to convict, and the remaining seven initially had reasonable doubts. Accordingly, the two undecided jurors could have assigned a “5” to the prosecution’s case, the mid-point between 0 and 10. By contrast, each of the three pro-conviction jurors could have rated the prosecution’s case a 9 if they were 90% certain of the defendant’s guilt, an 8 if they were only 80% certain, and so on. Lastly, each one of the seven reasonable-doubt jurors could have scored the prosecution’s case a 1 if they were 10% certain of the defendant’s guilt, a 2 if they were only 20% certain, and so on. Once each juror assigns a score or rating to the prosecution’s case, they would then add up all the scores, and Avery would have been declared guilty only if the sum of all the juror’s scores exceeded the threshold value.

How confident are you in the prosecution’s case?
Pingback: The Turing Test and “Making a Murderer” | prior probability
Pingback: A Bayesian Model of “Making a Murderer” | prior probability