Software system predicts 70pc of US Supreme Court outcomes

Computer specialist Chris Nasrallah claims to have created a self-learning IT system which predicts the result of Supreme Court decisions with a 70% accuracy rate.

Supreme Court

The system, CourtCash, uses three variables on which to make its analysis - number of words spoken by the justices to the lawyers for each side, the sentiment behind the words and the number of interuptions from the judges to each side. Views are mixed as to whether this is a breakthrough. The Court finds on behalf of the petitioner in 68% of cases - and specialist court reporters claim to have accuracy rates on their predictions higher than CourtCash.

Applied Maths

Mr Nasrallah is based at the Bioinformatics Research Center at North Carolina State University where he workds in the department of Statistical and Applied Maths. Source: ABA

Email your news and story ideas to: news@globallegalpost.com

Top