The observer effect in jurimetrics and the influence of predictive analytics on judicial behaviour

France banned analytics that single out judges. Regardless of its effectiveness and reach, this reaction raises several interesting questions, and this seminar will discuss some of them. In physics, the observer effect is an inconvenience progressively reduced by better instruments. Can we say the same about litigation analytics? Courts seem to be level 2 chaotic systems, but does this make them harder to predict? Can predictions become self-fulfilling prophecies?