So the algorithm is correct 70% of the time, while (later in the article) the Supreme Court reverses 65.9% of its cases?<p>This is not much of an improvement over always guessing "reverse".
Not very impressive, I would say, at 70%. Although there are some data points at 90% that interest me.<p><a href="https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds" rel="nofollow">https://en.wikipedia.org/wiki/The_Wisdom_of_Crowds</a>
Url changed from <a href="http://arstechnica.com/science/2014/07/algorithm-predicts-us-supreme-court-decisions-70-of-time/" rel="nofollow">http://arstechnica.com/science/2014/07/algorithm-predicts-us...</a>, which points to this.