Friday, November 11, 2011
The nine members of the US Supreme Court show patterns of common voting that have allowed researchers using network theory to guess how individual justices would have voted in past cases with an accuracy of up to 83 per cent.
According to two Spanish scientists, algorithms used in biology provide the predictions. They require no information about the Justices’ political leanings or the nature of the cases. But maybe “prediction” is not the most precise word. Because the predictions require knowing how other Justices voted in the case, the method cannot predict how the Justices will vote in future cases. Here is the article from the New Scientist.