Splitting Localization and Prediction Numbers

In this paper the work done by Newelski and Roslanowski is revisited to solve a question done by Blass about one of the possible evasion and prediction numbers. This led to define a variation of the $k$-localization property (the $(k+1{)}^{\omega }$-localization property) and the use of a forcing notion with accelerating trees.