At the end of december Urs reported differences between anaRecoil and anaQA in the results. I found two differences: anaQA was applying the acceptance cuts to the leptons while anaRecoil was not. More important, the default file format of anaRecoil was the april02 one. Unless Urs was explicitely setting -y2 (which I doubt), anaRecoil would have returned nonsensical results. Both these differences are fixed in VubAnalysis rf-010903 ciao ric