Virtually every clinician will agree that an informed assessment of the likelihood of success is a mandatory part of patient selection and planning the procedures themselves even though the clinical outcomes of an invasive procedure often can't be predicted precisely. It also goes without saying that this assessment plays a vital role in quality improvement undertakings involving operators and institutions, as well as determining the value of certain procedures. Unfortunately, we are poor at making such prognostications. This ability is particularly limited when it comes to transcatheter aortic valve interventions. For example, a transcatheter aortic valve implantation (TAVI) mortality model developed using >13,000 patients from the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy (TVT) Registry was well-calibrated, but its discriminative ability was modest, as evidenced by a C-statistic of only 0.66 in both US and Swiss populations , as was a subsequent attempt to predict periprocedural stroke . Attempts to characterize predictors of stroke using databases from large clinical trials and standard statistical techniques were even more limited and also failed to identify reliable predictors of stroke risk .