It seems that appropriate methodology in the development of prediction models is becoming more common in our field. These two recent studies caught my attention: Development of a model to predict the probability of incurring a complication during spine surgery and Prediction of 90-day mortality after total hip arthroplasty. Both studies report calibration performance and also interpret it. It was interesting to see that in the study model, discrimination for general complication after spine surgery was “only” 0.64. And in the second study the discrimination for the 90-day mortality was “only” 0.75-0.78. It could be easily thought that these would be relatively straight forward to predict but it is not definitely the case. There is so much unexplained variability in our practice which needs further research.