I should have added this one to my previous arXiv review: On the variability of regression shrinkage methods for clinical prediction models: simulation study on predictive performance.