Injury Prediction

Using machine learning to predict (& prevent!) injuries in endurance athletes: Part 1 Alan Couzens, M.Sc. I've talked about different ways that we can assess the individual'dose- response' relationship or, more specifically, how we can work out just "what it takes" for a given athlete to reach a given performance level. I have suggested that, given recent advances in machine learning, the current models are largely out-dated and that we can find more accuracy in models that look at the independent impact of volume & intensity rather than wrapping these variables into one'training stress' metric. But there is another addition to the current performance models that is far more important and has the potential to be even more powerful in its application than load- fitness modeling: Turning the focus of our models to those things that prevent us from ultimately doing more load! This is the flipside of the'more is better' dose- response model.