Nothing derails a championship faster than an injury to one or two of your key players. Everyone knows this much:

Certain positions and even certain players get injured at a higher rate. We see players reaggravating hammy and ankle injuries all the time. And if a player has had four concussions, he's far more likely to get another.

So no doubt there's some likelihood of injury trends among players. But how and why can we use that to try and predict the future? First, take the most comprehensive injury database detailing injuries for every major NFL player over the last 30 years. Then apply advanced machine learning to define and track these trends.

You get a model that uses over 1,000 variables to predict injury outcomes. Meaning, you've cracked the scientific code for predicting injuries. It is impossible to predict every injury. But we can predict which players have a higher injury probability AND the length of time they're likely to miss. We know because we've actually done it.