The injury predictor is a patent pending algorithm that works out which players are the most likely to be injured over the course of a season.
The algorithm is based on several factors that we tested to be true signals of an injury occurring. Some of these factors are things like a players age, their position, what injuries they have had in the past and what their projections are for touches in the coming game. All of these data points are weighted and placed in an equation that figures out which players are most likely to get injured.
Yes. Very accurate as a matter of fact. Last year we got it right 2 out of 3 times for players we thought would get injured as well as the ones we thought would not be injured.
No. The injury predictor will only tell you how likely a player is to get an injury. Of those injuries a player is likely to get we are able to parse out which ones are the most probable and that information is available on request but not currently part of the feed.
Last year Trent Richardson was your most likely to be injured player yet he did not get injured. What happened there?
A few things happened T-Rich that prevented him from getting injured. First off lets be clear – he was injured as he needed surgery on his labrum in his shoulder in the offseason but he was able to play through it. Secondly once he was traded to the Colts in week 3 last year his workload took a dive as he was clearly in over his head learning a new system. The probability we had assigned him was based on the huge workload of the previous year and this shift changed his likely injury outcome.
We track any injury that will take a player out of a game. We look for signals as to how severe the injury that includes things like missed practices, missed games as well as public statements from teams, players and other sources the media are able to quote. We also track surgeries players undertake in the offseason as these are strong signals that a weakness exists and will elevate the potential of injury to occur