Tweet
Something I thought someone on here would like.
To read the rest of the article and get a full grasp of what Bradley Woodrum did, click the link to the article.It has models and such.
http://www.mlbtraderumors.com/2016/0...surgeries.html
From Derick Velazquez in January to Lance Lynn in November, there were 112 ulnar collateral ligament (UCL) injuries requiring reconstructive surgery — commonly called Tommy John Surgery (TJS) — in the 2015 season. Once a career-killer, UCL injuries have become a much more survivable injury over the last 30 years. And while more and more players are successfully returning from TJS, the procedure itself is a catastrophic event and requires a minimum of a year to recover.
That makes predicting UCL injuries a valuable and worthy endeavor. From the GM to the fantasy owner, being able to steer away from players with early warnings signs of UCL injuries can save a team’s season. The red flags for UCL injuries are not big, though, and many UCL injuries appear from nowhere. But using a large data set, culled from a variety of valuable resources, we can find the tiny red flags, the little baby red flags.
For the past seven months, I have been working with Tim Dierkes and his staff to develop a model to predict Tommy John surgery. The creation of this model required, quite literally, hundreds of thousands of lines of data and hundreds of man hours to combine and connect and test data from a variety of disparate sources. The project also took, as a sacrifice, one of my computer’s CPUs, which burned out shortly after completing some herculean computations. Fare thee well, i7.
The Results
The following is an attempt to quantify the risks that foreshadow potential UCL injuries. It is a combination of FanGraphs player data, Jeff Zimmerman’s DL data, PITCHf/x data, a bunch of hard work, and the keystone data: Jon Roegele’s TJS data, as stored on Zimmerman’s Heat Maps. We also checked our numbers against Baseballic.com, which houses arguably the most comprehensive injury data online.
And while most efforts at quantifying TJS risk have focused on recent appearances or recent pitches, our research takes a step further back and examines injury risks on an annual basis. It seeks to consider the problem from the GM’s view, and not the game manager’s.
The following names are sorted by greatest risk to least. For more details about the columns and the model that has created this data, continue reading after the embedded data.
That makes predicting UCL injuries a valuable and worthy endeavor. From the GM to the fantasy owner, being able to steer away from players with early warnings signs of UCL injuries can save a team’s season. The red flags for UCL injuries are not big, though, and many UCL injuries appear from nowhere. But using a large data set, culled from a variety of valuable resources, we can find the tiny red flags, the little baby red flags.
For the past seven months, I have been working with Tim Dierkes and his staff to develop a model to predict Tommy John surgery. The creation of this model required, quite literally, hundreds of thousands of lines of data and hundreds of man hours to combine and connect and test data from a variety of disparate sources. The project also took, as a sacrifice, one of my computer’s CPUs, which burned out shortly after completing some herculean computations. Fare thee well, i7.
The Results
The following is an attempt to quantify the risks that foreshadow potential UCL injuries. It is a combination of FanGraphs player data, Jeff Zimmerman’s DL data, PITCHf/x data, a bunch of hard work, and the keystone data: Jon Roegele’s TJS data, as stored on Zimmerman’s Heat Maps. We also checked our numbers against Baseballic.com, which houses arguably the most comprehensive injury data online.
And while most efforts at quantifying TJS risk have focused on recent appearances or recent pitches, our research takes a step further back and examines injury risks on an annual basis. It seeks to consider the problem from the GM’s view, and not the game manager’s.
The following names are sorted by greatest risk to least. For more details about the columns and the model that has created this data, continue reading after the embedded data.
To read the rest of the article and get a full grasp of what Bradley Woodrum did, click the link to the article.It has models and such.
http://www.mlbtraderumors.com/2016/0...surgeries.html
Comment