The NFL Wants to Better Predict Injuries
06 December 2019 - 7:54AM
Dow Jones News
By Andrew Beaton
Imagine an NFL player who receives a crushing blow to the head.
Far away, a computer detects a trend showing that other players
have suffered injuries on similar plays. That information could
rewrite rulebooks to eliminate the play in question, changing the
way football is played.
That's the idea behind a new deal between the NFL and Amazon Web
Services, a subsidiary of Amazon.com Inc., that aims to leverage
the tech giant's data operations -- using the league's growing
array of data sets -- into a new approach to solving football's
thorniest and most troubling long-term issue.
Over the last decade, participation in youth and high-school
football has declined as more has become known about the injury
risks inherent to the sport. The NFL's handling of those problems
has been at the forefront of the discussion, casting a dark light
on America's most popular sport. A 2017 settlement between the
league and retired players established an uncapped fund,
potentially valued at $1 billion, to compensate them for ailments
such as CTE, ALS, Parkinson's and other neurocognitive
impairments.
The NFL, in recent years, has overhauled the league's concussion
protocol and implemented stricter standards for helmets to improve
the game's safety. The league has also tweaked the rulebook to
whittle down some of the game's most dangerous plays, such as wedge
blocks on kickoffs.
But the league's previous attempts to identify these trends have
been laborious and antiquated. Staffers trawled through the film of
every game to manually identify the plays that produced an abnormal
injury risk, jotting down every helmet impact in every game without
a mechanized or modern process.
Now the partners are trying to change the league's approach to
these problems by utilizing the same influx of data that has
provided new measurements about the game on the field. AWS already
works with the NFL on its "Next Gen" stats, which use chips
embedded in every player's jersey to provide advanced
player-tracking metrics. The league hopes artificial intelligence
and machine-learning technologies can more successfully parse
through this data to understand football's most dangerous
situations.
"We'll be able to model NFL players and be able to understand
their behaviors and potentially predict injuries in ways we can't
now," said Jeff Miller, the NFL's executive vice president of
health and safety.
The parties say they will use the data to gain a newfound
understanding of factors that cause injuries. That could include
the type of route a wide receiver runs or the types of blocks that
linemen execute. It could also bring in other factors, such as the
variety of cleats players wear and the surface the game is played
on. And, they hope, the data culled from this new project can be
combined with game film to understand football's risks like never
before.
"It's a really unique opportunity in the sense of having so much
data available," said Dr. Allen Sills, the NFL's chief medical
officer.
Write to Andrew Beaton at andrew.beaton@wsj.com
(END) Dow Jones Newswires
December 05, 2019 15:39 ET (20:39 GMT)
Copyright (c) 2019 Dow Jones & Company, Inc.
Amazon.com (NASDAQ:AMZN)
Historical Stock Chart
From Mar 2024 to Apr 2024
Amazon.com (NASDAQ:AMZN)
Historical Stock Chart
From Apr 2023 to Apr 2024