Dougherty: Packers poised to profit from NFL data dump

Pete Dougherty
Packers News
View Comments
Green Bay Packers general manager Brian Gutekunst watches practice for the Senior Bowl in Mobile, Alabama on Tuesday, Jan. 23, 2018.

The NFL is officially embarking on an era of big data, and the teams willing to invest the money and effort to make sense of it all are going to come out winners.

That should include the Green Bay Packers, if under President/CEO Mark Murphy and new general manager Brian Gutekunst they continue the progressive approach the franchise has shown in the NFL’s analytics movement the last 15 to 20 years.

The big news came last week with reports that the league soon will share player-tracking data from games over the past two seasons. The league has been tracking players on game day with extremely precise RFID (radio frequency identification) technology for several years, but for the last two years was giving teams the data only for their players.

Starting in April, though, the league will provide all data from the last two years, and starting this season will make available new data from every game, every week.

“It holds the potential to be a game changer,” said Mike Eayrs, the Packers’ former director of research and development, and one of the first great analytics experts in league history. “This is like building the Grand Coulee Dam of football.”

By that, he means this is an enormous project for NFL teams. We’re talking an overwhelming amount of data here. To make it of any use to coaches and scouts, teams will have to amp up their computing power and storage, and hire more analysts. (The Packers already have four people in their football technology analyst department: director Mike Halbach, analysts Ryan Feder and Connor Lewis and assistant Chris Gaines.)

DOUGHERTY: Aaron Rodgers' road to record-setting deal

RELATED: New Packers QB coach Cignetti must win over Rodgers

The potential implications of this data dump are profound, though it will take several years, if not longer, for teams to figure out how best to mine all this new information. But one way or another, it should save teams time and manpower in game planning, help scout players for free agency and eventually the draft, and aid monitoring their own players for fatigue over the course of a game and susceptibility to injuries.

Some quick background first. With this RFID technology, chips embedded in players’ shoulder pads are read by sensors placed all around the stadium that can pinpoint where every player is on the field at any moment, as well as how fast each is moving and the power he’s generating.

Now think about how much data the league will be dumping on teams in April. There are 22 players on the field for about 160 plays (from scrimmage and special teams) a game. Every player participating is tracked on every play in every game.

Teams that want to dive into the data will have to make a big investment – Eayrs estimated well into the hundreds of thousands of dollars, if not more. They will have to make major upgrades in computer hardware and storage, and develop far more sophisticated software than they’ve been using for analytics and videotape up to now. 

“The people who understand how to use (tracking data) are in the aerospace industry and aircraft industry; they’ve done this to develop jets and missiles for years,” Eayrs said. “My guess is you’ll see some new faces in the (Packers’) building, and they’re going to be really competent engineers that are able to work off platforms that can deal with these file sizes.”

Eayrs said there are essentially three different ownership approaches on such matters: highly competitive teams that invest freely to win, teams with a spend-with-responsibility approach that invest as costs come down and teams that won’t spend until the league essentially mandates it.

Eayrs counts the Packers in the first group. This is where not having an owner, and thus investing all profits back into football and the franchise’s financial future, can be to their advantage.

Teams have been getting their own game data the last two years, so the more progressive already have been working on analyzing the information for a year or two.

“My best guess is (the Packers) are way ahead of the curve,” said Eayrs, who retired in September 2015.  

So what are some of the possibilities for utilizing this data? Eayrs provided several possibilities.

First, it can help in game planning. Currently, assistant coaches (usually quality control) convert videotape into play diagrams for players by watching video and then drawing the plays with a computer app. Now, teams (or third-party companies) can develop apps that will use player tracking to draw the plays, and they’ll show precisely what everyone does, not just approximate. Coaches then can spend their time on other duties.

But that’s only the start. The data will be most useful for scouting opposing players – for weekly game plans in-season and free agency in the offseason.

Up to now, coaches and scouts have relied on comparative analysis – they draw similarities between a new player with someone they’re all familiar with from the past. This new data will allow for what Eayrs calls “precision analysis,” that is, the ability of software to compare and contrast players with concrete numbers.

“It changes the way you evaluate football,” Eayrs said.

So, for instance, the data can show how much power a defensive lineman such as Mike Daniels generates at the point of attack. Over time, teams should learn how much power a defensive lineman needs to, say, defeat a double team.

They can show how many strides it takes for an explosive player such as Clay Matthews to get to top speed.

They can show how fast an offensive lineman moves in a short space, or how much his punch slows a pass rusher. Or how much a cornerback’s jam slows a receiver, or his closing speed on the ball.

When two players collide, they can quantify how much leverage one gained on the other.

They can help identify if a player really did loaf on a play, as well as how much players get fatigued over the course of a game.

And chips implanted in the football reveal actual quarterback arm strength in miles per hour and accuracy on all types of throws.

This data capture will filter down to college football quickly, if it hasn’t already at the bigger Division I programs. Eventually NFL teams will be able to buy the data to help evaluate and project college players as well.

The Packers have been one of the more progressive teams in the league in implementing technology and analytics – they hired Eayrs in 2001, and coach Mike McCarthy was one of the early adopters using GPS tracking in practice, mainly to monitor workloads for injury prevention.

But McCarthy also has warned in the past of information pollution. Too much information can overwhelm, so the challenge for teams will be working with their analysts to bring meaning to the mounds of raw data.

For the teams that do it right, it should allow for a competitive advantage as an augment for human judgment, though not a substitute.

“What’s the big upside?” Eayrs said. “You’re gaining amazing precision, and you have the ability to quantify everything that’s happening on the field. … (But) at the end of the day you still have to decide whether you’re going to take Player A or Player B, and there are going to be intangibles that come into this.

“Who’s got the better emotional stability? Who’s got the better attitude? Who’s going to be more cohesive in the locker room? Who’s going to alienate the guys around them to the point that they’re not going to consider them a good teammate.? Those things, you can’t capture with player tracking.”

View Comments