Big Data Sacks the NFL

As a diehard football and Patriots fan, I watch as many games as humanly possible from September until February. Usually, the best game of the week is reserved for Sunday Night Football on NBC and traditionally the Patriots have hogged the spotlight there. The broadcast team of Al Michaels and Chris Collinsworth does a great job of presenting the game and the elements that go into it. One thing that I have noticed more and more often over the past couple of seasons are the “Next Gen Stats powered by AWS” features. The broadcast team will display highlights such as how far a player travels, their maximum speed or the likelihood a quarterback completes a pass. I often find myself saying, “how can they possibly get this data and analyze it?”. To learn more, I decided to dig into how Next Gen Stats started, how they work and how they benefit not just the fans but the actual teams as well.

How it works

Starting in 2013, the NFL sought out ways to obtain more data from on-the-field play. The best way to do it was to place data trackers on the things that actually moved during the game: the players, balls and referees. They partnered with a company called Zebra Technologies to first install RFID chips into the shoulder pads of players in 2014. The NFL was able to share the statistics they complied during the games on the broadcasts in 2015, which greatly enhanced the experience for the fans at home. Fans could get a statistical measurement and representation of what made their favorite players so great.

By 2016, the NFL pushed along its efforts and tagged chips in the footballs used in the preseason and Thursday Night Football games. A year later, the NFL decided that the data was so useful that the chips must be placed in every football used in every game. The NFL also placed RFID tags in the pylons, chains and on the referees to get as much data as available. To capture the data, each stadium has 20-30 ultra-wide band receivers and there three operators at each game to ensure they are working correctly. The end result is more than 200 data points created on each play that can be analyzed by the broadcasts, fans and teams.

The RFID tags in the player’s shoulder pads are about the size of a nickel

Impact on the game

With all this data available, the question becomes how do teams use it to their own benefit? The Philadelphia Eagles created the first known analytics department in the mid 1990’s. At that time the front office, which decides which players will be on the team, were the primary users of the analytics. Data was viewed as something that could enhance the work that the scouting team would normally perform. The New Orleans Saints were the first team to individually use player tracking to their own benefit in order to understand trends in gameplay. That Saints team eventually went on to win the Super Bowl. In a copy-cat league like the NFL, this led other teams to start their own research on how they can get an edge through analytics.

When data analytics first made its way into the NFL, the teams that adopted it were reluctant to share their knowledge. The NFL leveled the playing field in 2018 when every club received the entire leagues Next Gen Stats. This provided teams the ability to judge the strengths and weaknesses of opposing players in a much more complete way than before. Teams can dive into the statistical breakdown of how specific players block, run routes, drop into coverage and tackle ballcarriers. All of this was previously done by scouts and coaches who had to make subjective observations and report their findings. These advanced analytics provide definitive numbers that can be used for easier decision making on how to attack other teams.

In the future, teams could conceivably use all of the data they built up for immediate predictive capabilities during the game itself. It’s not hard to imagine that a team could use rapid analytics to read a defense or offense based on personnel on the field and make a prediction on the exact play the opponent will run. If it ever gets to that point, it will be interesting to see how the NFL views that in terms of competitive gameplay and whether it is legal.

Teams have learned that “going for it” on fourth-down is worth it

Is it worth it?

The question that I find myself coming back to in all of this is whether the data analytics capabilities are making the game better. It is clear that there have been teams that have had success and were driven by analytics. The 2017 Philadelphia Eagles, which went on to win the Super Bowl, used analytics to drive their decisions on to “go for it” on fourth-down plays when many other teams would either kick a field goal or punt it away. The Baltimore Ravens and San Francisco 49ers are two of the teams most heavily linked to data analytics and were the top seeds for the playoffs in the 2019 season. However, the Ravens ended up losing the Tennessee Titans; a team that bucked current analytics trends and ran the ball heavily instead of passing it.

As a fan, I want my team to be using analytics to help bolster their chances of winning. However, I do not want them to solely use it to make decisions and judgements. In sports there has to be a place for making human decisions based on the current moment. Analytics can only go so far because there are several other variables that impact the game. How do you statistically measure the momentum your team or the opponent has at the time and factor that into data models? Additionally, how do you measure the experience that a coach or player has and its impact? It sounds impossible to me based on my experience playing the game and from listening to others who have played professionally. An over-dependence on either data analytics or “gut feeling” will leave the team in a weaker position. A balance between the two should be taken to put players in the best position to succeed and make evaluations of opponents.


  1. conoreiremba · ·

    Great post Michael. Reading it has definitely helped to fill the gap that football has left in our lives in the week after the Super Bowl. I do agree that the jury is still out on the impact it has on team success, though, and often times there is too much data at teams’ disposals. I think the technology still has a way to come in order to be consistently effective in in-game situations, to accurately measure the intangibles you mentioned. Off of the field, I do think it has huge merits in recruitment, especially in current times where pro-days and combines are restricted. You might be interested in this piece on the role data plays in how Liverpool F.C sign new players.

  2. ritellryan · ·

    As a big baseball fan and Math major, I have loved the merging of analytics and sports and went to the MIT Sloan sports analytics conference in Boston in 2014, an awesome experience I highly recommend if you are interested.

    I mentioned on Twitter how analytics has made some sports a lot more interesting like football and and basketball as it has encouraged more offense. However, we saw this year in baseball, in Game 6 of the World Series last year a pitcher who gave up 1 base runner was pulled because it was the 3rd time through the order (or the concept of an opener, which is smart from a numbers perspective but terrible for an excitement perspective).

    You mentioned in your post it can be easily identifiable what play people are going to run depending on who is on the field. At this point, I believe that is already the case. Everyone has a pretty good idea of what everyone is going to do, it is just a matter of execution. This is great in some cases because we are seeing the best of the best, but it also kills the idea of gut feel as the game’s strategy become more and more predetermined

  3. Scott Siegler · ·

    I noticed the AWS Next Gen Stats this season but I had no idea how it worked and what the process was leading up to it. Thanks for covering this topic. Ever since I read the book Moneyball, which details how the Oakland A’s leveraged data to win 110+ baseball games in a season with a bargain basement payroll, I’ve always been fascinated by the insights that this type of tracking can open up.

    Of course, there is also over-analyses paralysis, and that is why i agree that it is essential to find a balance between data and gut feeling. If nothing else, this type of info is another tool in the tool box that can be applied when it is helpful. Just not universally.

  4. Great post. I’m a Collingsworth fan too. Of course, I would note that he still maintains that the Pats got screwed and should have won the SB vs. the Eagles. I’m curious how much the analytics helps inform the gut feeling instead of replacing it, though. That would seem like the best use of data in a complex game such as NFL football.

  5. therealerindee · ·

    Thank you thank you for posting about the NFL. I am lost without it. I too have been fascinated by the use of data and analytics throughout the professional sports space. I also agree with you that there should be a balance between decisions made based on data and decisions made based on “gut feel”. I think, at least recently, the teams that relied heavily on their instincts and talents had a bit better of a 2020 season, and teams that were leaned more towards data driven were left behind *cough the Packers when they threw 3 times to the same receiver in the end zone cough*. Also as an aside, I know that the gaming industry is tapping into some of the data being collected on players to make the playing experience more realistic, which I am here for.

  6. lourdessanfeliu · ·

    This is a really great and insightful post. I have not always followed football, but have been getting into it more in the recent years. I was always curious to see how all the metrics were calculated and used in broadcasting in real time and this post helped me understand how everything is done.

%d bloggers like this: