Quantcast
Channel: FT data » Jason Abbruzzese
Viewing all articles
Browse latest Browse all 7

Understanding American football, one standard deviation at a time

$
0
0

If you watched every American professional football team play every snap, analysed its success, compared it to all other plays in that situation, and then weighed it according to the defence it faced, you would reach an important conclusion: the Denver Broncos have the best chance of winning the Super Bowl on February 3.

That is according to FootballOutsiders.com, whose popular DVOA rating puts the Broncos as a slight leader (24.3 per cent) over second-favourite New England Patriots (23.6 per cent).

Founded by Aaron Schatz, an ESPN columnist, the site provides an in-depth analysis of every single play of the NFL season along with various ratings based on that analysis.

“For me, personally, I started this because I wanted to make better commentary,” says Mr Schatz. “The bigger gap that needed to be closed was between reality and the nonsense of colour commentators.”

Baseball’s “sabermetrics” movement – popularised by Moneyball (both the book and movie) – led the way increasing the sophistication of the statistics used in sport. It was primarily developed by people outside the baseball establishment to help teams properly value individual players.

In contrast, Football Outsiders does provide some predictive qualities, but the data analysis offers a different utility from the of baseball: to educate and inform football fans about a complicated and sometimes misunderstood sport.

“The development [of data analytics in football] has worked very differently from baseball. In baseball, what you had was a lot of guys on the outside that were fans,” says Mr Schatz. “They built this community of people … Gradually the ideas slowly filtered into front office.”

While baseball executives scoffed at the idea that numbers could be more accurate than their own perception, executives of American football teams had been breaking down film and using data to inform their decisions for years.

“In football, front offices have always had far more information than outsiders have had due to coach’s film, and they’ve needed to know economics for quite a long time due to the salary cap,” says Mr Schatz.

Football teams keep a tight lid on their information, lest a competing team get a look inside and figure out their weaknesses. This secrecy, along with the fact that film of the entire field (commonly known as the all-22) was unavailable before 2012, kept fans in the dark apart from the basic, widely-available statistics. There is even greater transparency now, with data on every NFL play from the ten season available online.

Mr Schatz’s desire for a better way to understand football bore DVOA, short for Defence-adjusted Value Over Average. While Football Outsiders produces a variety of different analyses, DVOA is their standard-bearer. It is calculated by following every single play a team runs, analysing its success, and weighing it against the quality of the defence.

What emerges is “a general rating of a team’s efficiency”, Mr Schatz says. This comes in handy particularly with the NFL playoffs entering their divisional round (or what could be considered the quarterfinals round). Instead of judging a team purely by its record, the average fan has a number that boils down every single play that team has run over the course of the season.

Of course, DVOA is far from perfect. Mr Schatz emphasises that taking in a variety of analyses of a team’s strengths and weaknesses provide for the best picture. And at the end of the day, these stats give fans a more realistic view of the factors that can influence a game – even luck.

“I think that people do understand a little bit more that the running back is not as important as people used to think,” says Mr Schatz. “And I think that slowly people are starting to understand the effect of luck on the game.”


Viewing all articles
Browse latest Browse all 7

Latest Images

Trending Articles





Latest Images