Predicting victory in the NFL

Kirby Lee-USA TODAY Sports

With no national or local attention, here's a look at the value of the time of possession statistic, and a theoretical discussion of the ramifications.

Any given Sunday. An adage embedded into the football lexicon, in normal contrarian thought I wondered how we could beat it. What is the best metric for predicting victory? There will never be a catchall way to predict a win or loss in the NFL. There are too many variables at hand. But what comes closest?

In the past five to ten years hockey has undergone a renaissance in advanced statistics. The big two stats, Corsi and Fenwick, and their subsidiaries all deal in possession of the puck. But because in hockey there is no reliable and easy to come by way to measure possession, it is indirectly tracked through shot attempts, blocked, wide, and otherwise. Football doesn't have that problem at all. It's said that possession has the highest correlation with winning in the NHL. So what about the NFL? If one team has the ball, the opponent's odds of scoring are microscopic, inversely related to time spent on defense. Hold the ball and your chances of victory increase.

Going through each game from the 2013 season, I recorded which team won the possession battle, checked against the contest's winner and loser. It goes without saying that correlation does not prove causation. Holding the ball in and of itself does not lead to victory. Teams run out the clock. Maybe Dallas dominates possession for three quarters, but Kansas City comes back in the fourth quarter, takes the lead and then sits on the ball for ten minutes. Dallas did most of the heavy lifting, but couldn't hold on in the end.

It's not necessarily about pace either. Whereas Carolina, the slowest moving team in the league last year, held the fifth best mark for possession (53.13%) last year, division rival New Orleans was third with 54.38% possession.

Here's a look at the chart I made to keep track. Viewer discretion is advised.

Mark_ii_medium

Green represents games in which the victor won the possession battle, red has the losing team achieving supremacy. Red games with gold lettering represent games where the winning team lost the possession battle, but won the game via a fourth quarter comeback. The sole gray game is Green Bay's tie with Minnesota.

Accounting for all 256 regular season games last season, winner of the possession battle won 66% of all games.

Naturally there are outliers. To my eye, the two teams which most defied this metric were Houston (2-14) and the Philadelphia Eagles (10-6). The Texans won the possession battle eight times, which would portend an 8-8 record. On the contrary, Philadelphia won the time of possession just twice over their 16 games. Houston possesses a strong defense, and a solid running game, enabling high possession; however, the missing ingredient, it would appear that quarterback deficiencies held the Texans back in 2013. Crippling turnovers in the red zone, or extended drives which end in punts or field goals, can be lethal. On the other hand, the Eagles, operating at breakneck paces, were able to win often, despite holding some of the season low times of possession.

For the record, Carolina (12-4) won the possession battle in 12 games last season, which would predict a 12-4 record. The contrast comes with predicted victories against Buffalo, Arizona, and New Orleans (Week 14), exchanged for losses to New England, New Orleans (Week 16) and Atlanta (Week 17).

Some of the criticism thrown to OC Mike Shula points to the offense's struggles down the stretch, the near defeat in Atlanta, and the performance in the de facto NFC South Championship Game. Perhaps not coincidentally, over their last three games the Panthers put up a possession rate of 42.69%, the second worst number in the NFL over that time period, rubbing shoulders with Tampa Bay and the Jacksonville Jaguars. The offense was unable to control possession.

In the playoffs, with a smaller sample size, time of possession becomes much less important. Not exactly a huge decline, 55% of teams winning the possession battle won their contest. Although, in the first two rounds just three of eight winners were able to control possession. A clean sweep of the championship and Superbowl games raised the bar from 38% to 55%.

Of course, the all-important clarification, correlation does not prove causation. However two-thirds is a fairly high correlation for a such an intricate game. Investigating overall team possession, over the last five years, 47 of the 60 playoff teams (78%) had total possession rates over 50%.

From a theoretical standpoint, perhaps this is the optimal way to build a team for longterm regular season success. A strong defense, a QB who will minimize mistakes, and a capable running attack. If you eliminate turnovers, and keep your offense on the field (by sustaining drives), the opposition cannot score.

Through this perspective, new schemes are ephemeral. Eventually, through the maw that is the NFL season, opposing coaches develop effective counter-schemes. Improvements to the game are adopted, adapted, and fine-tuned. Counter and adapt, it's an immutable process. Unlike playing styles, time of possession, like points, or yards, is a constant.

X
Log In Sign Up

forgot?
Log In Sign Up

Forgot password?

We'll email you a reset link.

If you signed up using a 3rd party account like Facebook or Twitter, please login with it instead.

Forgot password?

Try another email?

Almost done,

By becoming a registered user, you are also agreeing to our Terms and confirming that you have read our Privacy Policy.

Join Cat Scratch Reader

You must be a member of Cat Scratch Reader to participate.

We have our own Community Guidelines at Cat Scratch Reader. You should read them.

Join Cat Scratch Reader

You must be a member of Cat Scratch Reader to participate.

We have our own Community Guidelines at Cat Scratch Reader. You should read them.

Spinner.vc97ec6e

Authenticating

Great!

Choose an available username to complete sign up.

In order to provide our users with a better overall experience, we ask for more information from Facebook when using it to login so that we can learn more about our audience and provide you with the best possible experience. We do not store specific user data and the sharing of it is not required to login with Facebook.

tracking_pixel_9341_tracker