It is frequently said that success in the NFL is based on the ability to run the football on offense and to stop the run on defense. To those who follow the NFL this may be obvious. I was curious as to whether the numbers actually bore this out. So I went to the kick-ass website Pro-Football-Reference.com, took the rushing (for and against) stats for each team through week 14 and compiled the differential yards per game for all the teams. (At the time I did this, the Cleveland-Philly Wk14 game was not yet done---but the way the game looked in the third qtr it appears that its stats are going to be in line with our observations here).
Anyway, I compiled differentials for the passing stats too, and threw in (amongst a bunch of other stats) the won-loss record for good measure. Then I sorted the worksheet in different ways to see what, if anything, stood out. Well, one sort did stand out----the one sorted for rushing differential in descending order, and I have shown the table below.
|TEAM||RUSHING DIFFERENTIAL YARDS/GAME||PASSING DIFFERENTIAL YARDS/GAME||WIN--LOSS|
|NEW YORK GIANTS||58.3||14.8||11--3|
|NEW YORK JETS||38.3||-21.8||9--5|
|NEW ENGLAND PATRIOTS||29.2||28.8||9--5|
|TAMPA BAY BUCCANEERS||-1.2||51.9||9--5|
|SAN FRANCISCO 49ERS||-4.3||8.1||5--9|
|SAN DIEGO CHARGERS||-6.7||7.5||6--8|
|NEW ORLEANS SAINTS||-11.3||82.5||7--7|
|GREEN BAY PACKERS||-29.0||42.0||5--9|
|KANSAS CITY CHIEFS||-40.7||-38.6||2--12|
|ST. LOUIS RAMS||-58.0||-33.2||2--12|
Pretty impressive correlation, huh? If you out-rush your opponent you are almost assured of having a winning record. The exceptions to this correlation are Washington (7&7; so at least they don‘t have a losing record) and Jacksonville (5&9). But Jacksonville has a positive differential of only 0.1 rush yard per game, so that is hardly a meaningful advantage and maybe their won-loss record isn‘t necessarily against the trend.
Also noteworthy--- Only four teams with a negative rush differential have winning records. They are Tampa Bay (9&5), Denver (8&6), Arizona (8&6) and Indianapolis (10&4). Of these, Denver and Arizona are the beneficiaries of playing in horrible divisions and thereby playing weak schedules. They do not fare well against good teams; indeed, in both cases, only two of their wins have come against opponents who have a winning record.
Indianapolis is a fascinating outlier. Looking at the stats one would try to explain their won-loss success with one or more of the following speculations:
1) Opponents are moving the ball really well against Indy but not scoring proportionately, i.e. the defense is keeping the team in games by not allowing TDs;
2) Indy is being very efficient with its drives;
3) Indy‘s offense is not doing very well, but shows up when it counts and manages to win close ballgames.
If you have followed the season so far, the above pretty much summarizes Indy’s season. The defense has allowed the third-fewest TDs in the league so far and has kept many games close; indeed, it has pretty much won 2 games for the team. And Peyton Manning has been responsible for some wins in other games by being absolutely clutch in the fourth quarter.
As for Tampa Bay---They had a really anomalous game that skewed their stats. A couple of weeks ago, Carolina ran for 299 yards against them. Now, 200-yd rushing games don't happen frequently and 300-yd rushing games are very rare in the NFL. If Tampa had only allowed a bad 200 as opposed to a horrible 299, they'd have a positive rushing differential (+5 yds/gm or so) for the season---more in keeping with their 9&5 record.
I guess my point is that wins and losses are decided by so many factors, and maybe every team has an anomalous game or two in some regard every season. Ultimately, the points scored vs points allowed decide the games, and important factors like turnovers, average field-position, red-zone offensive/defensive efficiency etc affect the outcome of every game. But given all this, one single stat---the rushing differential---is able to separate the good teams from the bad, for the most part.
All in all, interesting stuff. Run, and stop the run. Pretty simple, huh?
Update: Here's a follow-up article to this post.
/ Print this post