The League

Smarter Stats

The Value of Rushing Titles


Going into Week 6, Atlanta running back Michael Turner led the NFL in rushing yards with 543 on only 103 carries, for a 5.27 per-carry average. These numbers sound very solid until you break them down game-by-game, and opponent-by-opponent. 220 of Turner's yards came in Week 1 against the Detroit Lions, who currently rank last in the league in defensive rushing DVOA* -- in no small part because of Turner's efforts, but also because the Lions defense just stinks.

In his second game, Turner ran for 42 yards on 14 carries against the Tampa Bay defense, currently ranked third in defensive rushing DVOA. Week 3 saw a 104-yard game against the Chiefs, currently ranked 25th, and Week 4 saw another blockade -- 56 yards on 18 carries against the 16th-ranked Carolina defense. Week 5, Turner was up against another down defense, racking up 129 yards against the 29th-ranked Packers. And last Sunday, against the 8th-ranked Bears, Turner gained a "whopping" 54 yards on 25 carries.

After his performance versus the Bears, Turner also lost his NFL rushing lead to Clinton Portis. Still, his up and down performance got me thinking -- do NFL rushing champions require that the majority of their yards come against creampuff defenses, and is there a way to look at this on a per-game basis? I then went back ten years to 1998, when Terrell Davis rushed for 2,008 yards, and listed the per-game performances against the season total rushing DVOA of the defenses they faced. Based on their numbers and the quality of those defenses, backs were given success points on a scale from zero to five:

Five Points -- 150+ yards against top 10 defensive rushing DVOA
Four Points -- 100 - 149 against top 10 DVOA
Three Points -- 100+ against 11-20 DVOA
Two Points -- 75-99 against top 15 DVOA
One Point -- 100+ against 21-32 DVOA
No Points -- less than 75 against top 15, less than 100 against 16-32 DVOA

This way, we're not giving the same value to yards piled up by elite backs against terrible defenses, while rewarding productive performances against better opponents, This goes back to one of the main Football Outsiders tenets -- that all stats should be adjusted for opponent. Based on these numbers, here are the most-to-least "valuable" rushing titles of the last decade, based on success points. Click on the link below:

View image

Now, let's talk about the numbers, and the values behind them. James has the top two seasons of the last decade in terms of sheer value because he put up so many good games against fine defenses. In 1999, he rushed for over 100 yards in ten different games, and five of those came against top 10 defenses in rushing DVOA. He had only four games under 75 yards, and two of them were against the best (San Diego) and fifth-best (Buffalo) defenses that year. 2000 featured similar excellence -- of his nine 100-yard games, six came against top 15 defenses.

Priest Holmes' 2001 season may have been the spikiest of all -- in the first nine games of the season, he put up five "zeroes" and two "fives". Mixed in with clunkers like the 55-yard game against 31st-ranked Indianapolis and the 71-yarder against the 22nd-ranked Jets were two bravura performances, when he totaled 331 yards against the Steelers and Chargers, the two best run defenses that year.

It's perhaps fitting that the rushing title claimed by the enigmatic Ricky Williams was a bit of a mirage. His one truly transcendent game came in week 14, when he put up 218 yards on the 8th-ranked Bears. Other than that, there were a lot of high numbers in low places: 151 yards against the 31st-ranked Jets, 185 yards against the 26th-ranked Patriots, and 228 yards against the 24th-ranked Bills. Ricky was truly a beneficiary of his division -- the AFC East had a bunch of bad run defenses, and it was a great time to be a Dolphin.

Speaking of divisional beatdowns, how about the 295 yards gained by 2003 champ Jamal Lewis against the Browns? That may have been a single-game record until Adrian Peterson broke it last year, but huge games against the NFL's 31st-ranked run defense mean little in our system. In 2004, Martin had a nice run of games against top ten defenses, but he also benefited from games against the lesser lights -- witness the 196-yard game against the 21st-ranked Bengals. Alexander used the NFL's best offensive line and a cushy schedule in 2005 to glide to his title, though his 110 yards against the second-ranked Giants was fairly impressive.

And that leaves us with the two-time defending champ. How is it that LT2 can lay so low, when the consensus has been that he's the league's best back? Well, I'm not going to argue with his greatness, but let's take a look at that interesting 2007 number. Tomlinson put up eight "zeroes" and six "ones" because he never gained big numbers against elite run defenses. In fact, he faced five different top-ten defenses, and never gained more than 77 yards. When he did run free, it was against the league-worst Raiders (198 yards) or the 24th-ranked Chiefs (177 yards).

As for Michael Turner? His six games, as boom-and-bust as they've been against good and bad defenses, leave him with three "ones" and three "zeroes". And that's why we like to adjust performance numbers for opponents -- because we can then see who really earned their stat titles, and who received gifts from their opponents.

*All DVOA numbers as of 10/13/08, before Week 6 updates.

By Doug Farrar  |  October 14, 2008; 2:00 AM ET  | Category:  Statistics Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati  
Previous: Why Scouts Need Stats | Next: The Curse of 370


Please email us to report offensive comments.

These stats do not take into account defensive game planning at all. LT faced defenses designed to take him out the game. Defenses couldn't do that against Edgerrin James, since he played with future Hall of Famers Peyton Manning and Marvin Harrison.

Posted by: AllScott | October 17, 2008 1:46 PM

Great insight! Edge benefitted greatly from Indy's high-octane offense in his back to back rushing title seasons. Curtis Martin's season was the most impressive because of his 30 year old age. The age that is conventionally perceived to begin an inevitable decline.

Posted by: Motown Prowess | October 15, 2008 9:35 PM

I had no idea LT sucked so bad or that Edge was so good. Maybe DVOA is off

Posted by: fram | October 15, 2008 2:42 AM

How about Clinton so far??? Not bad huh

Posted by: Meraj | October 15, 2008 2:17 AM

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company