<
>

Playing with the numbers: Disproving a cold-weather theory

11/7/2007

Every year I get similar e-mails right about now from owners who are sitting 5-4 or 4-5 and are trying to figure out how to ensure their team gets to the playoffs. Their e-mails usually begin like this: "With the colder weather now upon us, teams obviously will shift their focus to the running game. Because of this, do you think it's wise for me to trade (insert name of elite wide receiver and fringe No. 2/flex running back here) for (solid No. 2 running back and No. 3 fantasy receiver)?"

It makes me want to scream. First, you never trade two players from your starting lineup for two new starting lineup players if the latter do not score more points (or project to score more points) than the former. Never ever. Second, the much-ballyhooed theory that teams run significantly more during the latter part of the season because it's cold is garbage. As in, it doesn't happen. Take a look at the following chart, which illustrates by week the percentage of NFL rushing yardage to total NFL offense over the past 10 years.

OK, if you want to be completely technical, Weeks 15, 16 and 17 are among the highest weeks, and Weeks 1 and 2 are the two lowest. However, the difference between the highest week (16) and the lowest week (1) is only nine yards for the average NFL team. When you consider that most teams employ an 80/20 split for the starting running back and his backup, you are now looking at an average of about a seven-yard differential for your fantasy player.

As I write this, I imagine that many a cynical reader is getting ready to call me an idiot, probably in ESPN Conversation (Beta!) below, because a global look like this shouldn't be used. Instead, I should be looking at only the franchises in cold-weather cities because those are the teams that would be most affected. First, I'll state that those cynics are wrong because a majority of the teams would be affected by weather based solely on geography. Second, I actually did the statistics for each NFL team and broke the stats into four-week blocks to show that there is no direct correlation of rushing yardage to cold temperatures and geography. Those results are shown here:

The largest variation versus a team's season average within any four-week period over the second half of the season is owned by the Houston Texans, with plus-8 percent during weeks 13 through 16. Not only is Houston's average daily high in December is about 66 degrees Fahrenheit, but they also play in a retractable roof stadium.

The Jaguars post the second-highest variation at plus-4 percent. Daily high temperature in Jacksonville during December: 67 degrees Fahrenheit.

The standard deviation for all the four-week groupings is 3.7 percent. The Jaguars and Texans are the only teams that display a four-week grouping during the second half of the season with a variance from their season average greater than that margin.

The Giants and Jets play their home games in the same stadium. During weeks 13–16, the Jets differential is minus-3 percent to their season average, and the Giants are plus-1 percent. The Raiders and 49ers play a mere miles from each other in the Bay Area, but during weeks 9–12 the Raiders' run split has dropped three percent while the 49ers increase two percent.

If you still believe that temperature has a significant impact on a team's offensive philosophy, then feel free to trade Greg Jennings for Cedric Benson. Pay no attention to the fact that Jennings has outscored Benson on average by 6 points per game. After all, it gets cold in Chicago in December!

Ken Daube is a fantasy football analyst for ESPN.com