I was having fun with numbers again.

I wanted to know how important having a top 5 RB in rushing yards is to winning in the NFL. Both regular season and post season.

I did a little checking and put the numbers in a spread sheet.

Here is what I came up with.

Actual, Stat type...... Ave for any random 25 teams
9..... Ave wins/year.. 8
11.... Post season.... 9.375
7..... Post wins...... 8.590
11.... Post loses..... 8.590
0..... Super Bowl..... 1.600
0..... Super Bowl wins 0.800
0.563 winning %...... 0.500
0.389 post win %..... 0.500
0..... SB Win %....... 0.031
0..... Conf win %..... 0.063

First I must appologize. I tried loading an image of the spread sheet so I could insert it into the body of the post, but I was unable to for some reason. No big deal, I worked around it.

My assumption was that NFL leading running backs do not make a team great. They don't win games like a league leading QB would.

The first thing I noticed after collecting the data was that they win slightly more than average. For the 5 best running backs every for each year, you would expect more than a .563 winning percentage. In fact, the top RB for each of the last 5 years averaged .500 for the season. The 5th ranked backs fared better with a .613 winning %.

The second thing that I noticed is that, while they went to the post season slightly more than average, they lost more than average. They also rarely made it to the Conference playoff game and never won.

It seems that Dominating running games in the NFL are not effective in the post season. They are hardly better than average in the regular season.

I will probably continue to go back a few more years and see if there is a pattern. Trends are not easy to spot with a short sample like this.

---

Attach the image and I'll show you how to embed it. Also, I went through what you wrote and didn't see what merited a running back to be top five.

---

Rushing Yards. It was clearer in the spreadsheet.

I have done that many times. I usually uploaded them in my album and then use the picture link button on the tool bar. But I couldn't save the picture in my album.

---

I would think a more interesting examination would be a regression of total team rushing yards (independent variable) vs. wins (dependent). You could do it by season, playoffs, super bowls, etc, and then combine several seasons. Regressing with just total team rushing yards will explain how much of the variability in wins is explained by the rushing yards (by the coefficient of determination, or R^2).

I'd love to see something like that, but I don't have enough time to do it myself. If the data was into a spreadsheet, I could pop out some graphs and such pretty quickly, though. Even a true statistical analysis in SAS is easy one the data is collected. Hint hint! heh. 😉

If I had to guess, it would be a weak to moderate linear correlation, at best.