A Football Statistic I'd Like to See: Median Yards Per Rushing Attempt

During last week's game against the Cowboys, the Titans' Chris Johnson averaged 6.9 yards per rush. For the entire 2009 season, Johnson averaged 5.6 yards per rush. (Tonight he averaged 4.3.) These are great numbers, but anyone who watches the Titans each week knows that these numbers tell us little about how Johnson runs. I would venture to guess that the majority of CJ's runs are for two or fewer yards. But the occasional 60- or 70-yard run give his yards-per-rush average a boost.

CJ is a great back whom opposing defenses must always account for. But, despite what his statistics suggest, he isn't a back the Titans can count on to pick up four or five yards on first down or three tough yards on third down.

Johnson's yards-per-rush average is much better than that of Eddie George, one of CJ's predecessors in the Titans' backfield. While Eddie never put up 170 yards in a single game or broke off an 80-yard touchdown run, he was more likely than Johnson is now to rush for at least three yards on a given play.

The yards-per-rush stat may be a good way to evaluate a back's overall efficiency, but it is a poor predictor of what one can expect a back to do on any single attempt. Thus I would like to suggest that the Elias Sports Bureau add a "median yards per rush" statistic. If a back were to have runs of -2, 0, 1, 1, 2, 3, 5, 11, and 70 yards, he would average an impressive 9 yards per rush; but his median yards per rush would only be 2. Likewise, a back with runs of -1, 1, 2, 3, 4, 4, 5, 7, and 11 yards would average 4 yards per rush (much less than 9) and also have a median yards per rush of 4 (much better than 2). Using both the average and median stats would give the coach, broadcaster, or fan a more accurate assessment of a back's skills and strengths.

But the median stat would not be a pure median. I would not include any rush of two yards or less in which the rusher picks up a first down. These plays are designed for short yards in specific situations and should not count against a back's stats. Also, rushes resulting in a touchdown would only figure into the median if they were to raise the median. When a team snaps the ball on its opponent's 1-yard line, a rusher can only pick up a single yard (and only a single yard is needed). This single yard shouldn't count against the rusher's median yards per rush. If a back with a median of 4 ran the ball into the end zone from the 3-yard line, this 3-yard rush would not figure into the median statistic. (A 5-yard rush resulting in a touchdown, however, would.)