Four traditional box score stats that are relatively useless

[Todd Pence]

FIRST DOWNS
Yes, the achievement of getting first downs is essential to maintaining possessions and even scoring. But tracking the total number of first downs achieved by a team can be a misleading statistic in judging the efficiency of their offense.
Consider the following two cases: Team A gets the ball at their own twenty, and embraks on a long sustained drive in which they earn five separate first downs. The drive eventually bogs down in the red zone, and they settle for a field goal.
Team B starts on their own twenty, and on their first series complete a long bomb for a touchdown.
Now, obviously in the end result, team B has been more successful than team A. But in terms of first downs, team A has a whopping five to team B’s mere one!
I’ll punctuate the uselessness of first downs as a stat by pointing out that in the most lopsided NFL game of all time, the Bears’ 73-0 rout of the Redskins in the 1940 title game, both the Bears and Redskins finished with an equal number of first downs (17).

THIRD DOWN CONVERSIONS
A statistic related to first downs. Yes, a team that is better than its opponent in converting third downs has a better chance of winning the game than their opponent. But the problem with the third-down conversion statistic is what it leaves out. In the first place, it doesn’t give credit to the teams who do the job on first and second downs and so don’t put themselves in bad third-down situations. In the second place, the third down statistic completely ignores the times the offense elected to gamble by going for it on fourth down and succeeded, keeping the drive alive. In the box score, each of these instances are recorded as failures! There is no stipulation for a team who made good there third-down shortcomings by converting on fourth down and keeping the drive alive.

FUMBLES THAT AREN’T LOST
Every traditional football box score team stat section has a line reading FUMBLES-LOST. The second number is a vital stat, the first is meaningless. If a team fumbled the ball, and then recovered their own fumble, it is an inconsequential non-event. The non-lost fumbles stat is more insidious at the end of the year indiviual totals than it is in the individual game box scores. In a given year, we might be told that your favortie team’s leading rusher had five fumbles for the year. Yes, but how many of those five actually hurt his team?

TIME OF POSSESSION
Yes, there is a high correlation between the team that leads in TOP and the team that wins. But this is mostly due to the fact that the team leading in the fourth quarter attempts to control the ball more. In other words, TOP is a byproduct of, not a reason for, victory. And of course every year there are plenty of instances of blowouts in which the losing team dominates in TOP simply because the winning team scores so quickly every time it has the ball. I know TOP didn’t exist as a stat in 1916, but I have a sneaking suspicion that in Georgia Tech’s 222-0 win over Cumberland, it was those historic losers who might have won the TOP battle.

[Mark L. Ford]

It's an interesting observation, and I'm surprised that it hasn't generated a lot of responses, considering the number of stats fans we have in the Forum. I disagree with the assessment of relative worth (other than time of possession).

While the final score is the actual important statistic, I think that the stats on first downs and third down conversions raise the more important question of why did a team lose if it had a significant advantage over the other in getting a first down (or having 3rd down convert to 1st down rather than 4th down)? Going for it on 4th down doesn't happen unless a team is losing the game, and it generally doesn't occur until late in the game. Regarding number of overall team fumbles (followed, of course, by how many of those were lost), the first number is an indicator of how good the opposing team was successful in forcing fumbles, and of the possessing team's efficiency in keeping control of the ball.

But I agree, time of possession doesn't mean much-- the extreme example was a Colts/Dolphins game on MNF in 2009. The Dolphins had control for more than 45 of the 60 minutes; the Colts won, 27-23.

[JuggernautJ]

Statistics (like everything else) must be viewed with context in mind.
While it's true that Time of Possession may be less relevant to some teams (those that throw the deep ball more often) it was a hallmark of some ball-control offenses (Bill Parcell's Giants team come to mind here).

So, while one is making broad general statements it must be remembered that the devil, as always, is in the details.