When Jason Garrett was asked a few years back which stats the Cowboys pay the most attention to, he answered turnover differential and 'winning the fourth quarter,' because he believes they are the two stats most significantly correlated with winning in the NFL.
We looked at turnover differential in a post last week, and today we take a look at 'winning the fourth quarter.' Here's Garrett's exact quote:
There were a lot of games in the NFL that were within one score in the fourth quarter, and the teams that went to the playoffs won a lot of those games. Teams that did not go to the playoffs lost a lot of those games. It's really important to be good in the fourth quarter.
Most of the time when you read or hear about close games, the stat you'll be presented with is the final score. So let's take a look at that most obvious stat of all. Last season, 65 of 256 games (25%) were decided by three points or less, and 123 of 256 games (48%) were decided by seven points or less. Garrett argues that playoff-bound teams did better in those games decided by a small margin, so let's see if that statement holds water. The table below shows the W/L records for close games split by playoff and non-playoff teams, and compares their records in close games against their overall record.
|2013 regular season||Games decided
by 3 points or less (65 of 256)
by 7 points or less (123 of 256)
|Non Playoff teams||35-49-1||.412||72-88-1||.447||122-197-1||.381|
And as you can see, there's really isn't that much to see here. The playoff teams in 2013 had largely the same win percentage regardless of whether the games were close or not. Same picture for non-playoff teams, they generally struggled to win games, whether those games were close or not.
But that's not the precise metric Garrett was talking about. He was talking about games that were within one score at some point during the fourth quarter. And by looking at that specific number, Garrett highlights a key dynamic about NFL games: on average, scores in the NFL are much closer during the fourth quarter than the final score indicates.
Let's take a look at the data: While only a quarter (65 of 256) of last year's regular season games were decided by three points or less, at one point during the fourth quarter, the score differential in almost half of all games NFL games was three points or less (126 of 256). Two thirds of all games were within seven points at one point in the fourth quarter (166 of 256). Here's the full breakdown.
|Games decided by one score, 2013
||Games within one score
at any point in the 4th quarter, 2013
||Number of Games||Percent||Point Difference||Number of Games||Percent|
|8 or Fewer||130 of 256||51%||8 or Fewer||174 of 256||68%|
|7 or Fewer||122 of 256||48%||7 or Fewer||168 of 256||66%|
|3 or Fewer||64 of 256||25%||3 or Fewer||126 of 256||49%|
Allow me to repeat what I just wrote, for full effect: Half of all NFL games last season were within three points at some point in the fourth quarter. And Garrett is right in pointing out that playoff teams do better in those games: Playoff teams had a .671 winning percentage (57-27-1) in these "three-point" games, non-playoff teams had a .407 winning percentage (68-98-1). Those percentages are in line with what we saw above, which just reinforces the point that good teams find ways to win and bad teams find ways to lose, regardless the circumstance.
I once heard someone say that you could significantly simplify the NBA by having teams start with the score tied at 100 - 100 and then let the teams decide the winner in the next five minutes. At the time, I thought that was pretty funny, but I find it a lot less funny when applied to the NFL: Play just one quarter and give one team a three-point advantage - half the games last year were exactly that.
Back to the Cowboys: Garrett is emphasizing fourth quarter performance for a very good reason. You may find this hard to believe given how we're constantly being told that the Cowboys were seemingly overmatched in many games over the last three years, but in 29 of the 48 games the Cowboys played over the last three seasons, the difference in points between the Cowboys and their opponents at one point during the fourth quarter was three points or less. That's sixty percent, or almost two thirds, of all games that were there for the taking for the Cowboys. And one of the reasons the Cowboys finished all three seasons the way they did is because their record in those 29 "three-point" games was a disappointing 13-16.
Last year, the Cowboys played in seven games decided by three points or less (in which they recorded a dismal 2-5 record), but they played 10 games in which the difference in points between the Cowboys and their opponents was three points or less at one point during the fourth quarter. The Cowboys' record in those close games: 4-6. Here's a summary of those ten games.
|Week||Opponent||Closest 4th Quarter Score||Final Score||Result|
It's not hard to imagine that even a slight improvement in your fourth quarter performance could have a pretty strong impact on your overall W/L record, and for the Cowboys this year, winning the fourth quarter will be the difference between making the playoffs and another top 15 pick.
The Cowboys kept ten games very close up until the fourth quarter last year. In the end, they lost six of those games. And when you think back to those games, the difference between winning and losing usually came down to one play the Cowboys made or didn't make.
It will not take a top ten defense to win the fourth quarter. It will not take a record-breaking offense to win the fourth quarter. But it will take better playcalling and better execution to win the fourth quarter.
It is clear why fourth quarter performance is Garrett's focus. Improved play here in 2014 on offense, defense, and special teams could easily mean a two- or three-game swing in the Cowboys' final W/L record. And you know what that means: