TCU has played much better in the second half of games this year. What other schools have seen dramatic shifts of play after halftime?
13 Oct 2004
Guest Column by Jim Armstrong
Most basketball fans are familiar with the concept of pace. The Dallas Mavericks and Sacramento Kings are well known for their fast paced, running style of play. Other hoops teams have more of a slow-it-down, half court game, for example, any team coached by Mike Fratello. There has been a fair amount of research published on pace in basketball, perhaps most notably by Dean Oliver in his recent book Basketball on Paper.
But what about pace in football? Football Outsiders briefly touched upon this subject last season, but there's been very little written on about the idea. Which offenses push the pace the most, and which tend to crawl down the field with the speed of a turtle? Are there any strategic elements to setting the pace? Does a so-called "ball control" offense really help when a team is overmatched? In this two part series, I will attempt to answer these questions and a few more that come up along the way. Part one will give a definition of pace, look at the fastest and slowest teams, and determine whether pace relates to winning and scoring. Part two will break down pace into greater detail and test the usefulness of some pace-related game strategies.
The first thing that becomes apparent when studying pace is that it can be greatly affected by game situations. A team with a lead late in the game is likely to slow down and grind out the clock as much as possible. And a team down by a couple touchdowns in the fourth quarter is probably running a fast paced no-huddle offense in an attempt to catch up before time runs out. For the purposes of this study, I was not really interested in situation-induced pace, but rather, I wanted to examine each team's pace as dictated by its game plan or style of play. So I decided to focus on pace in the first half of games only. This should eliminate most game situations where a team alters its pace due to the score and time remaining, yet still leave a reasonable sample size of data with which to make conclusions. From this point on, unless otherwise noted, when I refer to "pace" I am really talking about first half pace. Also, some writers and broadcasters use the term "tempo" rather than pace. To me, the two terms are synonyms in the context of football. I simply chose to use "pace".
Pace is generally thought of in terms of possessions, or drives, as I will use interchangeably here. So my data set for this study is the collection of drive charts from all regular season NFL games from 1998-2003. After throwing out obvious take-a-knee drives, this data set includes over 17,000 first half drives. The drive charts list the number of plays and time of possession for each drive. Thus, the definition of pace used here is time (in seconds) per play. A high value indicates slow pace, whereas a low value indicates fast pace.
Before we look at some numbers, there are limitations of this definition that should be mentioned. There are several ways an offense can control the pace of the game. In his book Football Clock Management (reviewed here), John T. Reed listed twelve ways an offense can work the clock, but in general they fall into three categories: how much of the play clock is used before the next snap (if the game clock is running), play calling (in particular, passing plays are more likely to result in stoppages of the game clock), and what the players do during the play (for example, run out of bounds or commit penalties). There is definitely some overlap among these three. Thus, it is probably difficult to separate completely the effects each of these have on pace under our definition, but is is also not clear how important it is to do so.
So which offenses were the fastest and slowest in 2003? Here are the rankings of all 32 teams, from the slowest to the fastest:
2003 Offenses by Seconds per Play,
It's not too surprising to see Tennessee among the slowest, as they are known for their grind-it-out style, particularly during their Eddie George three-yards-and-a-cloud-of-dust running game era. In fact, Tennessee ranked among the top four slowest the past five years in a row, and in 1998 they were 7th. Cleveland is another team that pops up repeatedly near the top of this list. They have ranked among the top seven slowest teams four years running. The Jets have also been a slow team under Herman Edwards, ranking 6th slowest in 2002 and 10th slowest in 2001 after being the fastest team in the league in 2000 under Al Groh. But most teams do not show as much consistency at either end of the rankings. For example, Denver was 3rd slowest in 2003, but 5th fastest in 2002, and all over the board prior to then, despite many years of consistency in the coaching staff.
On the flip side, Seattle has ranked among the fastest paced teams in all five years under Mike Holmgren. Bill Walsh taught Holmgren how to script the first fifteen plays of the game, and apparently that script also dictates an aggressively paced offense. Philadelphia also shows up in the fast lane during every year of Andy Reid's tenure. Of course, Reid studied the West Coast Offense under Holmgren in Green Bay, but curiously, the 1998 Packers were an average paced team, ranking 14th.
I was not surprised to see St. Louis as a fast paced team in 2003, but Mike Martz's offense didn't show up as consistently fast as the Seahawks or Eagles did. Only three of the six years in this study were they among the ten fastest teams.
Looking at the list, however, there doesn't seem to be any apparent connection to the success of the team. Some good teams are fast and some are slow. We can measure this quantitatively by computing the correlation coefficient of a team's pace to winning. I've also figured the correlation of pace to various drive stats during the entire six year period, shown below. Keep in mind that the drive stats used are season averages for the first half of games only, since we're only talking about first half pace. But of course, wins is determined by looking at the score at the end of the game.
|Correlation with pace, 1998-2003|
Recall that a correlation coefficient near zero indicates no correlation, while a value close to 1 or -1 indicates strong correlation. As the numbers show, there's seemingly no obvious relationship between a team's pace and winning, moving the ball, or scoring points. The only values that show any indication of correlation are those relating turnovers, suggesting that perhaps a faster (lower) paced offense may be more susceptible to turning the ball over.
But there is another possible explanation involving turnovers, as we'll see when we break down the numbers further in part two of this series. We'll also look at some strategy aspects of pace in part two, particularly whether it makes sense to employ a slower, "ball control" offense right off the bat against a more powerful team, as some commentators have suggested.
Jim Armstrong wrote about consistency in turnovers back in April, and is now contributing his drive stats to our site each week. If you have questions on this article you would like to see answered in part two, feel free to post them in the discussion thread. If you are interested in writing a guest column, something that takes a new angle on the NFL, please email us your idea at email@example.com.
6 comments, Last at 08 Jan 2008, 2:52pm by DavidH