How Offense Has Changed

You are viewing free content provided by Why not consider subscribing today?

by Mike Horn, Statistical Analyst

Published, 8/2/14 

This is not so much about fantasy football as it is just plain football. I’m working on several things to do with coaches’ offensive identity, such as the Norv Turner article. To put them in context, I threw together some simple graphs.
This first chart shows total plays by offenses in three periods: the beginning of the full free agency era from 1994-2001, the several seasons after the expansion to 32 teams (2002-2008), and the last five years (2009-2013). These time periods are completely arbitrary. The number of wins by individual teams is across the X-axis and the number of plays each team ran is on the Y-axis.
Rather than show all the individual team plots, which would be a big jumble, I graphed (and by this I mean “Excel calculated”) the best straight line “fit” of wins vs. plays.
What we see is, regardless of era, winning teams typically run more plays than losing teams. We can also see that from 1994-2001 teams were more “up-tempo” than they were in the following period from 2002-2008. In the last several years, losing teams have gotten off more plays than in either of the previous two eras, but winning teams are right about where they were 10+ years ago.
Obviously, this is a broad-brush description and, for the statistically inclined, the R2 values of these curves are pretty low. Which means that the relationship between winning and number of plays is not strong – in other words, “other factors” are more important. And if you look at the blue line, it has a lower slope and does not ascend as fast. So the difference in the number of plays – which is connected to fantasy opportunity – is not as great between winning and losing teams as it used to be. So while there may be more fantasy production on winning teams, that may not be as important as all those “other factors” either. The difference between a 6 and a 10-win team is around 2-3 plays a game – one or two of which is often a kneel-down that has no (or worse, negative) fantasy value.
As you might expect, the number of plays an offense runs correlates (somewhat) with how often those plays are passes – the clock stops on around 40% of all passing plays. That’s why winning teams tend to rush more – they want to kill the clock. The “Pct Rushes” shows that trend.
Not surprisingly, the percent of running plays was higher in the 2002-2008 period; that’s when we saw the lowest number of total plays. The 1994-2001 period had a lower percentage of rushing plays across all win-levels, but the relationship between “Pct Rush” and wins was almost exactly the same in the next period – the red line parallels the black line, just at a higher level. Since 2009, not only is a smaller share of plays rushes, the relationship between the percentage of rushing plays and wins is the weakest of the three periods. I think this indicates that teams are more likely to keep throwing when ahead than they used to be.
If you put the previous two charts together, you get the next two:
Notice there was almost no difference in the total number of rushes, on average, by teams in the two older periods. Bad teams are slightly more likely to run the ball now than they were, and good teams are somewhat less likely to do so. Remember, this is on the order of 1-3 plays per game difference, so these are not huge changes on the game level. 30 more rushes per year, if 60-70% of them go to a team’s RB1, translates to 80-100 yards (20 extra carries at 4-5 yards a pop) or 8-10 total fantasy points. So the shift in offensive trends has not been dramatic in affecting RB scoring – not nearly as dramatic as the effect of more RBBC situations in the league. From 2002-2008, there was an average of nine RBs a year with 300+ carries. Over the last five years, that is down to around 4.5. Last year, there were two. Fewer total rushes at the team level matter, but it’s marginal compared to how those rushes are distributed.
And of course Total Passes are up, and that has been increasing in each time period. Again, notice how much flatter the decline of the blue line is compared to the other two. Said a different way, these lines imply garbage time production in the passing game is not as big a bonus as it used to be before 2009.
Two last graphs which show how the average number of plays and run-pass mix has evolved since 2002:
2008 and 2007 stand out as anomalies on the first and second charts respectively. But the trends are obvious, and probably no surprise.
As I said at the beginning, this article doesn’t have a lot of fantasy insight other than to say the relationship between wins and total number of plays, run or pass, is not a highly significant factor in affecting fantasy production. It makes a difference, particularly if a team goes from 2 to 11 wins like the Chiefs from 2012-2013, but I think the overall landscape of the league does not matter nearly as much as understanding what individual coaches are going to do with their offenses. I also think that what coaches do in the red zone is particularly important for fantasy, and I want to explore that more in my next article.


3,609 people think it's Tebow time.

Back to the top