The Fantasy Football Historian: Evolution of the NFL
So far in the Fantasy Historian series we’ve focused on epic performances, both from individual players and entire lineups. In this article, we study how the game has changed. This discussion will center around the evolution of the NFL with, as always, the resulting fantasy implications.
Data, unless otherwise specified, comes from nflfastR.
Let’s start with some Kirk Goldsberry-esque cartography charts to visualize the current state of NFL aerial attacks. The coordinate data is thanks to Ethan Douglas and Sarah Mallepalle et al. (you can find the data here). Base code for the charts is from Thomas Mock.
The below visuals show the distribution of ‘hot spots’ for NFL incompletions and non-TD completions in 2020. The redder the area, the more passes in that spot. The line of scrimmage is shown in blue at the ‘zero-yard line’, and all passes are relative to that location. Of course, in real life, these passes occurred at lines of scrimmage all over the field.
Note that completions did occur further than 20 yards from the line of scrimmage, there just wasn’t enough density to show up on this chart (less than 5% of completions).
Let’s take a look:
- What first strikes me about these charts is the symmetry. Although there is slightly more density on the right side of the field, quarterbacks overall don’t seem to strongly prefer throwing to their right side. This is in stark contrast to earlier years like 2017, which show a significant right-handed bias (chart not shown). This trend isn’t because there are more lefties taking snaps – Tua Tagovailoa is the only name that comes to mind – meaning that there has been a conscious shift towards symmetry in recent years.
- We can immediately see the ‘hot spots’ in the completion chart: the ‘peak’ of the data is about 5 yards downfield, slightly outside of the hashes on both the right and left side. In general, the area within about a five-yard radius of the pocket – including the ‘checkdown’ areas behind the line of scrimmage – are quite hot. Density drops off quickly after 10 yards.
- Unsurprisingly, incompletions are, on average, much further down the field: these are simply harder throws to make! Note that there isn’t a lot of density in the ‘checkdown’ areas for the incompletions chart. These are very, very high completion percentage passes, and there simply aren’t many incompletions in these areas. This in contrast to the ‘peak zones’ from the completion chart (around 5 yards downfield, slightly outside of the hashes): there are a lot of completions and incompletions in these zones!
Overall, to complement these charts, it’s important to remember that teams have been getting much better at passing. Glancing at the annual passing stats from Pro Football Reference shows how basically every major statistic (Att, Cmp, Cmp %, Yds, TD, INT) has improved over the past few years.
Indeed, possibly because passing has gotten better, teams are deciding to do it more often. Pass probability (% of plays that are pass plays) has slowly but steadily increased, paced by the total number of plays per game. These two factors combined mean, simply put, more passing plays overall! All told, this data implies that pass-catching RBs are more valuable than ever.
A perhaps surprising trend is the percentage of passes out of the shotgun formation. This has risen sharply since the early 2000s, and is now sitting above 80%. Just two decades ago, the probability was flipped: there was just about a 20% chance of shotgun formation on a passing play.
Running Back Committees
The dream of every fantasy manager is to roster a ‘bell-cow back’: the type of RB that dominates his team’s backfield. Of course, NFL head coaches and offensive coordinators aren’t playing fantasy football, and the ‘backfield by committee’ is a common rushing approach.
In this set-up, teams spread out their carries and targets to multiple running backs; often, one back will specialize on first downs, another on third downs, another as a ‘change of pace’ RB, etc. To put it simply, this can make fantasy football difficult. More ‘mouths to feed’ means that production is more opaque and generally harder to predict (and diluted across different players).
Unfortunately, committees are on the rise! Check out this chart, as well as some discussion points, below:
- The yardage for the average ‘leading rusher’ on an NFL team has decreased noticeably, while the second and third leading rushers are, on average, delivering more production. This makes dominant rushers today (like Derrick Henry) more valuable, and also confirms the trend that backfield committees have become the rule instead of the exception.
- In 2000, for example, the average leading rusher on a team ran for 1007 yards in a season. In 2020, that number was just 799! That decrease is almost completely mirrored by an increase in the production of a team’s second leading rusher: 398 yards on average per season in 2020 versus 201 yards in 2000.
- ‘Third leading rushers’ have tripled production from 55 (!!) yards a season to 168 yards. Unfortunately, this is a sort of ‘dead zone’: it doesn’t really make them fantasy relevant at 10 yards per game, and yet takes points away from more prominent rushers!
Size and Speed
Given the physical nature of the sport, one of the best ways to track the trends of football is to analyze the physical traits of the players themselves.
In this section, we visualize player height and weight, measured at the weekly level and aggregated to a season-long average, among players who scored 3 fantasy points or more (standard scoring). Let’s start with height:
- The QB trend is funny: an increase up until around 2016 and then a sharp decrease. Naturally, this is from taller QBs retiring (Peyton Manning, standing at 77 inches, hung up his cleats in 2016) and shorter QBs entering the league (Kyler Murray is 70 inches tall).
- Running backs have gotten shorter (by about .6 inches) while tight ends have gotten taller (by around .8 inches). Wide receivers haven’t changed that much; note how tight the scale on the y-axis is (gap of .2 inches).
The moves have been more significant – and aligned – in weights:
- Is everyone on Tom Brady‘s “Plant Man” diet or what? All of these positions have gotten noticeably lighter in the past decade. QBs are off from the 2010 peak by about 6 pounds, while RBs are down over 3.5 pounds on average since 2000. Tight ends have shaved off about 4 pounds, with wide receivers about 5 pounds from their peak in 2014. All of these moves make sense: the league has transitioned to a faster, quicker, more mobile style of play.
- For visual reference, Devonta Smith‘s famous weigh-in showed him at 166 pounds, which is far below the dots in the WR chart. I only see twelve wideouts in the dataset since 1999 at 166 pounds or lighter; the two most notable are Taylor Gabriel and J.J. Nelson.
- It’s funny to note that running backs, while a couple of inches shorter than wide receivers on average are over ten pounds heavier.
Let’s now consider actual fantasy scoring (Half PPR) since 1999.
- All of the major positions have been increasing in point totals over the past two decades. This is good news for fantasy: more points means more fun!
- Are you surprised that WRs are so high? Remember that wideouts essentially have all of the same yards and touchdowns as quarterbacks, but receive more points for them. Indeed, the QB and WR lines are highly correlated, and have been growing even faster over the past few years. Indeed, we’ve already seen in this article how the NFL has been transitioning to more pass-heavy approaches.
- Sadly, to the chagrin of Kyle Juszczyk, fullbacks have been used less and less in a fantasy capacity.
Kickers Have Feelings Too
I know what you and Kyle are thinking: where is all of the kicking analysis? Fear not!
Like passing, NFL teams have gotten much better at kicking. Teams have been sending it more often and from further out: 3 yards deeper on average over the past two decades. Kickers have stepped up to the plate, delivering more accurate performances and being blocked far less.
Any other trends you’d like me to investigate? See anything that looks off? Message me on Twitter.