The Sophomore Slump: Myth or Fact? NFL Edition

Sophslump1

The Sophomore Slump: Myth or Fact? NFL Edition

Plus: Bold Stat Predictions for RGIII, Luck, Wilson, Weeden, and Tannehill in 2013

If you are an NFL fan you know how a trio of rookie quarterbacks, RGIII, Andrew Luck, and Russell Wilson, seemingly took the league by storm this past season. However, on the dawn of their second seasons, many people are predicting a ‘sophomore slump’ for these electric young QBs. “Remember Cam Newton!” they will preach, trying to draw connections between last year’s rookies and the similarly explosive Newton. However, did Newton really have a worse sophomore year? For that matter, do most rookie quarterbacks undergo a sophomore slump like many people seem to believe? In this article I will try to get to the bottom of the sophomore slump myth, examine the curious case of Matthew Stafford’s first 35 games in the NFL, and make some predictions for the numbers that will be put up by last year’s trio of rookie stars as well as the other two rookies (Brandon Weeden and Ryan Tannehill) who will try to avoid the fabled sophomore slump.

Introduction

The term ‘sophomore slump’ itself can cause problems because it is never really defined, and what can be categorized as a ‘slump’ may vary by person. If a rookie has a particularly fantastic year, does a statistically worse sophomore season classify as a slump if the player still is well above the average at their position? This is an issue with no real answer, but the results actually make this problem mostly relevant only to Luck, Wilson and RGIII and will be discussed later when I try to make predictions for their sophomore seasons. To begin my research, I had to find the QBs who played in both their rookie and sophomore years and met a minimum number of pass attempts. The minimum number I selected was 224 passes, because this is the number that the NFL deems the minimum to qualify for certain efficiency ratings (it works out to 14 attempts per game over the 16 game season). I found 43 quarterbacks from 2011 to 1990 that met this number in their rookie year; however, only 31 (Excluding Matthew Stafford, more on him later) of these 43 met this minimum again in their sophomore year. Almost all of those excluded were those whose poor rookie seasons robbed them of a chance to start again.

Method

Using the ever-helpful pro-football-reference.com, I recorded a variety of stats for these 31 players in their rookie and sophomore seasons. While all of these rookies met the league minimum number of attempts for passes, many of them still didn’t play in all 16 games, usually due to an injury (like Michael Vick giving way to Nick Foles) or poor play from the player ahead of them. This means that the stats that are reliant upon the number of passing attempts varied significantly between the rookies (if you throw twice as many passes as another QB, you have twice as many opportunities to throw a touchdown or interception). As a result, rate stats hold the most weight in my analysis: Completion%, TD%, INT%, Yards per Attempt, Adjusted Yards per Attempt (AY/A), and RATE+. Completion% and Yards/Attempt are pretty self-explanatory, but some of the others might need a bit of clarification:

  • TD%: (Passing TDs) / (passes attempted)
  • INT%: (Passing INTs) / (passes attempted)
  • AY/A: yards per attempts adjusted to account for the bonus of TD passes and the drawback of interceptions
  • RATE+: a more advanced type of quarterback rating, 100 is average

While these stats are the most useful in this analysis, I also recorded several other, more traditional, stats for the QBs: TDs, INTs, Yards, Yards/Game, Yards/Completion, and Longest Completion. All of these stats were then complied into three columns: Rookie Year, Sophomore Year, and Change Between Years. I began with all 43 QB’s who qualified their rookie year, but if a QB didn’t qualify in his sophomore effort I removed his data and made a note that he didn’t meet the minimum attempts both years. The color scale in each of the 3 groups allows us to easily identify the best and worst player in each stat for each year and who made the greatest jump between the years. The deepest green is the best of the column while the deepest red is the worst performer, with varying shades in between. After calculating the change in every single stat for each of the quarterbacks between their rookie and sophomore seasons I added up these changes to see what the overall changes were; if the changes were mostly negative (i.e. fewer TDs, a higher INT%) then that would be evidence of a sophomore slump.

Sophslump2

ABOVE: A section of the collected data. Sophomore stats for last year’s rookies are predictions and will be discussed later.

Results

Out of the 12 stats that I tracked, only 2 displayed a net decline. Even more surprising was that these two stats, LNG and Y/C, are stats that say little about the competency of a quarterback and relate much more to play calling and receiver quality. The stats that are most valuable, Comp%, TD%, INT%, Y/A, AY/A, and RATE+ all experienced an overall improvement when the changes of each player were added together. Next, I used these total changes to find the average change per player, and then compared these numbers to the rookie numbers to find the percent difference from rookie season to sophomore season, which I find to be the most telling numbers.

Cmp%

Yds

TD

TD%

Int

Int%

Lng

Y/A

AY/A

Y/C

Y/G

RATE+

5.50%

17.72%

24.15%

7.57%

-9.24%

-20.60%

-3.47%

2.75%

10.82%

-2.48%

13.26%

10.34%

[Note: These are % differences from the average rookie numbers, so a 7.5% increase in TD% does not mean the TD% went up by 7.5 percentage points, but rather that it saw an increase of 7.5% (from 3.58% to 3.85%)]

The above table represents the average percent changes of various stats for the 31 quarterbacks from their rookie to sophomore years. On average, quarterbacks improved in nearly every statistical category, and the two decreases in performance (Lng and Y/C) are two stats that are not a huge indicator of a quarterback’s skill.

Analysis

Of course, these averages do not mean that every single one of these quarterbacks improved in each of these skills, but overall, rookies seemed to improve their all-around performance in their sophomore years. While few quarterbacks saw increases in every single stat, the RATE+ stat attempts to give an overall indication of performance, and only 3 quarterbacks saw a decrease in their RATE+ in their sophomore year. Interestingly enough, one of the huge changes was the 20.6% drop in the INT% of the quarterbacks. This indicates that sophomore quarterbacks clearly benefitted from a full year of NFL games and practice, making smarter decisions leading to far fewer interceptions. Obviously, all of the improvements imply quarterbacks gain better command and comfort in their second year, completing more passes of higher quality, fewer interceptions and more touchdowns. Another interesting thing to note is that Y/A increased, which could result from several factors. One is just as mentioned before: comfort, more practice, and improved confidence; however, it could also be a sign of play calling. Another general belief in the NFL is that offensive coordinators often put restraints on their young quarterbacks, focusing on short passes and easy reads, and then once they become comfortable they open up the playbook to more deep passes and complex plays, which could partially account for the rise in yards per attempt in general for these quarterbacks in their sophomore season. Overall, it seems clear that, in general, rookie quarterbacks do not experience any slump in their sophomore year; rather, on average, they seem to improve in nearly every important statistical category.

Matthew Stafford

As mentioned before, in the course of my research, Matthew Stafford stood out as a bit of an anomaly. While Stafford easily qualified in his rookie year, an injury cut his second season short after just 3 games and 96 passing attempts, less than half of the 224 attempts needed to qualify. The obvious response was to throw out Stafford’s numbers and move on to the next QB, but I felt compelled to look into his situation a bit more because he was the only quarterback who failed to qualify with so few passes due to a major injury. I wondered if his third season could, in a sense, be considered his sophomore year. Obviously his situation was unique, so I removed his data from all computations and averages, but I kept his rookie stats and compared them to his numbers in his third season. While his rookie year was just about on par with the averages of the other 31 rookies (a bit below in many stats in fact), his third year was outstanding, as he posted the biggest improvements from his rookie to “sophomore” season in 5 of the 12 stats measured, throwing 28 more touchdowns, increasing his RATE+ by 44 points (from a significantly below average rating to well above it), and more than doubling his passing yardage to 5038 yards. Stafford had barely any more game time than the rest of these quarterbacks, but his numbers were clearly some of the best. This seems to suggest just how important time off the field is for young quarterbacks in their development. Stafford’s unique situation adds to the conclusion that more time generally helps young quarterbacks.

Predictions

Predictions for Tannehill, Weeden, and even Luck to an extent, don’t seem to be too difficult. All of these players’ rookie seasons were very close to the average for the other 31 rookies I looked at, therefore I applied the average percent difference to these players to come up with their predictions (note: Luck threw a lot of passes, and so I am less confident in my predictions for his stats that do not account for number of attempts). However, when it comes to Wilson and RGIII, I am faced with a problem; both of these QB’s performed at such a high level last year that I find it unlikely that they would see the same growth in stats as the other rookies (for example, does it seem likely that RGIII would be able to complete 69% of his passes this year?). However, the purpose of this article is not to delve deep into regression to best predict these stats, so to make things easy, I decided to look at individual QBs and see if I could find any similar cases: an extremely impressive rookie season followed by less growth than the average. One QB who I found to use as my model is Ben Roethlisberger, whose excellent rookie season was followed up with a great sophomore season, but due to how successful he already was in his first year, he experienced less growth (or in most cases, essentially no growth) than the average rookie. Keep in mind that this shouldn’t be taken as a “slump”, but rather as a sign of how good these two rookies were in their first season; there simply is little room for them to experience the same jumps that other rookies had. Below is a snapshot of my rough predictions for this year’s sophomores; note that RGIII and Wilson still are mostly deep green, indicating that even with only a modest increase in performance they will still be among the best.

 Sophslump3

Conclusion

In conclusion, the numbers clearly indicate that the sophomore slump is mostly a myth. Sure, some quarterbacks see a decline in many stats in their sophomore years (see: Sam Bradford, Matt Ryan), but that number is dwarfed by the number and extent of which rookie quarterbacks tend to improve in their sophomore year. Even QBs who improved may have seen a decrease in some areas, but overall they generally improved, a statement supported by the RATE+ stat, which attempts to measure overall performance. Only 3 of the 31 quarterbacks analyzed actually experienced an overall decrease in RATE+ in their sophomore year. The unique case of Matthew Stafford’s tremendous 3rd season also provides evidence that more time in the NFL can dramatically improve a player’s performance, even if games aren’t played like in Stafford’s case (also think of Aaron Rodgers learning under Brett Favre). As for this year’s rookie class, I expect slight improvements in the stats of Weeden, Tannehill, and even Luck, although to a lesser extent in his case. I have to expect some natural regression in the case of Wilson and especially RGIII following his injury, but I don’t think such regression should be referred to as a slump. Overall, I think it is safe to say that in general the idea of a sophomore slump is just a myth and that, on average, most quarterbacks will see improvement in their stats rather than decline.

Max Roberts

Georgetown University Class of 2016

Follow Max on Twitter: @Maxroberts611
Follow GSABR on Twitter: @GtownSports
Like GSABR on Facebook

Data courtesy of:

http://www.pro-football-reference.com/

http://www.footballperspective.com/comparing-rookie-quarterbacks-to-their-teams-prior-years-stats/

About these ads

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 424 other followers

%d bloggers like this: