Writers of Pro Football Prospectus 2008

Most Recent FO Features

owie.jpg

» Turf Type and NFL Injuries: Part II

In our second look at turf and injuries, we look at specific stadiums, with some surprising results. It turns out the turf in Washington isn't nearly as dangerous as you may have been led to believe.

20 Apr 2017

Snap-Weighted Age: 2016 NFL Rosters

by Scott Kacsmar

Welcome to the annual internal struggle where we try to quantify age for NFL teams from this past season. Actually, the quantification part is quite simple. By accounting for how many snaps each player played in the regular season, we are able to give each team a Snap-Weighted Age (SWA).

The difficult part is determing what these numbers mean. Should a team want to rank very high (older, experienced) or very low (younger, potential) in SWA? In 2016, the seven oldest teams all failed to make the playoffs, though Atlanta (eighth) and New England (10th) were in the top 10 on their way to Super Bowl LI. That was the second year in a row where both Super Bowl teams were in the top 10 in SWA.

Meanwhile, the four youngest teams failed to make the playoffs. The Jaguars (Gus Bradley) and Rams (Jeff Fisher) fired their head coaches during the season after years of boasting young, talented rosters that could never get to .500 when push came to shove.

The trend towards younger players continued last season. The league-wide SWA for 2016 was 26.5 years old, the youngest yet for any season in our database, which goes back to 2006. Things have slowly sunk from an average SWA of 27.2 in 2006. There are a number of factors including more early retirements, fewer veterans in their 30s, and an eagerness to start rookies.

In fact, we wanted to check that last part with numbers from our master database. We looked at games started by rookies for each season since 2002, which has neatly featured 11,264 starts every season thanks to a consistent setup of 32 teams and 16-game seasons. We broke down the starters by offense and defense, as well as the percentage of total starts by rookies in the league that season.

NFL Games Started By Rookies (2002-2016 Regular Season)
Year SWA OFF Rookie GS DEF Rookie GS TOT Rookie GS %NFL Starts
2002 - 451 379 830 7.4%
2003 - 424 370 794 7.0%
2004 - 392 478 870 7.7%
2005 - 402 411 813 7.2%
2006 27.2 440 455 895 7.9%
2007 27.2 383 368 751 6.7%
2008 27.2 406 303 709 6.3%
2009 27.1 380 403 783 7.0%
2010 27.0 455 405 860 7.6%
2011 26.9 450 437 887 7.9%
2012 26.8 557 443 1,000 8.9%
2013 26.7 531 502 1,033 9.2%
2014 26.7 655 339 994 8.8%
2015 26.6 543 452 995 8.8%
2016 26.5 491 562 1,053 9.3%

In 2012, the second year under the new CBA, rookie starts hit 1,000 for the first time in this era, and have remained close to that figure in the years since, peaking at 1,053 rookie starters in 2016. Last season also had the most defensive rookie starts at 562, which might help to explain why the season didn't seem to feature any great defenses. With nearly 10 percent of all starts going to rookies, 2016 was the youngest season we have tracked yet.

2016 Snap-Weighted Age: By Unit

The following table shows SWA for the overall team (TOT) along with the unit breakdown for offense, defense, and special teams. We ranked the units from oldest to youngest, but that is really done for reader convenience rather than making any definitive statement of quality.

Team TOT SWA Rk OFF SWA Rk DEF SWA Rk ST SWA Rk
ARI 26.8 9 28.3 1 25.9 29 25.5 27
ATL 26.9 8 27.8 5 26.0 25 27.3 2
BAL 27.4 2 28.0 3 27.2 5 26.1 18
BUF 27.0 6 26.6 19 27.2 6 27.3 1
CAR 26.6 16 27.0 12 26.2 23 26.6 6
CHI 26.3 23 26.7 16 26.0 24 26.2 11
CIN 27.3 3 26.7 17 28.2 1 26.4 7
CLE 25.7 31 26.7 15 25.1 32 24.5 32
DAL 26.4 21 26.6 20 26.3 20 26.1 16
DEN 26.4 22 26.6 18 26.7 12 25.1 30
DET 26.2 24 26.2 25 26.4 18 25.9 21
GB 26.2 25 26.8 14 25.8 30 25.4 28
HOU 26.1 27 25.7 30 26.5 17 26.2 12
IND 26.7 12 25.8 27 28.0 2 25.9 20
JAC 25.7 30 25.6 31 25.9 28 25.5 26
KC 26.0 28 25.9 26 26.5 16 25.1 31
Team TOT SWA Rk OFF SWA Rk DEF SWA Rk ST SWA Rk
LARM 25.7 32 25.5 32 26.0 26 25.4 29
MIA 26.4 20 26.3 23 26.9 10 25.6 25
MIN 27.2 4 27.1 10 27.8 3 26.3 10
NE 26.8 10 27.3 7 26.6 14 26.3 9
NO 27.5 1 28.3 2 26.8 11 27.0 4
NYG 26.1 26 26.3 22 25.9 27 26.2 13
NYJ 26.8 11 27.5 6 26.4 19 26.0 19
OAK 26.6 14 26.5 21 26.6 13 26.9 5
PHI 27.0 7 27.0 11 26.9 9 27.0 3
PIT 26.6 15 27.1 9 26.5 15 25.7 23
SD 26.6 13 27.9 4 25.7 31 25.7 24
SEA 26.4 19 25.7 29 27.2 7 26.4 8
SF 26.5 18 27.0 13 26.2 22 26.1 17
TB 25.9 29 25.7 28 26.2 21 25.8 22
TEN 26.5 17 26.2 24 27.0 8 26.2 14
WAS 27.1 5 27.1 8 27.4 4 26.1 15
NFL 26.5 - 26.7 - 26.6 - 26.1 -

There is consistently a non-zero, positive correlation between SWA and DVOA. In 2016, that correlation coefficient was 0.38, which was up from 2015's mark of 0.27. Not strong at all, but not zero either. We typically observe practically zero correlation for special teams DVOA and SWA, and that was the case again this season (0.05). However, while the correlation is usually stronger on defense than it is on offense, this year it was minute at 0.03, compared to 0.30 on offense. That 0.30 mark is a result very typical of past seasons.

For the second time in three seasons, the Saints were the oldest team in the league, led by a 37-year-old Drew Brees, who continues to compile 5,000-yard passing seasons. Not even salary-cap hell could give the Saints a youthful appearance. Now the team has traded away Brandin Cooks, its second-youngest receiver, this offseason.

As I predicted last year, Arizona fielded the league's oldest offense thanks in large part to 37-year-old Carson Palmer and 33-year-old Larry Fitzgerald. The latter was still going strong last year, leading the team with 150 targets, or more than double the next closest wideout on the team (J.J. Nelson, 74). The Cardinals have often been linked to a first-round wide receiver in this year's draft. No team had a larger split between the SWA of offense and defense than Arizona, which fielded the fourth-youngest defense in 2016. San Diego was close with the fourth-oldest offense and second-youngest defense, but we know the Chargers also employ an ancient quarterback-to-receiver combo in Philip Rivers (35) and Antonio Gates (36). That young defense could be really good in 2017 should they keep players such as Joey Bosa, Denzel Perryman, and Jason Verrett healthy.

The Colts and Seahawks have taken SWA imbalance in the other direction. Andrew Luck and Russell Wilson are the centerpieces of very young offenses on teams that feature older defenses. The Colts have finally moved on from general manager Ryan Grigson, who built the oldest defenses in 2014 and 2015 after neglecting the unit in the draft for years. The Colts only finished as the second-oldest defense in 2016, getting dethroned by the Bengals. Cincinnati is not a favorite to lead the league in defensive SWA again next year after letting Domata Peko (32) go to Denver this offseason. Adam Jones (34 in September) is the oldest remaining defender, but the Bengals really need to see something out of former first-round cornerbacks Darqueze Dennard and William Jackson III (who spent his rookie year on injured reserve).

We've often talked about the young talent of the Rams and Jaguars, but the on-field results have simply not matched the hype. The Rams have ranked 32nd, 32nd, 31st, and 32nd in SWA in the last four years without breaking .500. They even managed to bring their coaching staff's age down with the hiring of Sean McVay, who turned 31 in January. He has his work cut out for him getting an offense with high draft picks Jared Goff, Todd Gurley, Greg Robinson, and Tavon Austin to start producing. Jacksonville has climbed from 32nd to 31st to 30th in SWA over the last three years, but Gus Bradley was shown the door after compiling a 14-48 record. The defense began to show signs of improvement last year, but the season was lost by a struggling offense with third-year quarterback Blake Bortles. Perhaps adding a running back like LSU's Leonard Fournette in the draft will help, but the Jaguars are likely to be one of the youngest teams again in 2017.

If we average snap-weighted age for all 11 years we have data, Jacksonville has fielded the youngest average team (SWA: 26.3) in the NFL since 2006. The Packers are surprisingly the second youngest at 26.4 years. The Redskins (27.6) and Steelers (27.4) have had the oldest rosters on average.

2016 Snap-Weighted Age: By Position

Finally, we measured SWA by position. This excludes fullbacks since some teams never use one. "ST" only includes the specialists on special teams (kicker, punter, long-snapper).

The method used was to create a z-score for each team's SWA by position. For example, the Jets had three quarterbacks combine to take 1,043 snaps last year. A 34-year-old Ryan Fitzpatrick took 765 of those snaps, or 73.3 percent. We multiplied each player's snap percentage by his age and summed them together for each team and position. New York's quarterback SWA was 31.6 compared to the NFL's average of 29.0 in 2016. We took that difference and divided it by the standard deviation in quarterback SWA (4.04) to get a z-score of minus-0.66 for the Jets.

The same was done for each team for each position, and we took the average z-score for all positions for a team to rank them in the table below. Instead of overloading you with z-scores for every position, we just posted where the team's z-score ranked for SWA for each position, sorted from oldest (first) to youngest (32nd). The 12 playoff teams are highlighted and each team's rank is sorted into quartiles: red (eight oldest), orange (ninth to 16th), blue (17th to 24th), and green (eight youngest). This should help highlight where each team is older, younger, or about average in regards to the rest of the league in SWA.

Through this table, we can see that the Falcons had the oldest offensive line and the Patriots had the oldest team among the 12 playoff teams, but there was a pretty good cluster of playoff teams in the 19-to-28 range too. The Ravens barely edged out the Saints as the oldest team when adjusting for position, and may have taken the SWA title outright had Benjamin Watson, a 36-year-old tight end who came over from the Saints, not missed the whole season with a torn Achilles. Also, instead of just accepting that a team like Dallas was 20th in SWA on both sides of the ball, we can see the Cowboys were balanced out by having a very young quarterback (Dak Prescott) and running back (Ezekiel Elliott), while fielding the oldest players at tight end (primarily Jason Witten).

Posted by: Scott Kacsmar on 20 Apr 2017

5 comments, Last at 24 Apr 2017, 12:08am by LionInAZ

Comments

1
by MEVK :: Thu, 04/20/2017 - 5:20pm

Looking at the rankings by position, I was curious which positions skew older (QB?) and which skew younger (RB?). Median ages might be more informative than averages, particularly if you only look at the 2016 data.

2
by Pen :: Thu, 04/20/2017 - 10:05pm

I remember Mike Holmgren saying you needed an older, experienced offensive line, but a young, hungry and quick defense. Now Seattle has the exact opposite, except strike the hungry and quick from the oline part.

3
by ammek :: Fri, 04/21/2017 - 3:32am

Is that the same Mike Holmgren whose Packers won the Superbowl with only one offensive lineman over the age of 26, and only three defensive starters under the age of 26? That young, hungry defense was anchored by Reggie White (aged 36), Sean Jones (35) and Eugene Robinson (33). It led the league in points and yards allowed.

4
by LionInAZ :: Mon, 04/24/2017 - 12:03am

A team with a veteran Reggie White on side and a veteran Brett Favre on the other side is probably a very atypical example. Don't cherry pick exceptions.

5
by LionInAZ :: Mon, 04/24/2017 - 12:08am

These numbers don't seem very useful without some kind of analysis of which units benefit most by veteran presence. It's obviously easier for skill positions with DVOA numbers, but not for linemen or defense unjits. FO hasn't even tried to break it down this finely.