Re-Evaluating Draft Success

I read a great article by Kevin Seiert which quantified team ownership/front office success based on how many draft picks were still on a roster.  In a nutshell, it looked at how many original draft picks used by a team were still on that team currently.  You can see the results below.

However, this got me to thinking: shouldn’t we look at this based on total picks still on a team.  What if a team had fewer or more draft picks?  Shouldn’t we factor that in?

I didn’t think it was fair to just arbitrarily select a time frame to study the number of draft picks, so what I did was looked back at the average NFL career.  As it turns out, for each player selected in the NFL draft, his average career length is 5 years.  Some are longer, some are much shorter, but the average is 5 years.

So I went back and grabbed the most recent 5 year span, from 2009 thru 2013, and studied all of the draft picks selected in those 5 years.  Because that is the average career length, so we shouldn’t penalize a team who drafted a player in 2005 (for example) who is no longer on that team.  We only want to look at players MAXIMUM 5 years back and more recently, as that is the career average.

Analyzing all the draft picks in the last 5 years and comparing those numbers to the total number of original draft picks selected by the team who are still on the roster, we find some interesting results:

As you can see, many of the teams with the most picks are still at the top.  Some takeaways:

  • Atlanta’s 86% is simply MASSIVE and far ahead of the rest of the pack.
  • The Chargers, Jets and Saints, despite a smaller number of draft picks still on their roster, are ELEVATED in my analysis because they simply didn’t have many to begin with.
  • Meanwhile, the Bengals, Seahawks and Eagles are all significantly lowered because they each had 46+ picks to work with, which was more than most teams, yet their picks-retained percentage is average to below average.

Also keep in mind the far right column, where its shows the average draft pick each team had over the 5 year sample.  It makes the results for some teams even better and some teams even worse.  Why?

According to the NFL, the average NFL career length for 1st round draft picks is 9 years.  My numbers show its more like 7.8 years.  Whereas my numbers show the average length for a player drafted in the 7th round is just under 3 years.  So teams who have better draft picks should have an easier time keeping them in the league and (thus) on their team.  Whereas teams with bad draft picks should see a greater turnover.

But as the chart above shows:

  • The Cleveland Browns, Denver Broncos and Jacksonville Jaguars had a VERY good average pick, but wound up at the bottom of this analysis, keeping 50% or less of their draft picks.
  • Whereas some of the best teams in the league, like Atlanta and Green Bay, had very bad average draft picks yet wound up at #1 and #3 in this analysis, finding a use for most of these lower draft picks.
  • Seattle keeping only 57% (19th in the league) doesn’t look as bad when you consider they had the 2nd worst avg draft pick in the league.

More: Top Analysts Say NFL Will Turn to Running the Ball but I Disagree

Read more NFL Draft and NFL Gambling Insight


Average NFL Career Length

In researching average NFL career length for a future article, I found there were two very different schools of though which were discussed back in 2011 during the lockout.

The NFL argued that “the average career length for a player who makes a club’s opening-day roster (active/inactive roster or injured reserve) in his rookie season is 6.0 years.” but went on to throw out statistics about average length for pro-bowlers at 11.7 years, and this was deemed a joke by many in the media.  In addition, the NFL used 1993-2002 as the measuring stick.  But that was in an era when teams may have been more patient, less cutthroat.  Its more relevant to use recent years.

Meanwhile, the NFLPA argued that the average career length for a NFL player is 3.2 years, although according to their own numbers, they are basing it on average “accrued seasons” and their number is 3.54 accrued seasons (definition is that a player must be on the roster at least 6 games in a season).

However, the NFLPA used VERY tricky math.  They ran their report PRIOR to the first game of the 2010 season, and included (as zero accrued seasons) players drafted in 2010 who didn’t have a chance to play in any game (since it was measured pre-week 1).  See their chart:

Obviously that data is tainted.  They should have run the numbers immediately AFTER the 2009 season, rather than pre-week 1 in 2010, after all the new draft picks of 2010 were part of the pool.  Note that if you remove all the “Zero Year” players from that analysis, the NFLPA average “accrued” seasons is 4.4 per player.  At any rate, we know their numbers are suspicious, but where is the truth?

I looked back at all the drafts since 2002.  I determined how many years each player played in the NFL (regardless if they “accrued” a year or not). Obviously we don’t want to include 2013 as a year, for instance, as a player who was drafted then could have only played 1 season, which would have screwed up the numbers.  But we don’t want to go back to 1993 like the NFL did in their study.

What I did (as you can see below) is looked back to 2002 and found that in the earlier years (when players who were drafted in 2002 would have 12 full years up to 2013 to factor into the average) players played an average of between 5 and 6 seasons.  Clearly, as you move forward in time, you give newer players less time to contribute and the numbers decrease.

As such, I decided to NOT include the 6 seasons from 2008-2013 as they would unfairly decrease the average, and I only used 2002-2007 for the study.

If you count players who were drafted but never made the team as the NFL getting 0 years out of that player, I conclude that on average

  • each DRAFTED PLAYER averages exactly 5.0 NFL seasons

If you exclude players who were drafted but never made the team, I conclude that on average

  • each DRAFTED PLAYER WHO MADE THE TEAM averages 5.7 NFL seasons

These numbers are closer to the NFL’s calculations than the NFLPA, but not surprisingly they are in the middle.

Depending on the point you want to prove, you could use either 5.0 or 5.7 as the average career length of a NFL player.

Top Analysts Say NFL Will Turn to Running the Ball but I Disagree

This past month I read two articles written by gentlemen who I respect in the football world, and both discussed the same notion:  the NFL is not going to be a passing league moving forward, in large part because its a copycat league – Seattle won it all last year with a strong running attack so teams will try to mimic their success by turning to the ground.  Here is what Greg Cosell wrote and here is what Pat Kirwan wrote.

First, let’s examine how and why the Seahawks won the Super Bowl.  I’ll submit it was their defense #1 and Wilson #2, rather than the running game #1.  Second, let’s examine how the 2013 Seahawks were built.  This team almost played in the Conference Championships in 2012.  They were very close to the Super Bowl that year.  What did they add that offseason:  pieces to focus on running the ball or pieces to focus on passing the ball?  And lastly, let’s actually look at some real statistics using regression and other analysis which should drive a final nail into the debate of the NFL and its direction as a running or passing league.

Why did Seattle win the Super Bowl?

During the season, and even in the 2 weeks heading up to the Super Bowl, the entire discussion was how Russell Wilson was not fit to win a Super Bowl.  Maybe not to that extreme, but there were disparaging comments comparing him to Peyton Manning, and indicating that the Seahawks had been winning despite Wilson, not because of him.

At the time, the entire discussion revolved around the strength of their defense.  This was the #1 defense in the NFL, going up against the #1 offense of the Broncos.  The talking heads analyzing the Super Bowl seemed to be from 1 of 2 schools of thought:

1.  Either that Russell Wilson would struggle and therefore the Broncos would win, or
2.  If Seattle was actually able to win the game, it would be on the back of their strong run game.

I disagreed with both points.  Of course, I backed up all of my opinion on solid fact, which you can read below in clips from my 8 page write-up that I shared with my clients the week before the Super Bowl. (Click to enlarge.)

The Super Bowl was won because of the Seahawks defense first and foremost, and secondly because of the quarterback’s performance.  There is even a point in Cosell’s article where he says “The bottom line: the Seahawks won a Super Bowl with the running game as the starting point.”

For some reason, suddenly these authors out there believe that Seattle won the Super Bowl because of their run game.  The point is not an “I told you so”, but I feel the need to remind them that:

  • Marshawn Lynch only ran for 39 yards on 15 carries (2.6 ypr).  Turbin gained 25 yds on 9 carries (2.8 ypr).
  • As I predicted, Russell Wilson was 18/25 for 206 yds, 2 TDs and 0 Ints.  That’s a passer rating of 123 and a QBR of 88.1 (out of 100).

I have nothing negative to say about Lynch.  I agree with Cosell that Lynch “possesses all the traits you look for in a feature back”, and Seattle got him for a decent deal. But I believe that Seattle won the Super Bowl with defense as the starting point, and offensively they won because Wilson in 2013 played efficiently and effectively.

How were the 2013 Seahawks Really Built?

So let’s back up for a minute.  I already showed that the way the Seahawks actually “won” their Super Bowl was defense and Wilson.  Not the running game.  But don’t take my word for it.  Let John Schneider, the GM of the Seahawks, tell you what they wanted to focus on with this 2013 team.

In 2012, the Seahawks were the #1 overall most efficient team in the NFL.  Ahead of Denver and New England already, the Seahawks boasted the #2 defense and the #4 offense from an efficiency standpoint.  But Schneider wanted to add something else to the team.  Something he felt they needed to get over the hump and win a ring.  So where did they turn, passing or running?

Schneider brought in WR Percy Harvin and paid him $12M in signing bonus and $10.7M average per season.  Even though Seattle already had their #1 receiver from 2012 Sidney Rice returning, along with their #3 receiver TE Zach Miller, throwing Percy Harvin into the mix would make this passing offense one to be feared.

Why did they need more firepower?  Because in games the Seahawks lost in 2012, the offense scored just 18 ppg.  And they lost every single one of those games by 7 or fewer points (an avg of 4 ppg).  So if the Seahawks offense could be just a bit more potent, Schneider thought this team would win it all.  After all, in their wins, they averaged 30 ppg.  Clearly, losing 6 games by an avg of only 4 ppg and scoring just 18 ppg shows they were 1 more TD away from winning every single game.

Schneider decided to build up the passing game in 2013, NOT the running game.  So they brought in a massive receiving weapon in Harvin.  But then disaster struck:

They got stung by the injury bug.  Harvin didn’t even play until the end of the season, so they lost their major acquisition which could help elevate their offense.  Then their leading receiver from 2012 (Sidney Rice) was injured most of the season and missed most of the games, recording just 15 total catches (vs 50 the prior year).

Seattle was now down to backup receivers.  Not plan A catching the ball, and not even plan B.  They still had the BEST defense in the NFL, and a very young quarterback.  But he didn’t have any real weapons at receiver any longer.

So the Seahawks had two options in 2013 given the injuries to their wide receivers:

Option 1 – Let their young QB air it out to “plan B” wide receivers and hope it goes well, and if it doesn’t, they put their strong defense in tough spots.
Option 2 – Let their defense dictate the strategy, and only force it through the air when absolutely necessary.

The Seahawks decided to go with the option which most of us would go with:  option 2.  Because of their decision, Seattle ran the ball on 60% of their 1st and 10 play calls, which was 2nd most in the NFL (behind the Bills) and well over the 50% NFL average.

But if you look at the 3 teams who ran it the most often on 1st and 10, you’ll see the Bills, Seahawks and Jets.  Were these teams running because they had ridiculous rushing offenses and it was their go-to way to destroy their opponent?  No.  The Bills ranked 17th in rushing efficiency and the Jets ranked just 23rd.  They ran so much because they all had young, inexperienced quarterbacks and top 10 defenses from an efficiency standpoint.  Sounds similar to the Seahawks in that respect.  So did these teams run the ball because philosophically they thought it is the way to win football games, or did they run it because the strength of their defense and inexperience of their offense and quarterback?  Therefore, would other teams without strong defense and inexperienced passers decide to run it as often, or view it as the best way to win games?

The Seahawks are letting Wilson grow into the offense.  But now they know he can handle it.  And when he enters 2014 with a full compliment of receivers, I would be shocked if the team didn’t allow Wilson to fill a larger role in their offense than he did in 2013.  The Seahawks are simply going to manage their talent in the best way possible to play as efficiently as possible.  If Wilson was a 4th year quarterback and he had healthy Percy Harvin and healthy Sidney Rice playing, do you still think Seattle is running the ball on 60% of their 1st and 10 play calls?  They did what they had to do in 2013 without their top 2 receivers, and they will do what they need to do in 2014 with them.

Why Most Teams Can’t Rely on Defense like the Seahawks Did

So now that we’ve established the Seahawks won their Super Bowl because of defense and Wilson, and we’ve established the fact that the 2013 Seahawks were built to be a much better passing team but injuries to their top 2 receivers derailed that opportunity, let’s take a look at the black and white numbers.  And my bigger issue with the under-riding theme of these articles I read.

The issue of the “future” of the NFL, and the notion that it will be returning to the ground.  Cosell’s article poses the following questions at the end:

“Will they attempt to build “teams” rather than putting all the focus on “maybe” quarterbacks? Is it back to the future: run the ball, play good defense, with the quarterback the only nod to the new era?”

I’m fully on board with the notion of “team building”.  As I wrote in my initial article related to optimal roster design, the average top-10 quarterback salary was $18.7 million this past season.  But the last five teams to win Super Bowls had quarterbacks averaging just $8.7 million annually.  Emphasizing the need to have well balanced teams, as opposed to sinking tons of money into one position and then having a defense filled with scrubs.

But here is the problem with Cosell’s line of thinking related to the “future”:  most of the best teams in the NFL are already committed long term to their quarterback.  They sunk a lot of money into that position.  I’ve covered this in my article about overpaying for a “franchise” quarterback, but looking at most of the playoff teams from the AFC (Denver, New England, Indianapolis, Kansas City, San Diego) and the NFC (Green Bay, New Orleans, Chicago), they are already committed long term to their quarterbacks and sunk tens of millions of dollars into that position.  The teams who haven’t, such as Cincinnati, Philadelphia, Carolina, Seattle and San Francisco, have young quality quarterbacks who more than likely will get signed to a similar deal when their rookie contract expires (with the Bengals being the biggest question mark).

I’ve gone on record saying that I don’t agree with re-signing quarterbacks just because they played “well”.  As mentioned in my article, the NFL is littered with horror stories and fired coaches from quarterback extensions gone wrong.  But in today’s era, where quarterbacks make so much money, you have teams like the Ravens, Lions, Cowboys and Bears sinking a significant percentage of their cap into one player, because they don’t want to start over at that position, even if their quarterback really isn’t at that true elite “top 5″ level of production.

What all of this means is that most of the NFL’s best teams CAN’T play “good defense” because they have too much invested offensively.  It’s impossible to pay enough defensively to have a great defense while overpaying a QB and a few offensive skill players.

Take a look at the top 5 defenses from 2013 based on efficiency:  Seattle, Arizona, Carolina, Buffalo and Cincinnati.  In 2013, ALL of these teams were paying pennies to their quarterbacks compared to the NFL average.  Looking at their QB’s cap hits and their team overall cap hit:

  • Russell Wilson ($681K) and $1.5M as a team to all QBs (30th in NFL)
  • Carson Palmer ($4M) and $6.2M as a team (22nd in NFL)
  • Cam Newton ($6M) and $7.5M as a team (20th in NFL)
  • EJ Manuel ($1.6M) and $4.7M as a team (26th in NFL)
  • Andy Dalton ($1.4M) and $2.8M as a team (29th in NFL)

Consider that in 2013, the NFL averages were cap hits of $10.1M per team to all QBs and $3.4M per QB, we see that most of these great defenses were built by NOT having that money tied up in their quarterback.  As such, looking a the teams who recently paid big money to their quarterback, It will be very hard for Dallas, Detroit, Chicago  or Atlanta to build solid defenses when they are paying $18 to $21M annually for their quarterback.

To me, the thought that the “future” of the NFL is playing “good defense” and “running the ball” is an interesting assertion given the huge importance put on obtaining a “franchise” quarterback and paying him accordingly.  The best teams in the NFL have that franchise quarterback.  Its highly unlikely that they have the money to also pay for a top defense, and it’s just as unlikely that they will take the ball OUT of their quarterback’s hands and run it 60% of the time just because Seattle ran it that often in 2013.  Picture Manning, Brady, Rodgers or Brees just handing the ball off all game long while their mediocre defenses struggle to stop the opponent!

Do Teams Know that Passing Contributes Over 4 Times More than Rushing to Winning Games?

This article could end here, but you expect more from me.  Its not enough to use common sense and a few facts, we need to dig deeper into the analysis.  As we know, the NFL and Rodger Goodell changed the rules in 2010 to make it much easier and beneficial to pass the football.  So I looked at a few comparisons since 2011 (3 seasons) to see if I could find anything worth noting related to rushing, passing and winning.

First, I ran three regressions:  a regression on rushing yds/attempt to wins, a regression on total rushing yards to wins, and a regression on passing yds/attempt to wins.  It didn’t surprise me one bit that there was absolutely no relationship of any strength between either rushing metric and wins (rushing ypa to wins, for instance, had a P-value of 0.9 and a R^2 of 0.02%, both of which are TERRIBLE).  But looking at passing yds/attempt to wins, we find a very strong relationship (passing ypa alone explain over 37% of all wins, with a P-value of 4.5E-11).

As such, the last 3 years we find NO connection, relationship or correlation between teams who run the ball well and teams who win the most games.  Running is used to compliment passing, but passing is what determines wins and losses in today’s NFL.  [Do you really think teams would pay quarterbacks over $20M a year while paying their RB peanuts if passing the ball didn’t determine wins and losses?]

Secondly, I looked at the average cap hit per RB by team, and compared that to team wins.  What I found is the relationship is NEGATIVELY correlated.  What this means is that the more money a team pays to its running backs, the fewer wins they gain.  Look at 2013 as an example:  none of the top 5 teams in terms of cap hit per RB (Tennessee, Minnesota, Chicago, Oakland and Baltimore) finished the season with a winning record.

Lastly, I looked at the 4 teams who made the Conference Championship games the last 2 years.  I looked at both the average cap hit per RB and the total cap hit for their running backs on the roster.  No team who ranked top 5 in amount of their salary cap spent on RBs made it to the Conference Championships.  In fact, the average ranking of the 8 teams who made it the furthest in the NFL playoffs the last 2 years was 17th.  Meaning that these teams spent the NFL “average” on their running back position.  None of these teams paid to have the best running backs.  None of these teams had the highest paid running backs.  They won a lot of games in the regular season and made their great playoff runs with the average salary going to their running backs.

Of course, on top of these numbers, we have my analysis from my initial article on optimal roster design, which showed that 60% of the teams that posted 13-plus wins the last four years called run plays fewer than 45 percent of the time; vs. just 23% of the 13-plus win teams from 1989-2009.  Indicating the NFL is indeed getting more pass heavy, but not just that, the BEST teams are passing the ball much more than they did in prior eras of the NFL.

Lastly, the fact is based on looking at a regression analysis of Football Outsider’s efficiency metrics for the last seven seasons, offensive passing efficiency contributes over four times more than rushing efficiency toward winning games!


Am I on board with teams overpaying for their quarterback and fielding bad defenses?  No.  If you want to see read the optimal way to build a NFL team, click here.  I agree in building a balanced roster while focusing on the key positions I indicate in that article.  I also think reloading through the draft is vital, or teams will be left paying too much money on 2nd and 3rd year contracts.

But I do not believe the best passing teams in the NFL (such as the Broncos, Patriots, Packers and Saints), who passed the ball on average over 60% of play calls in 2013, will decide to run the ball 52% of the time in 2014 because that is what Seattle did in 2013.  Similarly, those teams called passes on 51% of their first and 10 plays, and I doubt they call runs 60% of the time in 2014 because that is what Seattle did in 2013.

The bottom line is while the Seahawks were the best team in the NFL last year, they ran because it was in their best interest given their beastly defense, the inexperience of their young QB and the injuries to their wide receivers.  If we are looking for things that teams might “copy” from the defending champions, it’s NOT their running back or decision to run often.  It’s their defense.  We’ve already seen teams like New England and Denver pay big money in free agency on the side of the ball they’ve recently ignored, the defense, in part because it helped Seattle so much in 2013.

This is a copy cat league, but when teams look at the Seahawks, they will try to mimic their defense.  Based on the circumstances surrounding the Seahawks in 2013 as well as the regression analyses I ran, there is no doubt in my mind this IS a passing league and the BEST teams are winning through the air.  Needless to say, when teams pay $15-$20M annually on quarterbacks, as they do each and every offseason, they are not doing that because they envision running the ball to win games.  They know the numbers show passing wins games in today’s era of the NFL, and it will win games in 2014 too.