The Offseason Reading Series #11: So you want to draft a running back?

[21isSuperman]

Another ORS installment as the offseason rolls on...

 

Previous ORS installments

ORS1: The best Indianapolis Colts team ever

ORS2: Which Indianapolis Colt are you?

ORS3: Dissecting the 15th overall pick

ORS4: Choose your contract

ORS5: Which Simpsons characters are the Indianapolis Colts?

ORS6: The best trash-talking moments of Peyton Manning's career

ORS7: My favourite Andrew Luck throws

ORS8: Changes the NFL needs right now

ORS9: Projecting Moncrief's contract

ORS10: The NFL's MVP award

 

In the 2017 NFL Draft, much was made of the available running backs for the Colts.  Many mocks had the Colts taking Dalvin Cook from Florida State.  In fact, we even had a thread where some thought we should have taken Cook over Hooker at the 15th spot.  The counter to that claim is usually that taking a running back in the first round does not give you a strong return on your investment.  RBs take so much wear and tear that they don't last very long.  This got me to thinking, how long do RBs last in the NFL, and would it have been worth it for the Colts to take Cook at the 15th spot?

 

When do running backs peak?

 

There is a common acknowledgement that after age 30, NFL players begin to see a dip in their production.  The belief is that this holds even truer for running backs, whose jobs force them to take tons of hits and pounding throughout their careers.  However, there are many who believe using a first round pick on a running back is a good investment.  So when do running backs usually peak and are they worth first round picks?  Are good running backs able to last long enough in the NFL that they provide a good return on the investment of a first round pick?

 

The sample

 

I decided to analyze rushing yards per game, games played, and games started for the entire career of some running backs.  This would give me an idea of how good the running back is playing each year and how many games they miss due to injury or ineffectiveness.  To extract a good sample, I used running backs who were top 10 in the league in rushing yards at least twice in from 2000 to 2010; this allowed me to examine running backs who actually had success, but also avoid any one year wonders.  To make this a career-spanning study, I used data from 2000-2010 because that would allow me to examine the entire careers of players under analysis.  That is, if I had included players beyond that, I wouldn’t be able to examine their entire career trajectory.  And if I included players from earlier than 2000, rule changes and the nature of today's game might make that data not applicable.

 

This gave me a list of 27 running backs to analyze.  I excluded any season where the player had 4 or fewer games played.  For the graphs that I will present, I took the average of the stats for the 27 RBs.  The X-axis for all of the graphs is the number of years in the league rather than age of player.  I figured it wouldn’t make sense to compare ages since players come into the NFL at a variety of ages, so the better comparison would be to compare years in the league.  This would allow me to normalize comparisons to players who are at the NFL level, against NFL competition, in an NFL offense at the same time.

 

The results

Click on any graph to enlarge it

 

I devised three graphs from this data.  The first compared average YPG to number of years in the league.

 

 

As you can see, there is a consistent growth in average YPG from the rookie year to about the 4th year, then there is slower improvement until about the 6th year, where RBs tend to peak.  After that, it’s a downhill trend in production.  Once RBs hit the 10 year mark, their production falls off a cliff.

 

To see this data in a table…

Year in league	YPG	Change from previous year	% Change from previous year
1	48.5	N/A	N/A
2	65.4	+16.9	+34.8
3	74.4	+9.0	+13.8
4	78.5	+4.1	+5.5
5	79.0	+0.5	+0.6
6	81.3	+2.3	+2.9
7	73.2	-8.1	-10.0
8	71.5	-1.5	-2.0
9	61.6	-9.9	-13.8
10	61.6	0	0
11	41.9	-19.7	-32.0
12	31.0	-10.9	-26.0
13	25.0	-6	-19.4

 

Take home message from this graph and table: RBs are usually productive for about 8 years.

 

This next graph shows how long the RBs lasted.  That is, of 27 RBs under analysis, what percentages played in their # season in the league?  Reasons for missing a season could be injury that cost them at least 12 games (eg. Jamal Lewis in year 2, Ryan Grant in year 4) or careers ending/retiring (eg. Shaun Alexander’s last year was his 8th in the league).

 

 

In this case, we see a constantly high availability until year 5, then a sharp and consistent decrease from years 6 to 13.  RBs drop like flies after 7 years, with more than 20% of them not playing an 8th year.  The only RBs of the group under analysis to make it to 13 years were Ricky Williams, who deserves an asterisk because he wasn’t in the league for two years in between, and Fred Taylor.

 

Take home message from this graph: RBs tend not to last more than a decade.

 

The next graph shows the average of how many games these RBs started each year as a pro

 

 

As with previous graphs, we see an increase in starts made from years 1-6, then a consistent decrease.  By year 12, only 5 RBs were still around (Fred Taylor, Marshall Faulk, Ahman Green, Ricky Williams*, Thomas Jones) and they were only starting an average of 2 out of their team’s 16 games.

 

Take home message: After about 6 years, RBs starting seeing a dip in how many games they start, showing that their decline begins to be more evident to their coaches as well.

 

Where did these RBs come from?

Lastly, I thought it would be interesting to see where each of the running backs on this list were drafted.  Of the 27 running backs included, 13 were first round picks and 14 were not.  That means you absolutely do not need a first round running back for him to be a great player.  For anyone interested, I am more than happy to include the names of the 27 RBs used in this analysis.

 

What does it all mean?

 

When you combine all of this data together, we see that RBs peak around years 5/6, and are decently effective until about year 8.  This certainly does give credence to the theory that running backs hit a wall after age 30, as most players come into the league in their early 20s and an 8 year span of effective play would get them to around age 30.

 

The bigger question is how this relates to first round picks.  Under the CBA, first round picks have a 4 year deal with a fifth year team option.  Assuming a pick follows the historic trend, he will show consistent improvement until his 5th year.  Then come contract negotiations, where he will want a long term deal to reward him for the production of the last 5 seasons.  With this data in mind, we can see that these running backs only have about 3 more years of good football left in them at this point.  Furthermore, this data shows that successful running backs are first round picks only about 50% of the time.

 

At the end of the day, it's always up to the GM and coaching staff to make their own decisions.  If they feel it is worth it to take a running back in the first round, then they will do so.  But with: 1) many teams going to running back by committee, 2) the great success non-1st-round RBs have had, and 3) only about 50% of running backs making it to their 10th year in the league, my personal opinion is that it is not worth it to use a first round pick on a running back.  I think it would not have been worth it for the Colts to take Cook with the 15th pick, and I'm glad we didn't.

 

One last thing to note: there are lots of statistics and data one can analyze.  To make this analysis more thorough, it would be worth it to look at the performance of all first round pick running backs during that same time span.  The above data shows successful running backs are first round picks about half of the time, whereas analyzing all first round pick running backs would provide different data and conclusions that would complement what it seen above.

[coltsdan]

are you kidding me right now?  who is this 21guy?  this sup guy?  you think you can just post this stuff?   huh?  and then link the other posts?  really?  who do you think you are?   some moderator or something?   some all knowing poster guy?  

nice job man  

[21isSuperman]

21 hours ago, coltsdan said:

are you kidding me right now?  who is this 21guy?  this sup guy?  you think you can just post this stuff?   huh?  and then link the other posts?  really?  who do you think you are?   some moderator or something?   some all knowing poster guy?  

nice job man  

Thanks!

[WoolMagnet]

So i read draft mid rounds , keep refreshed thru draft with a rb drafted every 3 or so years.  Run them into the ground on their "cheap" rookie contracts and then let them walk b4 the average decline and bigger $.  You should have the next man up if you've drafted well. 

 This is for your "planned" rb #1 and #2.  Plus you can fill out the roster or get lucky with UFA and guys cut.

but it sure looks like the years 2 - 4 are the average "peak."  This really hints toward having a great RB scouting team to really improve your success.

then again, thats just good overall personnel planning.  You GOTTA build thru the draft.  The cheaper salaries give you such an advantage to sign or keep key pieces that demand bigger $.

sounds so simple, doesnt it?  

[Superman]

On 7/3/2017 at 0:49 PM, 21isSuperman said:

But with: 1) many teams going to running back by committee, 2) the great success non-1st-round RBs have had, and 3) only about 50% of running backs making it to their 10th year in the league, my personal opinion is that it is not worth it to use a first round pick on a running back.  I think it would not have been worth it for the Colts to take Cook with the 15th pick, and I'm glad we didn't.

 

Yup. The impact that individual RBs have on the outcome of games isn't big enough, in comparison with other positions. And the longevity, or lack thereof, is a problem as well. 

 

During the draft, defending drafting a RB in the first round, Bill Polian said 'look at Ezekiel Elliott.' And he's right, Elliott was fantastic for the Cowboys as a rookie. I don't think he was as important as he's made out to be, but there's no doubt he was a critical part of a really good team. But let's see what Zeke is doing three or four years from now. Let's compare him at that point with the players drafted after him -- Ronnie Stanley, Jack Conklin, Deforest Buckner, Jalen Ramsey, Eli Apple, Ryan Kelly, Karl Joseph, etc. And let's look at his second contract vs those other players. 

 

The draft is about building blocks for the future of your team, not so much what happens in Year 1. I'm sure the Cowboys are thrilled that Zeke played so well, and they should be. But I have trouble believing they'll be eager to pay him when his contract comes up. Those other guys, assuming they're healthy and play well, will be with their teams for 7-10 years, maybe longer. 

[OffensivelyPC]

Given that the average age of the NFL draftee is about 23 years of age (22 for higher drafted picks, I'm guessing Day 1 and Day 2 picks, which most of the guys you mentioned probably were), your results certainly prove the well known adage surrounding 30 years of age.  Their max productivity caps somewhere around 27-28, remains acceptable for another 2 years and then they are in free fall.  There's always the exceptions like Frank Gore, but by and large, 29 or 30 is the cutoff age.

 

The only bit I really disagree with is the part about starts made/missed - if only because we're talking about one play.  Some of that could be determined by game planning, matchups, personnel, etc.  I think you're right to point out there's a loose correlation, but I don't think it means the coaches sense decline is on the verge.  It could be they were knicked up, which you already addressed in a more meaningful way when showing the percentage of games played in the section before.  It may be obvious that production will decline in the coming years, but that's not my takeaway based only on the first offensive play of any game.

[21isSuperman]

18 hours ago, Superman said:

 

Yup. The impact that individual RBs have on the outcome of games isn't big enough, in comparison with other positions. And the longevity, or lack thereof, is a problem as well. 

 

During the draft, defending drafting a RB in the first round, Bill Polian said 'look at Ezekiel Elliott.' And he's right, Elliott was fantastic for the Cowboys as a rookie. I don't think he was as important as he's made out to be, but there's no doubt he was a critical part of a really good team. But let's see what Zeke is doing three or four years from now. Let's compare him at that point with the players drafted after him -- Ronnie Stanley, Jack Conklin, Deforest Buckner, Jalen Ramsey, Eli Apple, Ryan Kelly, Karl Joseph, etc. And let's look at his second contract vs those other players. 

 

The draft is about building blocks for the future of your team, not so much what happens in Year 1. I'm sure the Cowboys are thrilled that Zeke played so well, and they should be. But I have trouble believing they'll be eager to pay him when his contract comes up. Those other guys, assuming they're healthy and play well, will be with their teams for 7-10 years, maybe longer. 

Absolutely.  The one thing I always think of with Elliot is how much the Cowboys OL is helping him.  Elliot is a great and talented back, no doubt.  But put him behind Detroit or the Rams OL and see how he does.  At the same time, put a 4th or 5th round running back behind the Dallas OL and see how he does.

 

1 hour ago, OffensivelyPC said:

Given that the average age of the NFL draftee is about 23 years of age (22 for higher drafted picks, I'm guessing Day 1 and Day 2 picks, which most of the guys you mentioned probably were), your results certainly prove the well known adage surrounding 30 years of age.  Their max productivity caps somewhere around 27-28, remains acceptable for another 2 years and then they are in free fall.  There's always the exceptions like Frank Gore, but by and large, 29 or 30 is the cutoff age.

 

The only bit I really disagree with is the part about starts made/missed - if only because we're talking about one play.  Some of that could be determined by game planning, matchups, personnel, etc.  I think you're right to point out there's a loose correlation, but I don't think it means the coaches sense decline is on the verge.  It could be they were knicked up, which you already addressed in a more meaningful way when showing the percentage of games played in the section before.  It may be obvious that production will decline in the coming years, but that's not my takeaway based only on the first offensive play of any game.

Fair point.  I just thought there might be a loose correlation because you usually put your best guys out there to start, then sprinkle in the rest, save for a few exceptions as you mentioned.  The "Played this year" graph would be more meaningful, in that case.

 

18 hours ago, WoolMagnet said:

So i read draft mid rounds , keep refreshed thru draft with a rb drafted every 3 or so years.  Run them into the ground on their "cheap" rookie contracts and then let them walk b4 the average decline and bigger $.  You should have the next man up if you've drafted well. 

 This is for your "planned" rb #1 and #2.  Plus you can fill out the roster or get lucky with UFA and guys cut.

but it sure looks like the years 2 - 4 are the average "peak."  This really hints toward having a great RB scouting team to really improve your success.

then again, thats just good overall personnel planning.  You GOTTA build thru the draft.  The cheaper salaries give you such an advantage to sign or keep key pieces that demand bigger $.

sounds so simple, doesnt it?  

I think that's a fair assessment.  Run them into the ground and get every ounce of football out of them as you can on their rookie deals, then let them walk.  They'll begin to decline on their next contract and you don't want to be paying big money to a guy who is declining each year.