Raj Mehta-USA TODAY Sports
Advanced Analytics Reveal how Broncos Should Approach Free Agency | WRs
The duality of NFL free agency lies in the excitement fans find watching teams try to 'buy' a trip to the Super Bowl in March with big-money signings and the disappointment those fans feel when teams sit on their hands doing next to nothing.
Creating excitement is paramount to a fan-driven sport, but teams getting caught up in that clamor and throwing money around like candy at a parade is not a prudent philosophy. Doing nothing is also a flawed philosophy. Teams would be wise to use a measured approach to the free-agent frenzy.
When I use the term 'measured', I don’t really mean slow and methodical. What teams need to do is look to analytics to understand which players pose the most risk vs. those who will provide a more solid reward for their money. Once the latter are identified, the team should not be slow in offering a contract.
As a team entertains the option of signing a player in free agency, there are many things to consider, but most importantly, how old is said player? This is paramount because it is highly likely a player will not live up to the contract they signed as they age.
However, how old is too old? The answer to that question is — much younger than you think.
Analytics easily answer that question. The following analysis includes any NFL player who was 23 years old or younger in 1998 and went on to play at least eight seasons in the NFL. The reason for this is to include a full career of all those in the data set and with the knowledge that they could have received at least one free agent contract.
The metric to understand their 'value' at each age was Performance Value. This is a metric I have used several times in draft studies and it continues here in free agency.
To understand optimal ages to dole out free agent contracts, there are two pertinent questions that the analysis must answer:
1.) At what age does a player’s performance rapidly decline? 2.) Does it vary by position? Both are answered by data visualization analytics using Tableau.
The next five graphs will shed some light on players at the skill positions. Obviously, the performance of all four skill positions decline as they age, but the timing depends on the position.
The first graph depicts the percent of players who are above the mean Performance Value of their position over time. In this case, time is age in years. The percent of quarterbacks who can continue to perform above the mean declines rather gradually at 32 years old, then rapidly declines at 35 years old.
Running backs and wide receivers are not so fortunate. Both begin a rapid decline at age 30. Tight ends start to rapidly decline at age 31.
The second graph depicts the average performance value by age. Again, the Performance Value of quarterbacks declines much slower than the other position groups. Interestingly, all the positions peak at the same age — 26. The typical age they receive their first free-agent contract.
At age 29, both running backs and wide receivers are in rapid decline in Performance Value. Tight Ends, on average, don’t have as high a Performance Value as the other position groups, but they do tend to decline gradually as they age.
Tight ends surpass running backs in Performance Value between ages 32 and 33. Note: only ages with at least 15 players who played at that age are included in the below graph.
The next graph is a scatter plot depicting every player in the data, at every age they played and the difference from the mean at their position for each of those ages. It is quite obvious again that quarterbacks tend to enjoy a more productive career later in age, while the other three position groups decline rapidly after their 20s.
There are very few running backs with a performance above the mean after age 30. Only one, Tiki Barber performed at a high level after turning 30.
The next part of the analysis compares a player to his own performance throughout his career. In this part of the analysis, each player is compared to his mean performance over his entire career.
The following graph is another scatter plot that shows each player by his age and the difference in his Performance Value for that season vs. his mean.
Again, it is quite apparent that quarterbacks enjoy a longer lifespan in the NFL. Running backs perform the worst as they age. In fact, not one running back had a Performance Value greater than his mean after age 33.
Wide receivers didn’t fare much better. Only two players had a performance value above their mean after age 33.
The next graph depicts the difference in a player’s Performance Value versus his mean. Again, their performance peaks at 26 years old.
Quarterbacks and tight ends decline more slowly than running backs and wide receivers. After 30 years old, all positions are below the mean.
These metrics are important to understand when it comes to NFL free agency. A team must know when a player will start to decline so that they can give out free agent contracts at the right time.
If a running back is an unrestricted free agent at age 25, a team can take the risk on giving him a four-year contract because he will likely fulfill that contract with higher than average performance. However, a team is foolish to give that same running back a second contract of any length.
Also foolish would be to give a first contract to a running back who is an unrestricted free agent at age 27. The chance of that player giving the team high value over the span of that contract is low.
The same can be said for wide receiver. Both positions are huge risks when contemplating signing them to a second free-agent contract.
Tight ends are less of a risk. If they receive a first free-agent contract at 25, a second contract would not be ill-advised if it is three years or less.
There are always outliers, but if a GM constantly chases anomalies, it will come back to bite them, badly. Very apparent is that most NFL teams reward free agents for past performance when they should instead invest in future performance.
This means teams should go after younger players with four-year contracts and be very cautious of aging vets (28 or older) unless it is an absolute necessity or if they need a vet presence for a single year or two. This is not to say that a future Hall-of-Famer should not be given a contract as they approach 30.
Most players of this caliber rarely become free agents until they are well past their prime. They are an anomaly as a player and when free agency is concerned.
What it Means for Broncos
How does this impact the Denver Broncos? Well, by all accounts, the Broncos are in search of a wide receiver and possibly a tight end to back up Noah Fant. Let’s focus on wide receiver.
At this moment in time, there are 43 unrestricted free agent wide receivers ready to become available on March 18. Only 13 are 26 years old or less. Meaning, only 13 players of the 43 seeking a contract are at the peak of their Performance Value. The other players are already in rapid decline and many (those over 30) have little left in the tank.
There are several examples in the data of wide receivers who had their peak at 27 or 28 instead of 26. Usually, those receivers showed a gradual increase in Performance Value leading up to their peak season, which was followed by a large increase in Performance Value.
If the Broncos look to free agency for a wide receiver, there are a few that match this trend, although they are on the 'not going to be great' spectrum.
Those players are Breshard Perriman, Chester Rogers, Demarcus Robinson and Geronimo Allison. Of those, only two, Perriman and Robinson are likely to have productive seasons over the next three years.
Robinson doesn’t really fit what the Broncos need to complement Courtland Sutton, which leaves Perriman as the player the team should go after. With that said, it might not hurt to take a low-cost flyer on Rogers, just in case.
Next up: Defensive Backs
Follow Thomas on Twitter @ThomasHallNFL and @MileHighHuddle.
