Does Team Success Matter for Running Back Fantasy Production?

Michael Zingone Posted On May 11, 2018

Yearly production from the running back position can be a rollercoaster, to say the least. Most of the time in redraft leagues, the teams that bust on players like C.J. Spiller and Eddie Lacy in the first round are also the teams that don’t make the playoffs. Wide receivers can be far more predictable on a year to year basis; that’s why many dynasty analysts preach taking them early and often in dynasty startup drafts.

Although this website is indeed titled Dynasty League Football, and you’re likely not reading this unless you play that format, fantasy football is inherently a year-to-year game. I’m here today to provide you a statistical analysis of how you can systematically be better at predicting running back success in fantasy on a yearly basis.

This is a pretty bold claim, I know, and the answer may actually be somewhat intuitive. I set out with the question in mind: “Does NFL team success (wins) correlate to a higher fantasy output for that team’s running backs?”. The answer is a resounding yes.

[am4show have=’g1;’ guest_error=’sub_message’ user_error=’sub_message’ ]

I understand that there’s a lot baked into this question, and a simple analysis likely doesn’t tell the full story. Was a team successful because of its running back or was a running back successful because of his team? It’s kind of like the question: Which came first: the chicken or the egg? The chicken would not have been so without the egg, but the egg would not have been so without the chicken. The two are interconnected, and there’s merit to both sides.

The Study

I went back and looked at all team records and the top 50 ppr running backs each year until 2008, a ten-year sample containing 320 data points (data from pro-football-reference and fantasydata.com, respectively). From there, I linked each running back’s PPR finish back and total points scored to how their team did in that year to get an aggregated view of the average PPR rank and points for each number of team wins.

There was a lot of data to unpack. As one would expect, there weren’t many 16-0, 15-1, 14-2, 1-15, and 0-16 seasons within a ten-year sample. These seasons are relative outliers as compared to the more middle-of-the-pack records, combining for only nine total seasons of the 320 sample. For this reason, these seasons aren’t included in the overall study. The data is simply too thin to draw substantiated conclusions from, with none of those seasons occurring more than three times.

The Data

First and foremost, I want to share the data table that I will be referencing throughout this piece through various statistics and charts. I’ll explain what each column represents in full, but try digesting it yourself before reading on! I will also list column numbers for reference, so as to make things a little easier to refer back to. Hopefully, you can already see the high correlations between many of the fields and team wins.

screen shot 2018 05 01 at 21.05.20

Wins – The number of wins the team had that season.
Num Seasons – The number of times a season with that number of wins occurred in the sample.
Highest Best RB Finish – The best PPR finish by the best running back on a team with that many wins.
Lowest Best RB Finish – The worst PPR finish by the best running back on a team with that many wins. If a team did not have a running back within the PPR top 50, they were given 55 as their best finish.
Average Best RB Finish – The average best PPR finish of a team’s best running back.
Average # RBs in top 12, 24, 36, 50 – The average number of running backs on a team to have a PPR finish within the specific range.
Avg Best RB FPs – The average number of PPR fantasy points the best running back on a team with that many wins had.
Avg Other Top 50 RB FPs – The average number of PPR fantasy points scored by running backs in the top 50 on that team who were not the top PPR scorer on their team.
Pct Top RB is Pass-Catcher – The percentage of occurrences where the top scoring PPR running back on a team scored more than 50% of his fantasy points through the passing game (rec/yds/TDs).

The Results

At first glance, this data can seem like a lot to process. Let’s quickly think back to the question we wanted to answer at the onset of this piece: “Does NFL team success (wins) correlate to a higher fantasy output for that team’s running backs?”.

So let’s go through piece by piece!

Range of outcomes – Over the past ten years, teams that finished with at least four wins were able to produce a top three PPR finish fantasy running back. While teams with fewer than four wins did not have a high-end RB1 on their team, a top 12 finish is still in the range of outcomes. Additionally, each win bucket showed the possibility of having a complete fantasy dud on their team. There’s no statistically significant result to draw in the range of outcomes depending on team wins.

Average points and PPR finish – Where this study really starts to make strides is in analyzing the correlation between team wins and overall best RB performance on that team. Below is a chart illustrating the data above as it pertains to RB performance dependent on Team Wins.

word image

The blue dots above represents the data in column (5) of our data table. It is the average PPR finish for a team’s best RB. The blue solid line represents the linear trend between the data points. In this way, the visualization is clear that as team wins go down, average best RB finish goes up.

In fact, these two data sets (Avg Best RB Finish and Wins) have a correlation of -0.89, which represents a very significantly strong negative correlation. For those who might be unaware, correlation is a measure of how related two sets of data is. The closer the number is to -1 or to 1, the stronger the relationship between the two data sets is.

We can analyze the orange dots and solid line in a similar manner, which represents the data in column (10) of our data table. This data is very similar in nature to the blue lines, as it measures performance. One would expect these two lines to have an inverse relationship, and they do. The higher the fantasy points scored, the lower the rank should be!

Congruent with the blue lines, the orange lines show a very strong correlation to team wins. In fact, the correlation is even better coming in at 0.93! This part of the study without a doubt showed me that there was merit to my initial question. The results indicate with very reasonable certainly that team wins correlates well with best running back production.

The other guys – One fascinating part of this study to me was that not only did better teams yield better top running backs, but they also provided more fantasy value to their second or third best running backs. Perhaps this makes intuitive sense; if a team is winning more, they’ll run more often and perhaps want to spread out the workload. But, if this was true, then one might think that the top running back on that team would suffer due to the committee.

This data, shown in column (11) in the data table, provides a correlation coefficient value of -0.78, which is not quite as high as the previous results but is still a really strong correlation. So, we find that poor performing teams not only struggle to produce top level running but also struggle to produce other fantasy relevant running backs. The opposite is true of strong teams (think the Saints); they can sustain an elite running back as well as depth at the position.

Other interesting notes – Lastly, I want to list out some findings that aren’t necessarily worthy of a full section but are worth mentioning here:

There is no data that suggests that worse teams rely on a pass-catching running back more than good teams.
There are generally more top 50 RB fantasy points available to better teams.
Teams with worse records tend to have a higher percentage of their fantasy points go to their lead running back, though this doesn’t necessarily translate to better overall production for that player.

Summary

At the end of the day, fantasy football is a game of probability. The round a player gets drafted in is indicative of the general community’s belief of how likely that player is to succeed at a high level. In redraft formats specifically, this is emphasized. Being able to identify value players is difficult, but it can often lead to fantasy success more so than hitting on a first or second round pick.

The goal of my study was to help provide another data point to help in your own player valuation. Taking team situation into account is something us dynasty owners focus on so much during rookie draft season but is often left to the wayside when it comes to making our lineups. It is for these reasons that I’ll likely avoid David Johnson at his price this year or perhaps fade any running back starting for the Colts if Andrew Luck isn’t playing. I don’t believe either team will/would finish with more than five wins.

This is not to say these players can’t succeed in spite of being a part of a bad team; the elite players often do. There’s also the point, of course, that this data is interdependent, meaning two things: the performance of the running back could affect team success and vice versa. However, when deciding between two players on draft day to push that big red “DRAFT” button on, I’ll be picking the player on the team I think will do better. The odds say I’ll do better too.

[/am4show]