2022 RB Analytical Review part 1

Author: Billy Jones

Introduction

Welcome back to the blog! Today I begin my analytical review for running back fantasy football performance from 2022. As a fantasy football enthusiast, I have always been fascinated by the intersection of data analytics and football. However, I have come to realize that sometimes data analytics can get a little over complicated, and the vast amount of information available can be overwhelming when looking for easy actionable insights. That's what this series is all about, clean and simple analytics to help fantasy football fans heading into the 2023 season. I will be using basic data analysis techniques to uncover trends and patterns in the performance of running backs in the 2022 season and provide insights for fellow fantasy football enthusiasts. Whether you are a seasoned veteran or a newcomer to the world of fantasy football, I believe this series will provide valuable insights and help you make informed decisions when it comes to your fantasy team, so let’s get started!

Ground Rules:

Before we jump into the analytics, it is important to set some ground rules about how we will be calculating fantasy points. I am a big believer in the half PPR scoring system so that’s what we will be going with:

1. Rushing Touchdowns: 6 points for each rushing touchdown.
2. Rushing Yards: 1 point for every 10 rushing yards.
3. Receiving Touchdowns: 6 points for each rushing touchdown.
4. Receiving Yards: 1 point for every 10 rushing yards.
5. Reception: 0.5 points for every reception.
6. Fumbles Lost: -2 points for each fumble lost.

Additionally, I want to note we will be focusing on a pool of 58 running backs (shown below). The data used for this analysis was obtained from pro-football-reference.com, and all games to include where the running back snap share was greater than 30% to mitigate games where the player may have been injured, playing in garbage time, not yet incorporated into the offense (rookies), etc.. This will help ensure that the analytics aren’t skewed by anomalous game results and allow us to gain comfort in the conclusions we draw from the results.

Austin Ekeler	Aaron Jones	Dameon Pierce	Antonio Gibson	Samaje Perine
Josh Jacobs	Jonathan Taylor	David Montgomery	AJ Dillon	Javonte Williams
Christian McCaffrey	Rhamondre Stevenson	Cordarrelle Patterson	Melvin Gordon	Eno Benjamin
Derrick Henry	Leonard Fournette	Cam Akers	Isiah Pacheco	Zonovan Knight
Saquon Barkley	Miles Sanders	D'Onta Foreman	Damien Harris	Rachaad White
Nick Chubb	Ezekiel Elliott	Devin Singletary	James Robinson	Kenyan Drake
Tony Pollard	Jamaal Williams	Chuba Hubbard	Tyler Allgeier	Darrell Henderson
Breece Hall	Clyde Edwards-Helaire	Latavius Murray	Brian Robinson Jr.	Michael Carter
Joe Mixon	Alvin Kamara	Jerick McKinnon	Khalil Herbert	Chase Edmonds
James Conner	D'Andre Swift	Jeff Wilson	Deon Jackson	Kareem Hunt
Dalvin Cook	Travis Etienne	Raheem Mostert	James Cook
Kenneth Walker III	Najee Harris	J.K. Dobbins	Gus Edwards

This pool was determined by taking all running backs with a game average of greater than 5 points, a max individual game greater than 10 points, and requiring at least 6 games played. The only player not meeting these requirements included is Javonte Williams as I am interested in him as a long-term fantasy prospect. This set of games is shown below.

I do want to note that the removal of some of these games makes for a pretty small sample size for some players. The point of this series is to show what a player could be with a reasonable workload but keep in mind that a player like James Cook has all but 5 games removed from his figures.

Visualizations and Analysis

With the ground rules established and the data sources defined, we are now ready to dive into some analytics on running back fantasy performance from 2022 season.

Our first visualization is a histogram of total fantasy points for all games from 2022. The population average was 12.24 per game with a slightly skewed distribution. The x-axis is total fantasy points, the y-axis is frequency, and the bin size is 2 (e.g., 17-18 are grouped in 1 bar).

To provide a more granular analysis of running backs performance, I have split the RB pool into groups based on their average points per game rankings. Using this method, I am able to differentiate between high-end and low-end RB1s, RB2s, and RB3s, creating a more accurate understanding of the production profiles of these groups in later analytics. Specifically, my preliminary groupings are as follows: RBs ranked 1-6 (1a), 7-12 (1b), 13-18 (2a), 19-24 (2b), 25-30 (3a), 31-36 (3b), and 37+ in average points per game. By organizing the RBs in this way, we be able to understand the difference in production profiles for players in the same groupings in later analyses.

“Preliminary RB Groups”

1a - Austin Ekeler	2a - Dalvin Cook	3a - Travis Etienne	Cam Akers	Melvin Gordon
1a - Christian McCaffrey	2a - Aaron Jones	3a - Najee Harris	Jerick McKinnon	Raheem Mostert
1a - Josh Jacobs	2a - Jamaal Williams	3a - Samaje Perine	Devin Singletary	Rachaad White
1a - Derrick Henry	2a - Jonathan Taylor	3a - D'Onta Foreman	Brian Robinson Jr.	Tyler Allgeier
1a - Breece Hall	2a - Clyde Edwards-Helaire	3a - James Robinson	Damien Harris	Zonovan Knight
1a - Saquon Barkley	2a - Leonard Fournette	3a - David Montgomery	Isiah Pacheco	Javonte Williams
1b - Joe Mixon	2b - Miles Sanders	3b - Kenyan Drake	Antonio Gibson	Chase Edmonds
1b - Nick Chubb	2b - Alvin Kamara	3b - Gus Edwards	Eno Benjamin	Michael Carter
1b - Tony Pollard	2b - Dameon Pierce	3b - Khalil Herbert	Chuba Hubbard	Darrell Henderson
1b - Kenneth Walker III	2b - Ezekiel Elliott	3b - James Cook	Latavius Murray	Kareem Hunt
1b - James Conner	2b Deon Jackson	3b - J.K. Dobbins	Jeff Wilson
1b - Rhamondre Stevenson	2b - D'Andre Swift	3b - Cordarrelle Patterson	AJ Dillon

However, I didn’t love that I assigned tiers without reviewing any sort of data to validate that my generalized groupings were in fact reasonable. To do so, I reviewed each grouping by creating a bar chart of average points per game for all RBs in the pool. This will allow me to see if any adjustments need to be made to the current groupings.

After reviewing the bar chart of average points per game, it became clear that some adjustments were needed to the RB groupings. For example, it was apparent that certain players, such as Saquon Barkley and Breece Hall, did not belong in the same grouping as the other top-performing RBs. With this in mind, I have updated the groupings to better reflect the tiers in production output, as reflected in the new bar chart.

“Final RB Groups”

1a - Austin Ekeler	2a - James Conner	3 - Alvin Kamara	4 - James Robinson	Melvin Gordon
1a - Christian McCaffrey	2a - Rhamondre Stevenson	3 - Dameon Pierce	4 - David Montgomery	Raheem Mostert
1a - Josh Jacobs	2b - Dalvin Cook	3 - Ezekiel Elliott	4 - Kenyan Drake	Rachaad White
1a - Derrick Henry	2b - Aaron Jones	3 Deon Jackson	4 - Gus Edwards	Tyler Allgeier
1b - Breece Hall	2b - Jamaal Williams	3 - D'Andre Swift	4 - Khalil Herbert	Zonovan Knight
1b - Saquon Barkley	2b - Jonathan Taylor	3 - Travis Etienne	4 - James Cook	Javonte Williams
1b - Joe Mixon	2b - Clyde Edwards-Helaire	3 - Najee Harris	4 - J.K. Dobbins	Chase Edmonds
1b - Nick Chubb	2b - Leonard Fournette	3 - Samaje Perine	4 - Cordarrelle Patterson	Michael Carter
1b - Tony Pollard	2b - Miles Sanders	3 - D'Onta Foreman	4 - Cam Akers	Darrell Henderson
1b - Kenneth Walker III			4 - Jerick McKinnon	Kareem Hunt
			4 - Devin Singletary
			4 - Brian Robinson Jr.
			4 - Damien Harris
			4 - Isiah Pacheco
			4 - Antonio Gibson
			4 - Eno Benjamin
			4 - Chuba Hubbard
			4 - Latavius Murray
			4 - Jeff Wilson
			4 - AJ Dillon

Quick analysis: There were not 12 “RB1s” in 2022, there were only 10. Also based upon these tiers there were 9 “RB2s”. We commonly use top 12 or top 24 as analysis groupings when that is not the correct level of aggregation. I find this a profoundly interesting viewpoint when reviewing trades as the RB11 is commonly viewed as an “RB1” when it was shown here that that in 2022 it was actually an “RB2” disguised as an “RB1” based on its numerical rank. Also, this highlights the scarcity of the position by only identifying 19 players to fill 24 RB1/RB2 spots.

In the last visualization of the day, I show a bar graph breaking down the average fantasy points per game by point type, color-coded with dark blue representing receiving touchdowns, middle blue representing receiving yards, light blue representing receptions, dark green representing rushing touchdowns, light green representing rushing yards, and red representing fumbles. 

Analysis: It is interesting to note how certain RBs, such as Nick Chubb and Kenneth Walker III were able to produce solid fantasy numbers despite limited receiving production. Additionally, this also highlights the limitations of rostering running backs`1s who are primarily used in a 'running only' role in a half or full PPR format. For instance, Miles Sanders, Ezekiel Elliott, Jamaal Williams, and D’Onta Foreman all had solid rushing per game rushing production but ranked lower in points per game due to their lack of involvement in the receiving game. It is quite apparent that when looking for a high producing fantasy running back, receiving should be part of their repertoire unless they are the elite of elite runners on a run happy offense.

Conclusion

In conclusion, utilizing statistics and visualization can be a valuable tool in gaining insights on fantasy football. We've taken the first steps in understanding running back fantasy production from 2022. However, there is still so much more to explore as we continue our deep dive into the running back position. Join me in part two of our fantasy football analytics series as we delve deeper into the running back landscape and uncover new insights that can help us make more informed decisions in our drafts.

*This blog post was enabled by ChatGPT. The text was generated by me, and the content is my own, but some sentences and wording were provided by the model. I take full responsibility for all information produced in this blog. More information about OpenAI and their technology can be found at https://openai.com.*