As fantasy drafters, we are constantly striving to earn a positive return on our fantasy investments. This is for sure an impossible task, but nonetheless, we must do our very best to come as close to achieving this goal as possible. Identifying players that are likely to regress is an important step in this process.
In an attempt to find running backs that are regression candidates, I did some digging through points per opportunity (PPO). For the purposes of this post, ‘opportunity’ consists of rushing attempts as well as targets in the passing game. We will be reviewing the metric in an attempt to find unsustainable fantasy efficiency. I pulled data for the 2013 through 2015 seasons and found the average points per opportunity (based on PPR scoring), for the top 40 RBs from each season, to be .83.
PPO rates for the top 40 running backs from last season are included in the table below:
Notice anything? Pass catching backs have significantly better rates in terms of PPO. This is further evidence as to why they are so useful in PPR formats. Obviously, that’s a fairly intuitive concept; passing plays are more efficient and a higher percentage of these player’s production comes from a higher scoring play. But it does help in driving home the argument that pass catching backs, such as Theo Riddick and Charles Sims, can be useful fantasy assets even with limited touches. Though real-life efficiency doesn’t correlate strongly with fantasy success, measures such as PPO certainly have a place in fantasy analysis.
James White was extremely efficient in 2015, however, his overall usage was limited. With Dion Lewis returning, his role will remain ancillary and as such, he doesn’t figure to be a major contributor in 2016. In contrast, Theo Riddick has carved out a substantial role in the Detroit offense. He finished the 2015 season with 179 PPR points and was the 18th ranked running back based on total points. Is his rate of 1.46 points per opportunity sustainable? Perhaps. In 2014 (a season in which he saw only 20 attempts and 50 carries) he was good for a PPO rate of 1.35.
Of course, we need a little more context to evaluate the above data points. Here’s how Danny Woodhead, a player of a similar profile, has fared the last five seasons, excluding a three game 2014 season:
So it looks like players of this mold are able to put up relatively consistent PPO rates, even when switching teams. So Riddick doesn’t look like a major regression candidate, especially as his volume may increase in 2016.
David Johnson would figure to regress on a per game basis. However, his overall opportunity should drastically increase, negating the impact of any negative regression he may experience.
We’ll be discussing Lamar Miller in great depth in a future podcast, so let’s instead consider a player that will see substantial usage in both facets of his offense, Mark Ingram. His 2015 rate of .94 is well beyond the average of .83. In 2014, he managed a rate of only .73 and in 2013 he was good for a dismal .66. Ingram played in 12 games last season and was on pace for 271 PPR points, had he remained healthy for the duration. This point accumulation seems unsustainable as his efficiency last season was high for his career and the league at large.
Devonta Freeman is the poster boy for players labeled as 2016 regression candidates. This was to be expected as his production in the first half of 2015 was superhuman and his points per game dropped steadily after the mid point of the season. If a smaller percentage of his overall production was not related to the run game, we might be able to make a case for his PPO remaining stagnant in the coming season. However, rushing attempts accounted for over 73% of his 2015 opportunity. A rate of .94 points per touch, with such a high volume of carries is definitely unsustainable. I should note, however, that the market has corrected for this regression, baking it into his ADP.
Karlos Williams was a break out rookie in 2015. He was able to find opportunity early in the season and had a surprisingly solid year, averaging over 11 points a game. Granted, we only have one season of data to work with, but it seems highly unlikely that he will remain as efficient in 2016. This is problematic as an expansion of his role in the Bills’ backfield is likely dependent on the health of Shady McCoy. Williams’ workload as a rusher dropped by 8 carries a game when McCoy was healthy.
Just to confirm that PPO rates, on an individual player basis, do not always remain stagnant I pulled the following table together:
The important thing to remember when performing an analysis such as this, is that regression is okay. It’s unavoidable. Players can’t always have their best season. We don’t need to avoid these players at all costs. But, we do need to make sure that we factor this regression into the price tag we place on them.