Welcome to Regression Alert, your weekly guide to using regression to predict the future with uncanny accuracy.
Returning readers, you know how this works by now, but for new readers here's the deal. Every week I take a look at a specific statistic that is prone to regression and identify high and low outliers in that statistic, and then I wave my hands in the air and shout “regression!”
But since predictions aren't any fun without someone holding your feet to the fire afterward, I don't stop there. I lump all of the high outliers into Group A. I lump all of the low outliers into Group B. I verify that Group A is outperforming Group B. And then I predict that Group B will outperform Group A over the next four weeks.
I don't get to pick and choose my groups, beyond being free to pick and choose what statistics are especially prone to regression. If I'm tracking yards per target, and Antonio Brown is one of the high outliers in yards per target, then Antonio Brown goes into Group A and may the fantasy gods show mercy on my predictions.
And then, groups chosen and predictions made, I track my progress. That's this.
In Week 2, I outlined what regression was, what it wasn't, and how it worked. No prediction was made.
In Week 3, I listed running backs with exceptionally high and low yards per carry averages and predicted that the low-ypc cohort would outperform the high-ypc cohort over the next four weeks.
In Week 4, I looked at receivers who were overperforming and underperforming in yards per target and predicted that the underperformers would outperform the overperformers over the next four weeks.
In Week 5, I compared the predictive accuracy of in-season results to the predictive accuracy of preseason ADP. Outside of a general prediction that players would tend to regress in the direction of their preseason ADP, no specific prediction was made.
In Week 6, I looked at quarterbacks who were throwing too many or too few touchdowns given the amount of passing yards they were accumulating, then predicted that the underperformers would score more fantasy points than the overperformers going forward.
|Statistic for regression||Performance before prediction||Performance since prediction||Weeks remaining|
|yards per carry||Group A had 60% more rushing yards per game||Group B has 16% more rushing yards per game||None (Win!)|
|yards per target||Group A had 16% more receiving yards per game||Group B has 10% more receiving yards per game||1|
|passing yards per touchdown||Group A had 13% more fantasy points per game||Group A has 28% more fantasy points per game||3|
Today, we close up our first prediction of the season; yards per carry, your time has come. I predicted that Group A and Group B would both regress towards each other in yards per carry and that Group B's larger workload would result in more rushing yards once they did.
Little did I know just how much each group's respective yard per carry average would regress. At the time of the prediction, Group A averaged 5.7 yards per carry and Group B averaged 3.2. Since the prediction? Well, until Derrick Henry ripped off a 72-yard touchdown with the final carry of the sample, Group B actually had a higher average than Group A. In the end, Group A finished averaging 4.2 yards per carry, Group B averaged 4.0, and Group B's larger workload carried the day.
You may have also noticed that my most recent prediction is off to a bad start. Fear not; regression operates over longer timescales. Remember, in the week immediately after my yards per carry prediction, Group A outrushed Group B by 46%.
Now, on to the prediction.
Touchdowns Follow Yards (Redux)
As I mentioned last week, yard-to-touchdown ratios are not just my favorite regression statistic, they're actually the entire reason this column exists. The idea for this feature, (including tracking the predictions in real time), stems from a pair of articles I wrote back in 2015 on how Touchdowns Follow Yards, (but Yards Don't Follow Back)
The basic idea: some players are better than others at converting yards into touchdowns. Dez Bryant is heavily involved in the red zone. Julio Jones is not. Therefore Dez Bryant scores more touchdowns per yard gained than Julio Jones does.
With that said, all of this variation occurs within a reasonable range. The most prolific touchdown-scoring receivers of all time— guys like Jerry Rice, Randy Moss, Terrell Owens, Cris Carter, and Marvin Harrison— average a touchdown for every 100 yards or so. Bryant averages 98.4 for his career.
Meanwhile, the biggest yardage monsters of all time— Andre Johnson, Henry Ellard, Wes Welker, Eric Moulds, Art Monk— average a touchdown for every 200 yards or so. Jones averages 199.4 for his career.
This is the reasonable range. A yard to touchdown ratio somewhere between 100 to 200 is sustainable in the long term, (though in practice receivers mostly tend to fall between 130 and 200 with just a few “super-scorers” down around 100-110).
But right now Julio Jones has 367 yards and no touchdowns to show for it. Based on history, we'd expect him to have reached the end zone once or twice at least by now. And on the other hand, Jordy Nelson, (historically another “super-scorer”, averaging a touchdown per 111 yards for his career), has a remarkable 6 touchdowns on just 290 yards, a ratio of 48 that suggests he has at least three more touchdowns than we'd expect right now.
Why does this matter? Because the amount of yards a receiver has today does a pretty good job of predicting the amount of yards he'll have going forward... but the amount of touchdowns he has today does a terrible job of predicting the amount of touchdowns he'll have going forward.
In fact, the amount of touchdowns a player has today holds less predictive power for how often they're going to score going forward than the amount of yards a player has today. This is because players who are producing way outside of their established yard to touchdown ratios tend to revert to their career mean.
In short, guys with lots of yards but no touchdowns turn into guys with lots of yards and some touchdowns. Guys with very few yards and lots of touchdowns turn into guys with very few yards and few touchdowns. Owners who bought and sold on the basis of yards wind up happy, owners who bought and sold on the basis of touchdowns wind up sad.
With that in mind, 53 wide receivers currently have at least 30 fantasy points in standard scoring. Here they all are ranked by their yard:TD ratio.
|41||Odell Beckham Jr||4||302||3||100.7|
|45||Marvin Jones Jr||6||280||3||93.3|
|53||Will Fuller V||3||154||5||30.8|
There are 12 receivers who are averaging fewer than 100 yards for every touchdown scored. They are Will Fuller V, Jordy Nelson, Michael Crabtree, Chris Hogan, DeAndre Hopkins, Davante Adams, Nelson Agholor, Dez Bryant, Marvin Jones Jr, JuJu Smith-Schuster, Stefon Diggs, and Jermaine Kearse. This is our Group A.
Conveniently, there are also 12 receivers who are averaging at least 300 yards for every touchdown scored. They are Adam Thielen, Pierre Garcon, Demaryius Thomas, Julio Jones, Robert Woods, Marqise Lee, T.Y. Hilton, Keenan Allen, Kelvin Benjamin, Rishard Matthews, Antonio Brown, and Danny Amendola. This is our Group B.
To this point of the season, Group A averages 53.4 yards and 0.77 touchdowns per game played. Group B averages 71.7 yards, but a paltry 0.10 touchdown per game played. As a result of that touchdown differential, Group A averages 10.0 fantasy points per game to just 7.8 for Group B, a 28% edge.
Despite that advantage, I predict that Group B will score more fantasy points per game than Group A over the next four weeks. (There's a decent chance they'll even have more touchdown receptions in the process.)
Tune in next week to continue following Regression to the Mean's bid to go undefeated this season.
More articles from Adam HarstadSee all
More articles on: AnalysisSee all
Beating the Odds: Week 7 - Zamichieli
Buy Low, Sell High: Week 7 - Bloom
WR/CB Matchups to Exploit: Week 7 - Settle