Quote:
Originally Posted by DodgerIrish
I can't read it, seems like throwing to your running back has the duel benefit of getting your quarterback in a rhythm as well as keeping your running back out there for identical looks.
Isn't a high % pass that gets 4+ yards every play a good thing if your philosophy is built around continuously getting first downs? A flare out to the running back has to have a higher expected yards per play than like a ton of bubble screens, no?
I don't have a subscription to any of this stuff you guys are looking at to see what % of time guys lineup where or similar stuff.
Will just cut and copy the article here, can give you my thoughts to what you wrote but figured Ill give you a chance to actually look at the article first
couple other interesting links data point wise regarding some potential counter points/concerns of the article
https://twitter.com/benbbaldwin/stat...796586496?s=20
https://twitter.com/KentWeyrauch/sta...654892544?s=20
https://twitter.com/benbbaldwin/stat...371440640?s=20
Move over, rule of 53. You have company.
Brian Schottenheimer has identified another 50-something target. When discussing running back Chris Carson’s role in the passing game in the coming season, the Seahawks offensive coordinator was clear on the number of passes he’s hoping come Carson’s way:
“We need to get that number up around the fifties.”
Schottenheimer expanded on his thoughts, saying that ideally, “We are still getting his yards on the ground, and he’s still helping us in the passing game … People are understanding that getting the ball to your backs in space in the passing game is a good thing.”
These comments bring to mind Packers head coach Matt LeFleur, Broncos offensive coordinator Rich Scangarello, and Vikings advisor Gary Kubiak, all of whom have stated their intentions of emphasizing their teams’ running backs in the passing game. The Houston Texans are fresh off trading a conditional fourth-round pick for Duke Johnson, a receiving specialist, and the Jets signed the versatile Le’Veon Bell to a $52.5 million contract over the offseason.
But here’s the thing. Targets to running backs are by far the least efficient type of pass. This is important, so let’s say it again.
Relative to passes intended for other positions, running back targets are not efficient.
Here’s a basic overview of Expected Points Added (EPA) per play by targeted position on first or second down. (Throughout this piece, we’re going to be looking at first and second downs only. The concern with third downs is that running back targets on those downs could be disproportionately check-downs that are doomed to failure. In an effort to level the playing field, let’s look at these early-down attempts.) As always, a huge thank you to nflscrapR for making the data readily available.
https://cdn.theathletic.com/app/uplo...r-1024x664.png
Note the huge difference between running back targets and other targeted positions. What if we just looked at the percent of plays that gained positive EPA? Here’s what that looks like:
https://cdn.theathletic.com/app/uplo...s-1024x664.png
This looks better for running backs, but doesn’t exactly scream for teams to be diverting targets away from their tight ends and wide receivers towards their running backs. Plus, there’s no reason to privilege success rate over EPA anyway. The point of the game is to outscore one’s opponent, and EPA does a better job of capturing the impact of a play than a simple measure of whether a play’s EPA is positive (i.e., success rate).
So why are running back targets so inefficient? There are two factors at work.
Flaw No. 1: rarely attacking downfield
Let’s now group wide receivers and tight ends together because the results of their targets are similar and it makes the figures easier to read. Here’s what the distribution of targets looks like for running backs (red) and the other positions (blue):
https://cdn.theathletic.com/app/uplo...t-1024x664.png
A great deal of running back targets come within five yards of the line of scrimmage. (Note: I’ve truncated the figure outside of negative-10 and plus-50 air yards to make it easier to read, but these targets are rare.) This brings us to the next point: targets close to the line of scrimmage are the least valuable targets.
Flaw No. 2: worse at attacking downfield
Even when backs do get downfield, passes targeting them are less efficient than passes targeting other positions.
https://cdn.theathletic.com/app/uplo...1-1024x664.png
Within five yards of the line of scrimmage, targets to running backs are about as efficient as to other positions. However, the high-value part of the field is 10-20 yards downfield, where throws are likely to be completed for first downs and thus have a high expected value. And throws to wide receivers and tight ends are much more efficient passes in this region.
One difficulty when assessing running back targets is that some of them come late in the play to avoid a sack. Checking down to a running back is unquestionably better than getting sacked, but the above figure is important because even if we ignore everything within 5 yards of the line of scrimmage, running back targets are still the least efficient.
In sum: running backs are less likely to earn high-value downfield targets and less likely to convert on those attempts when they are targeted.
What about the Seahawks?
The numbers above show averages from all 32 teams, so one might worry about how they would generalize to the Seahawks. Perhaps Schottenheimer saw something in 2018 that convinced him Seattle had an edge in throwing to running backs. Let’s chart every single early-down target the Seahawks had in 2018, by position:
https://cdn.theathletic.com/app/uplo...s-1024x664.png
Each dot represents one play, and the bars are the average of each targeted position (the figure caps all plays at plus-3 and minus-1.5 EPA to make it readable, but the true values are used to calculate the averages shown). The basic pattern of running back inefficiency for the Seahawks is similar to the rest of the league, but — because they have Russell Wilson, one of the best deep-ball passers in the league — Seattle was substantially above average on passes to tight ends and wide receivers. In other words, the league as a whole is more efficient at throwing to non-RBs, and the difference is even more pronounced for the Seahawks.
Or to put it another way: the comparative advantage that the Seahawks have in Wilson is diminished when he isn’t passing to his tight ends or wide receivers. And it’s not like the Seahawks haven’t already been throwing to running backs. Their rate of running back targets on first and second down was 10th-highest in the league in 2018.
What is going on?
I find the case above to be compelling, but you don’t have to take my word for it. Using a bigger sample, PFF’s Timo Riske found that throws to WRs and TEs are more efficient at depths of target greater than five yards, a finding that Anthony Amico has since replicated. PFF’s Eric Eager and George Chahrouri showed that even when running backs line up on the slot, they are still less efficient than targets to WRs or TEs lined up in the slot, and this doesn’t appear to be driven by check-downs: it continues to be true when focusing on passes within two yards of the first down marker. Riske also examined the most commonly-run routes run by running backs and found that they are substantially less efficient than the typical routes run by slot or outside receivers. Still skeptical? Here is all the code used to create the figures above. You can do it yourself!
https://twitter.com/benbbaldwin/stat...242536455?s=20
The information is out there, and it is comprehensive. The best catchers of the football are the players who play positions primarily geared toward, well, catching the football. Perhaps this shouldn’t be surprising. Expecting running backs to catch passes downfield as well as wide receivers would be a little like expecting pitchers to hit as well as position players. Running backs are further disadvantaged on pass plays by lining up behind the line of scrimmage most of the time and thus taking longer to get downfield.
A couple caveats are in order: First, this is one area where tracking data — which teams have access to but the public does not — would be helpful. Excluding short passes to running backs that come late in the play could make the picture look less ugly for running back targets to some extent. Because of the inefficiency of running back targets on deeper passes and the work that PFF has done incorporating additional context into running back targets, I am skeptical that this would bridge the gap, but it’s a possibility that we should acknowledge. The evidence presented here isn’t as strong as, say, that NFL teams are too conservative on fourth down.
Second, if Seattle’s plan is to replace some running back carries with passes to running backs, that’s a net gain; just not as big as the gain would be if different positions were targeted instead. And we should be skeptical that this is the plan given Schottenheimer’s comments about Carson still getting his rushing yards (and his place as a coach on a Pete Carroll-led team).
Whenever I’ve asked why the Seahawks think involving their running backs more in the passing game is a good idea, a common response is that the Patriots and Saints have been successful while heavily involving running backs in the passing game. And indeed, Schottenheimer mentioned New England’s James White and New Orleans’ Alvin Kamara in his press conference. But while both rank highly in the percentage of early-down targets to running backs, so too do teams like the Giants, Jaguars and Raiders, who aren’t exactly high on the list of offenses to emulate. In fact, there’s a small negative correlation between a team’s EPA per play on first and second down and the percent of passes to running backs on first and second down (-0.08). This negative correlation holds even when looking at the share of passes over 5 yards downfield that targeted running backs.
It’s not hard to find great modern offenses that don’t heavily involve their running backs in the passing game: In 2018, the Chiefs, Steelers and Falcons were all in the bottom 10 in percent of early-down passes to running backs and top-8 in overall EPA per play on those downs. The Rams led the league in EPA per play on first and second down while being No. 21 in early-down running back target share. It’s better to be efficient than not efficient when targeting running backs, but there’s no evidence that doing so at a high rate benefits a team.
Conclusion
Passes to running backs aren’t inherently bad. The results are certainly better than rush attempts or sacks, and having the threat of a running back in the passing game — forcing the defense to account for every eligible receiver — can only be a good thing. To borrow an illustration from basketball, recent The Athletic addition Seth Partnow’s thoughts on mid-range shots are a useful parallel: “You want to take just enough mid-range shots that the defense honors that shot, and no more. Each additional one you’re taking, you probably could be getting a more valuable shot somewhere else.”
The same appears to be true for passes intended for running backs. The evidence suggests that intentionally diverting targets away from other positions in favor of running backs is harmful, especially if your quarterback is Russell Wilson.
Last edited by GrapesofRath; 08-21-2019 at 05:35 PM.
