August 9, 2017 by Anthony Staggs
The NFL is a game of creating mismatches, and find ways to exploit them in both the run and the pass game. Run games are about creating a numbers game advantage to a particular side, and trusting your running back to make one man miss in most schemes. The passing game, however, suffers from playing at a numbers disadvantage on most plays, as everyone pre-snap knows the number of eligible receivers.
Typically, the passing game creates mismatches in a variety of ways. Whether it is through play action, speed of receivers, size of tight ends or wideouts, or simply through play design, NFL teams work to exploit mismatches throughout the game. Last year, 4.51 players went out in pass routes on passing plays, so in rough terms, 1 of every two plays will feature an extra pass blocker, while all five eligible receivers (combination of wide receivers, tight ends and running backs) will be out in patterns on the other.
Some teams will even avoid the best corner on another team depending on their reputation and what they have put on film in previous games. Aaron Rodgers has done this numerous times in his tilts against Richard Sherman and Seattle as well as in other matchups throughout the league. Shadow corners are rare in todays NFL with just a handful of players tasked with following around an opponents top wideouts, the ones that do though are typically worth avoiding. Among all corners who saw at least 400 snaps in coverage last season, just 17 of them allowed a pass rating of sub-80 into their coverage according to Pro Football Focus.
So how else can offenses pass the ball more efficiently in this day and age? Well one idea is to attack the backside of the defense typically covered by a defenses second corner. Commonly, a team’s top corner will lineup in what is typically referred to as the left cornerback or strong side corner as most quarterbacks are typically right handed. In the NFL today, there is only one quarterback who throws with his left hand and that is Kellen Moore, and for the sake of good football, let’s hope we don’t have to watch him play on Sunday’s anytime soon. The rest of the league’s quarterbacks are all right handed and primed to attack the left side of the field on backside in-breaking routes. Tom Brady is a master of the quick shotgun fake to the running back to pull up the linebackers before quick striking a slant on the backside for chunk plays, but that’s just narrative, so lets examine the metrics in depth.
While passing to the short left and right each have similar success rate percentages in relation to league average (at 50% success rates each), there are some signs that attacking the backside is more effective. Using one of our tools, Passer Rating (by Target Depth & Location), we can examine the passer rating for each zone on the field. Passing short left yielded a 3% greater passer-rating league wide than short right, and an even starker contrast on third down where short left yields a passer rating of 87 compared to 81 when throwing short right. Now lets use this tools dynamic updating to look at some other statistics like yards per attempt and completion percentage.
On all passing plays, short left yielded a completion percentage of 1% higher than that of short right. Looking at yards per attempt yields a .35 higher total for short left than short right. Touchdown rates are nearly identical no matter the side of throw with short right having just a .006% advantage. Interception rate is slightly better for short left, over the course of 6,140 attempts to the short right, it would shave about 3 interceptions off the 82 thrown in that direction. There are signs that teams need to throw more to the backside than the front side that is attacked on 3% more of all attempts.
Let's look at a few examples from 2016. For this analysis, let's ignore passes to RBs, even though those are extremely strong plays which should be utilized more often. Looking only at passes to WRs and TEs, there are several young quarterbacks who excelled throwing to their backside (short left) but who need their coaches to call more plays to maximize targets to that location. Marcus Mariota posted a 8% stronger success rate short left as compared to short right, yet he threw almost equally (122 att vs 119 att) to the short left as he did to the short right. Dak Prescott was 12% above average throwing to the short left, and his success rate was 3% better to the short left than the short right. He posted a 113 passer rating short left vs a 93 passer rating short right. Yet he threw threw 154 times to the short right (to WRs and TEs) and only 111 times to the short left.
In addition to attacking the left, teams should continue to attack the middle of defenses. Attacking the middle of defenses both short and long yields the highest passer ratings in their sets. In terms of league wide success rate, attacking short middle is six percent more effective than attacking either side of the field. When throwing deep, teams are successful 50% of the time over the middle, but when going deep left just 37% of the time and 38% to the right. Short and deep middle also produces significantly higher completion percentages, yards per attempts, and touchdown percentages than compared to either direction. Offenses need to continue to find players that can help them over the middle as the tight end position is becoming more and more of a pass catching threat. Overall, a teams success rate is correlated to passing yards in a big way, as seen in the below chart.
NFL teams can flip their top wide receiver to the other side of the formation (if he's productive in on that side of the field) in order to use their skill sets in a more efficient manner. All pass catchers can work on getting more comfortable over the middle, as when this area is attacked it typically yields big results for the passer and the receiver. Each team can examine their past success rates to different areas of the field and increase their efficiency by spreading their pass attempts more to the right side, even if it is sometimes to the disdain of their number one receiver. In a never ending quest to increase efficiencies of passing games, teams can continue to scheme how to get players open to all sides of the field, but attacking the left and middle of the field can help teams be even more efficient.
About the author: Anthony Staggs (follow: @PyroStag) was crowned co-winner of the Sharp Football Stats 2017 Writing Contest. He will share articles featuring his analysis throughout the 2017 NFL
August 14, 2017 by Warren Sharp
Hopes are high for the Titans this season. Their win total was set at an incredibly high 9.5 wins, before being bet lower down to 8.5. But that alone is a massive increase from the 5.5 projected wins the Titans were lined at heading into the 2016 season. Off of 3 wins in 2015, Marcus Mariota's rookie season, expectations were for another subpar season. But the team went 4-0 in games decided by a FG or less, and instead of splitting those games and finishing 7-9, the team won all 4 and finished 9-7.
It would be a massive understatement to suggest that Titans fans are optimistic about the season. They believe the team has found the franchise star, their foundation and cornerstone for years to come. And with added weapons at the receiver position, there should be no reason to believe the Titans won't improve in Mariota's 3rd season. But there is one element of Mariota's game which has been lacking. And if he fixes it this season, look out.
Marcus Mariota has been terrible when passing to his right. He was bad in 2015, his rookie year, and he was bad his sophomore year last season. How bad was 2015?
Unfortunately, 2016's stats look eerily similar, but more egregious:
The main thing you'd want to consider immediately is sample size. But in over 300 attempts to his right over two seasons, this isn't a sample size issue. Mariota's success everywhere else, in over 450 attempts, is a consistent 66% with almost 8.7 YPA and well over 100 rating. To the right, that completion percentage drops to 58%, the average yardage is 2.3 yards per pass worse, and the rating is over 20 points worse.
Here is how Mariota's passing looked visually, thanks to the dynamic Receiving Success Rate Over Average (SROA) tool at Sharp Football Stats:
While first and second down still reflected these bad patterns, third down was especially concerning. Mariota fired over 60 attempts to his right on 3rd down and just over 25% recorded first downs. To his left, he fired 57 attempts on 3rd down and 58% resulted in first down. That's a massive difference. If you were defending Mariota on 3rd down, you'd try to flush him to the right and hope he'd throw to his right, rather than allow him to throw to his left. Here is what his SROA looked on 3rd down:
And when the game is close, within 1 score up or down, Mariota's success to the right was even more below average. In almost 100 attempts to the right in one-score games, Mariota posted a 35% success rate, whereas it was 20% better (55%) when passing to his left:
I wrote about this issue from Mariota in my 2017 Football Preview. When trailing, Mariota threw nearly 100 attempts to the right side of the field and averaged a 67 rating (31st in the NFL).
Despite the consistency in problems from 2015 to 2016, I still was hoping this was an anomaly. Some type of odd occurrence that, despite the stats (both basic and advanced) showing a major issue existed in his ability to throw to his right, there was no concern.
But then I watched his week one preseason game. Yes, he threw only 3 passes. But the only one he threw deep right was very inaccurate and nearly intercepted. Hopefully this was just another random throw which shouldn't be read into. But it was enough to prompt me to watch every single pass attempt to the deep right in 2016.
However, I decided to study the film of every one of his passes to the deep right. He clearly struggled to his right in general, but I focused on the deep right passes as these are where his larger gains would occur, and where the misses would result in the most negative-EV. Below is a reel of most of Mariota's passes to the deep right from 2016. I cut only a few throws which were from extreme bootlegs/wild scrambles or had him tossing it downfield just before running out of bounds. You can see for yourself some of the issues. The first pass is the one from the 2017 week 1 preseason game this past weekend, followed by his 2016 attempts:
Mariota has better receivers this year, in Corey Davis, Eric Decker and Taywan Taylor. Last year, the receiving options left a lot to be desired. Perhaps this will help Mariota in 2017. Because while many of the passes in the above reel were not accurate, rather than receiver drops, perhaps better receivers would have run better routes and put Mariota in more positive-EV situations to complete the pass.
However, it should be noted that we're not isolating only these passes to the right and saying Mariota's receivers were bad. Mariota and these receivers were outstanding to the left. They were far above average. Exceedingly so. But there were notable struggles to the right.
There are other areas Mariota needed to work on apart from passing to the right, although that is the most critical. On early downs in the first quarter, due in part to play scripting, Mariota recorded a 100 passer rating, averaged 8.2 yards per attempt and completed 69% of his passes. However, after the first quarter, his early down passer rating dropped to 81 (25th in the NFL), he averaged just 6.3 yards per attempt and completed just 57% of his passes. Those numbers became even worse when trailing, as his 68 passer rating ranked 34th, he averaged 5.6 yards per attempt and he completed just 54% of his passes. Early down passing correlates tremendously toward winning in the NFL, and Mariota’s passing on early downs must improve, particularly after the script runs out and it becomes the 2nd quarter.
But let's not paint Mariota in this totally negative light. What are some positives from the passing game in 2017? Unlike 2016, where the Titans played the 14th rated pass rushing defenses, which helped parlay into the 4th best defenses preventing explosive passes, it’s much easier this year. The Titans play the 26th rated pass rushing defenses and likely as a result, these defenses allow the 2nd highest rate of explosive passes. In Nashville, that’s literally music to the ears of Mariota, Matthews and Davis. Matthews was the most dominant deep threat for the Titans last year, and delivered a 138 rating deep left and a 141 rating deep middle.
While I'm extremely optimistic in Mariota as a passer and believe the Titans are certainly headed in the right direction (no pun intended) with him under center, I really want to see him develop his ability to be consistent when throwing to his right. If Mariota can develop those traits and deliver more efficiency on those attempts to the right, the sky is the limit for him and the Titans offense in 2017.
About the author: Warren Sharp ( @SharpFootball ) owns and operates SharpFootballAnalysis.com and SharpFootballStats.com
Customize | Visualize | Process | Retain
For best results, view on computer (not mobile devices)
August 7, 2017 by Connor Allen
It’s more important to pass efficiently than to run efficiently, because passing efficiency contributes more towards winning games in this modern era of football than does rushing. Some have confused this point as that teams should be “pass heavy”. They can still be balanced in their approach, but their efficiency from passing must be strong. And that is why studying passing efficiency is an extremely important investigation to undertake.
To aide in that investigation, Warren and I created called a visualization last week on Sharp Football Stats. This tool introduces for the first time ever, dynamic Success Rate Over Average (SROA). Previously, there has never been a way to track whether a quarterback or a receiver is performing above average based upon changing (dynamic) on-field situations, such as down, distance, lead/deficit, time remaining, field location, etc.
Not only does this tool dynamically show us these averages and a particular receiver or quarterback’s performance, it also shows it directionally, meaning deep left vs deep right, for 6 different zones of targeting.
The way it calculates the “over average” portion is simple. For example, if Amari Cooper had a success rate of 50% deep right, and the league average success rate deep right is 38%, his success rate over average (SROA) would be 12%.
Going team by team, I used this tool to make 70 observations on how each team in the AFC can pass more efficiently. I’ll tackle the NFC in the next installment.
1. When Derek Carr played a full game (Week 1-15), Oakland was surprisingly poor when throwing short passes grading negatively in all areas (-6% left, -3% middle, and -7% right.).
2. On the contrary, Carr was very good when throwing deep with a 12% SROA deep middle and 8% SROA deep right. The Raiders need to continue taking deep shots and not dink and dunk as they were much more successful when going deep.
3. This deep success was even more apparent on 3rd down. When they took shots downfield on the Raiders had a 22% SROA deep middle and a 20% SROA deep right.
4. Michael Crabtree has been Carr’s security blanket but was just 1% more successful than the league average on 3rd down and most of his success came left, both deep and short. He needs to be targeted more to the left instead of right (9 targets left vs 23 targets right).
5. Crabtree is also the only main pass-catcher who could consistently be successful in passes of 15 yards or less (3% SROA left, 2% SROA middle, 2% SROA right).
6. Amari Cooper was very good deep with a 16% SROA deep left and 13% SROA deep right. Cooper needs to improve on passes of 15 yards or less as he graded negatively in all areas.
7. The Broncos must target the middle of the field more in the red zone because it’s the soft spot of defenses. In the red zone the league average is a 51% success rate, highest among all areas.
8. The Broncos need to throw deep more because both of their top receivers excelled there. Demaryius Thomas posted a SROA of 2% or greater in all deep zones.
9. Similar to Demaryius, targeting Emmanuel Sanders deep middle (17% SROA) and left (21% SROA) was highly successful. Sanders was more successful than Demaryius with short passes and should be the Broncos preferred target within 15 yards.
10. Both Sanders and Thomas’ deep dominance lead me to believe the Broncos would be better off by taking a shot with Paxton Lynch under center or dialing up more long plays with Trevor Siemian who actually saw his highest SROA percentage when throwing deep.
Kansas City Chiefs
11. The Chiefs were surprisingly successful both deep right and left with Alex Smith at the helm. However, Alex Smith was atrocious deep middle (-21%) and when throwing short he was below the league average in all areas.
12. Travis Kelce is by far the Chiefs most efficient option all over the field with a 13.2% SROA overall. He needs to be targeted even more, becoming the true focal point of the Chiefs offense.
13. Kelce is especially efficient to the left side of the field posting a 21% SROA short left and 22% deep left. He was targeted short left, middle, and right at a similar rate but the Chiefs should try and target him more to the left because of his hyper-efficiency.
14. Tyreek Hill performed at exactly the league average overall and was very inconsistent by field zone. A highly explosive player, Hill needs to work on consistently helping his team produces successful plays if he wants to develop into the true #1 WR the Chiefs expect.
Los Angeles Chargers
15. The Chargers were slightly above average when throwing short and were most successful when throwing deep right, posting a 4% SROA.
16. When the Chargers throw deep left or middle it should be to Tyrell Williams who posted a 12% SROA left and 17% SROA middle. With Keenan Allen healthy and the growth of Hunter Henry both will be drawing defensive attention, meaning Williams should continue his deep success this season.
17. Hunter Henry was spectacular in the red zone last year albeit on a small sample size. He posted a perfect 100% success rate in 3 different zones and a 75% overall success rate inside the 10. The Chargers need to expand his role instead of relying on father-time Antonio Gates.
18. Melvin Gordon was very successful at catching passes out of the backfield (6% left, 18% middle, 4% right) and was used most on the outsides (23 targets left and 26 targets right). The Chargers should try and get Gordon the ball more in the middle of the field because it has the highest league average success rate. He only received 8 targets there and had an 18% SROA.
New England Patriots
19. Tom Brady’s SROA last season was extraordinary, grading positively everywhere combining for 2.9% above the league average. It’s scary to think he will likely get better when throwing deep with the addition of Brandin Cooks.
20. Julian Edelman was successful at all depths right and left but struggled in the middle. However, on 3rd down he was very successful as Brady security blanket with a 10.8% SROA overall.
21. Rob Gronkowski dominated all over, especially short middle (27%). He was also very good deep left (80% Success Rate, 42% SROA). Gronk’s success rate is so good that he should be targeted more instead of their stable of pass-catching running backs.
22. Dion Lewis was surprisingly underwhelming in the pass game and only graded positively in SROA short left. He had zero successful plays on passes longer than 15 yards (0/5).
23. James White had 8% SROA short right and was 100% (2/2) deep right. To the contrary, he was -9% short left and 0/3 deep left. I’m not sure why he was so much better to the right than left but it should be something the Patriots take note of.
24. For not being known as a good deep passer Tannehill was surprisingly fantastic in comparison to the rest of the league. Tannehill had a 3% SROA deep left, 17% deep middle, and 9% deep right. He needs to target deep left less as this was least successful out of all deep areas but was the most highly targeted by 20 attempts.
25. Devante Parker was very sporadic all over but performed well deep middle and deep right. With Tannehill at QB he was better than with Matt Moore according to SROA.
26. Jarvis Landry had a 9% SROA when targeted short left, but a -5% on his other short zones. The Dolphins need to go away from targeting him on 3rd down as he was poor everywhere except to the left (16% SROA). This may be difficult as Landry is clearly Tannehill’s security blanket but targeting him anywhere besides left is a –EV move.
27. LeSean McCoy was very good at passes short left and right (6% and 5% SROA) but struggled over the middle -11% SROA. Having Sammy Watkins healthy and passing more on offense due to the change in coaching should open up the middle of the field more for McCoy to take advantage.
28. Tyrod Taylor was significantly better at passing right (both short and deep) than to the rest of the field (3%), and attacked that side often attempting 55% (52) of his deep passes there. The Bills should continue this especially now that Watkins is healthy. They also need to attack the deep center more because it was their most successful part of the field (17% SROA) yet only had 9 attempts there.
29. However on 3rd down and 5 yards or less the Bills didn’t complete any of their deep attempts and were best when they threw to the outside posting a 2% SROA short right and 14% SROA short left.
30. While he was a very successful runner, Mike Gillislee was awful in the passing game posting a -22.4% SROA in all. Now that he’s on the Patriots, who have an incredible arsenal of pass-catching options, it’s unlikely he will be used in the pass game.
New York Jets
31. Speaking to the few players the Jets retained this offseason, targeting Matt Forte out of the backfield was good short left (5% SROA) but poor short center (-7% SROA) and right (-7% SROA).
32. For all the hype Bilal Powell has gotten as a great pass-catching back, he wasn’t very efficient last season posting a -6.4% SROA overall including a -39% SROA short center.
33. Quincy Enunwa was horrible on passes less than 15 yards grading negatively in all directions, but on plays longer than 15 yards he graded positively in all areas. An offense can’t rely entirely on long passes to one player but it’s certainly encouraging that the Jets have one potentially reliable option.
34. Big Ben was surprisingly average everywhere besides the right side (7% short and 12% deep).
35. On 3rd down Ben was good to both the outside zones and was able to do well taking deep shots but he struggled in the center of the field deep and short. This was probably a result of the lack of a tight end and solid #2 receiver.
36. Antonio Brown was good everywhere except short left and deep center posting a 1.7% overall SROA.
37. Brown had the 2nd highest percentage of his targets go short left, yet he had his worst SROA there (-11%). He needs to be targeted more short center and to the right side of the field where he scored positively in both (2%, 8%).
38. Le’veon Bell was really good short right (16% SROA) and that’s where he was targeted the most (44) by the Steelers. He was surprisingly very unsuccessful in the middle of the field compared to the rest of the league (-11%).
39. Cody Kessler was actually decent last season. I didn’t expect this conclusion coming into the process but in weeks he started and wasn’t injured (Week 3,4,6,9,10) he graded positively in half of the directional zones, more than many other starting quarterbacks. He was best when throwing short center (8% SROA) and deep right (10% SROA).
40. Corey Coleman was horrible last season grading negatively in all areas except deep center, where he was exactly average. He has a lot of growing to do as a player this offseason if he wants to help turn the Browns offense around.
41. Duke Johnson, a back known for his pass-catching skills was surprisingly poor (-2.4% SROA) and was especially bad on the outsides. He posted an 8% SROA short center exploiting the most vulnerable part of the field.
42. Isaiah Crowell was the exact opposite as Duke, performing slightly above average on the outsides (3% right, 0% left) yet was awful when passed to in the center of the field (-27% SROA).
43. While the Bengals offseason moves have suggested a switch to a “quick pass” approach, Andy Dalton was very successful when passing deep center (18% SROA) and deep left (5%). Dalton was very average when throwing short, posting a SROA of 1% left, -3% center and 6% right. John Ross and Joe Mixon will have to step up in the short passing game if they are to succeed at this new approach.
44. A.J. Green was really good deep center (50%) and left (17% SROA). He also had a 13% SROA short right which happened to be where he was targeted most.
45. Green was especially successful in the red zone where he had an SROA of 75% to the right and 67% to the left. The Bengals should continue to target him in the red zone, expanding their repertoire and targeting him in the softest part of the field, the center.
46. On the other hand, Tyler Eifert exploited the center of the field in the Red Zone posting a 75% success rate, 24% above the league average.
47. Like other elite tight ends I talked about previously, Eifert was much more successful than the rest of the league on passes all over the field with a 9.9% SROA.
48. Generally regarded as an elite deep threat, Mike Wallace was actually 24% below the league average on throws deep left and deep right, only grading out positively deep center.
49. Joe Flacco had a poor success rate all over the field (-4.4% SROA), especially deep.
50. Steve Smith was a very successful player for the Ravens last season (5.2% SROA) and will be missed by an already unsuccessful team.
51. Breshad Perriman was woeful last season (-12.7% SROA) and was one of the most inefficient receivers in the league. Jeremy Maclin will need to step into the #1 role because Perriman didn’t seem capable last season.
52. While most people thought of Marcus Mariota as an efficient QB last season, he posted a -1.2% SROA all over.
53. On early downs Mariota’s SROA was actually even worse (-3%), only grading positively deep left.
54. Mariota targeted deep right just as frequently as deep left (42 targets) yet had a positive 12% SROA to the left and a -12% SROA to the right.
55. Rishard Matthews was very good last season everywhere (5.5% SROA) except deep right. However I don’t know if Matthews is to blame as Mariota was bad at throwing there and Matthews had a positive SROA everywhere else. Hopefully the additions of rookie Corey Davis and Eric Decker can help Mariota.
56. Andrew Luck was a prolific deep passer posting a 25% SROA deep right and a 13% SROA deep left. He graded negatively deep center but it was only by 3% and deep center has the highest average of all the deep zones (50%).
57. Luck was very average when passing short, all of his SROA percentages ranging between 0-3%. He needs more out of Donte Moncrief when passing short.
58. When playing with Luck, Moncrief was pretty poor within 15 yards or less with a 0% SROA short left, -7% short center, and -11% short right.
59. TY Hilton needs to be targeted relentlessly by Luck, he is more successful than the league average everywhere except short right and had a combined 4.4% SROA.
60. Jack Doyle was very successful when targeted center or right and had a combined SROA of 13.4%. Doyle should be Luck’s main focus short center and right, something Luck struggled with overall last season.
61. Blake Bortles was poor everywhere (-4.1% SROA) but was especially awful deep left (-14%) and center(-13%).
62. On 3rd downs Bortles was similarly as poor with -3.8% SROA all over.
63. In the red zone Blake Bortles only glimmer of hope was his 30% SROA short center (17 attempts). Bortles was bad last year in pretty much every situation. The Jags need to hide Bortles with a heavy dose of running and some good defense if they want to win games.
64. Allen Hurns graded positively both deep right and left, an area Bortles really struggled with. Hurns should be targeted more deep on the outsides instead of the center (-30% SROA) next season.
65. Allen Robinson needs to rebound to his 2015 as last year he was painfully unsuccessful everywhere (-7.6% SROA) and was especially awful when going deep.
66. Marquise Lee had a great season and Bortles needs to look for him more within 15 yards. He was one of the only Jacksonville targets who was able to be successful in all short locations, posting a 6% SROA left, 5% center, and 13% right.
67. Leonard Fournette should be given a chance for pass-catching duties right away based on how poor T.J. Yeldon was last year. Yeldon had a -11.4% SROA and graded negatively in every zone he was targeted.
68. Deandre Hopkins was good short left (5% SROA) and center (7% SROA) despite abysmal QB play. When going deep Hopkins was extremely poor (-13% SROA or worse in all deep zones), which was likely due to quarterback play.
69. Lamar Miller should be targeted more often short left (10 targets, 10% SROA) and short center (10 targets 23% SROA) instead of short right (27 attempts -13% SROA).
70. Will Fuller is expected to miss 2-3 months due to a broken collarbone but his presence may not be missed that much. It’s important to have a field stretcher like Fuller, but he was not efficient posting a -11.8% SROA everywhere only grading positively short left (10% SROA).
About the author: Connor Allen (follow: @Fantasy_Matrix) was crowned co-winner of the Sharp Football Stats 2017 Writing Contest. He will share articles featuring his analysis throughout the 2017 NFL season.
Copyright © Sharp Football Stats. All Rights Reserved.