2018 ALEX: Season Review
by Scott Kacsmar
We'll have plenty of time after the Super Bowl to do statistical studies of a pass-crazed 2018 NFL season, but one of the first things to know is that this was the year of completion percentage. Teams completed 64.9 percent of their passes, smashing the former record of 63.0 percent set in 2016. Interestingly enough, the only qualified passers who did not complete at least 60 percent of their passes were rookies Josh Allen (52.8 percent), Josh Rosen (55.2 percent), and Sam Darnold (57.7 percent). Better luck (and hopefully fewer drops) next year. Sack rate (6.8 percent) was also at its highest since the 2004 season, indicating there was some trade-off between taking sacks and forcing the ball that could have spiked the numbers.
It's also safe to say that checking down for easy completions was a popular decision around the league in 2018. That likely impacted the numbers on ALEX on third down. For those new to this metric, it stands for Air Less EXpected, or ALEX for short. ALEX measures the average difference between how far a quarterback threw a pass (air yards) and how many yards he needed for a first down. If a quarterback throws a 2-yard pass on third-and-8, then that would be -6 ALEX. The best application of ALEX is to look at third down, when it's really crucial to get 100 percent of the needed yards to extend the drive.
The league-wide ALEX was +0.9 this year, the lowest for any season for which we have data going back to 2006. The previous low was +1.1 ALEX in 2009. When we last checked in on this in October, the league was at +0.7 ALEX through Week 5, so the season started low and finished low.
Let's look at defenses first, followed by individual quarterbacks for the 2018 regular season.
As always, the following numbers are subject to change after we clean the game charting more in the offseason.
Defenses: 2018 ALEX Rankings Through Week 17
The following numbers are for third down only. Since this is defense, the teams are ranked from lowest to highest for the various stats. Lower DVOA is better. A lower ALEX means the defense is keeping the play short of the sticks. Short% is the percentage of third-down attempts that were thrown short of the sticks (negative ALEX). ALEX and Short% generally have strong correlation (typically near minus-0.85), but Short% is a good way to account for outliers.
|2018 Defenses: Third-Down ALEX Splits|
|Rk||Team||ALEX||2017 ALEX||2017 Rk||CONV%||Rk||DVOA||Rk||Avg. Need||Rk||Short%||Rk||Passes|
We are used to seeing a lot of year-to-year variation on defense for various statistics, but the correlation in ALEX was just 0.13 from 2017 to 2018. While three of the top five teams from 2017 were in the top five again this year, the Buccaneers (32nd to first) and Chiefs (30th to third) made huge leaps despite starting 2018 with the same defensive coordinator. Tampa Bay fired Mike Smith during the season, which led to some improvement in pass defense. In the end, the Buccaneers finished with the 32nd overall defense in DVOA. Generally speaking, defenses in 2018 didn't share a lot in common with what they did in 2017. The Bears and Bills were the top two defenses in DVOA after finishing 14th and 15th a year ago. The Texans, Cowboys, and Colts all finished in the top 10 after finishing in the bottom 10 a year ago. There are likely to be a lot of stats we'll look at for 2018 that don't match up well with the 2017 results.
In Tampa Bay's case, in the rare event the Buccaneers forced a team into third down, they had the highest average of need yards (9.1) to convert. That helped the Buccaneers lead the league in Short% (51.3 percent), though the conversion rate allowed was still pretty average (17th) and the DVOA (27th) even worse. Todd Bowles is reportedly coming to Tampa Bay as Bruce Arians' defensive coordinator. Bowles' Jets defense this year had the fifth-best conversion rate allowed and seventh-best DVOA allowed on third-down throws. The Vikings, Browns, Bears, and Ravens were the only other defenses to rank in the top seven in both stats, but none of them forced as high of an ALEX (+2.6) as the Jets (28th).
Oddly enough, three of the four defenses with the highest ALEX averages are linked to Bill Belichick: his Patriots, Mike Vrabel's Titans, and Matt Patricia's Lions. The Patriots finished 29th in ALEX for the second year in a row. The Lions had the worst of it though. Despite a +4.0 ALEX, quarterbacks converted 45.6 percent of the time (29th) and Detroit was 30th in DVOA.
Vic Fangio's Bears were middle-of-the-road in ALEX (14th), but had the best DVOA in a turnover-heavy season. In fact, the Bears had more interceptions this season (27) than in the last three seasons combined (24). Chicago's 10 interceptions on third down only trailed Denver (11). The beneficial picks can help explain why Denver had the largest disconnect between its ranking in DVOA (third) and conversion rate allowed (21st).
In terms of pure stops, Minnesota may have been as good as any defense in 2018 against third-down throws. The Vikings finished first in conversion rate allowed and second in DVOA. They were first in both categories in 2017. Atlanta had arguably the worst of things as Dan Quinn's injury-ravaged unit finished 31st in conversion rate allowed and 32nd in DVOA.
While Bob Sutton's Kansas City defense leaves a lot to be desired this season, at least the Chiefs rank 14th in conversion rate allowed. They'll get a great test against Andrew Luck on Saturday. The worst defense by these numbers still in the playoffs is New Orleans, which ranked 25th in conversion rate allowed and 29th in DVOA. Meanwhile, the Eagles ranked second in ALEX after finishing first in 2017, a Super Bowl year for Jim Schwartz's defense. The Eagles have a banged-up secondary this year, but still ranked fourth in conversion rate allowed. If they can keep the play in front of them against Drew Brees this week, that could help in pulling off a huge upset.
Quarterbacks: 2018 ALEX Rankings Through Week 17
The following table looks at all 36 quarterbacks this season with at least 50 third-down passes through Week 17, ranked from highest to lowest ALEX.
Alex Smith's conservative nature was the inspiration behind looking at ALEX in the first place. In Football Outsiders Almanac 2017, I wrote that Patrick Mahomes was the Anti-Alex as his ALEX was +4.2 at Texas Tech in 2016. Sure enough, in his first year as a starter Mahomes led the NFL with +4.5 ALEX. He also had the highest conversion rate and DVOA in a magnificent season. As for Smith, before he broke his leg in Washington, he was 22nd in ALEX (+0.3) and had poor efficiency numbers. That's the second year in a row Smith finished 22nd in ALEX.
Aaron Rodgers finished second to Mahomes in ALEX (+4.0) and DVOA; he has ranked in the top six in ALEX in every season since 2008. However, unlike Mahomes, Rodgers was 21st in conversion rate. It's the biggest difference any quarterback has between the two stats other than San Francisco's Nick Mullens (12th in conversion rate, 33rd in DVOA), and that one can be explained by Mullens having the highest interception rate (7.4 percent) on third down. In Rodgers' case, he was basically hit-or-miss on third down as his completion rate (56.8 percent) was 25th in the league. Since he attacked beyond the sticks so often, his completions often produced first downs and big plays. His yards per attempt (9.6) on third down only trailed Mahomes (9.7).
While Mahomes and Rodgers played to their reputations with ALEX, so did Josh Allen (+3.3) and Jameis Winston (+3.0). They had the two lowest Short% rates, as both love attacking deep even if they are off target more often than they should be. Allen obviously struggled too much as a rookie in Buffalo, but Winston finished second in conversion rate and fourth in DVOA. This isn't a fluke by any means. In 2017, Winston was third in ALEX, third in conversion rate, and fifth in DVOA. He has a history of moving the chains at an elite level even on money downs, but it's those pesky interceptions and ill-advised sacks that get him in trouble. With Bruce Arians coming to Tampa Bay, Winston's vertical cravings should be satisfied, but it's hard to say if he'll get to eat any more W's since Arians' system that will still open him up to hits and mistakes.
Tom Brady (+2.8) was leading Mahomes in ALEX through Week 5, but his numbers dropped as the season wore on. More alarming is that Brady finished 28th in conversion rate and 24th in DVOA, his worst season for as far back as we have charting data (2006). Brady has previously never finished lower than 20th in conversion rate. Of course, he is 41 years old and hasn't looked like the only aging member of that New England offense this year. For what it's worth, Brady's opponent this week, 37-year-old Philip Rivers, also had his lowest season yet in conversion rate (20th). Finally, 36-year-old Ben Roethlisberger led the league in passing yards (5,129), but his +1.0 ALEX is the lowest of his career, and he only ranked 26th in DVOA.
Let's focus on some younger quarterbacks. Mitchell Trubisky had the lowest ALEX (-2.4) as a rookie and he was 28th through Week 5 (-0.6). But under Matt Nagy, Trubisky changed his style to finish seventh (+2.2) in both ALEX and conversion rate. He also went from having the most need yards (8.8 yards) on third down under John Fox to the fourth-lowest average (6.9 yards) this season, so that always helps. Baker Mayfield finished ninth in ALEX (+1.7) and had an encouraging rookie season for the Browns. Sam Darnold was getting "Checkdown Charlie" comments in the preseason, but he had some decent moments early and late in the year. He finished 14th in ALEX (+1.1), but he definitely needs to be more consistent at converting (31st in rate).
Next, let's give some credit to superior offensive design by the likes of Sean McVay (Rams), Frank Reich (Colts), and Sean Payton (Saints). Of the bottom 15 quarterbacks in ALEX (0.5 or lower), only Andrew Luck, Jared Goff, and Drew Brees ranked in the top 10 in conversion rate. DVOA wasn't as kind with none ranking in the top 10, but those offenses will look to keep moving the chains on third down this weekend.
Jason Garrett may have saved his job again, but not everything is peachy in Dallas. Dak Prescott is Dallas' franchise quarterback, but he has been trending downward statistically since his stellar rookie season in 2016. His ALEX (-1.1) was only higher than Jeff Driskel's, but at least his conversion rate was 22nd in the league. Still, the lack of aggressive play on third downs combined with the high sack total this year do a lot to explain why Dallas struggled to score at times and why Prescott's advanced metrics aren't so hot.
Cincinnati backup Jeff Driskel easily had the league's worst ALEX (-3.1) and Short% (58.5 percent), but he also faced the longest third-down situations (9.8 yards). He also barely had enough attempts to qualify, and he is not any team's serious answer at the position, so we'll forgive his Blaine Gabbert-esque numbers there. Andy Dalton was clearly more aggressive and successful for the team, which will finally go forth without Marvin Lewis as the head coach in 2019.
As for the other teams with multiple quarterbacks featured, Ryan Fitzpatrick, like Winston, was aggressive and effective (turnovers aside). Brock Osweiler actually had better numbers than Ryan Tannehill in Miami, though both feasted at times on huge YAC plays in Adam Gase's offense. Gase has been fired and the Dolphins may want to start fresh at quarterback too. It's interesting that Tannehill faced the second-longest third downs (9.3 yards) while Osweiler faced the shortest (6.6 yards), but neither is the long-term answer there. Neither was Gase, for that matter.
By far the most intriguing quarterback controversy in the NFL today is between Nick Foles and Carson Wentz in Philadelphia. Like last year, Wentz is hurt and Foles is leading a playoff run for the Eagles. Also like last year, Wentz had more aggressive regular-season numbers than Foles, but Foles really turned things on in the playoffs. Wentz also was way down this year from his standards in 2017, when he finished second in ALEX and conversion rate and first in DVOA. This year he was 27th in ALEX, 19th in conversion rate, and 25th in DVOA, which is more in line with his 2016 rookie season. As for Foles, his numbers include the great outlier of 2018. Against Tampa Bay, the Eagles were stuck in a third-and-41 after a series of penalties. Foles ended the drive with a little pass to the running back to bring out the punting unit after a 15-yard gain. The play was -45 ALEX, the worst of the season. The biggest plus-ALEX pass of the season was when Derek Carr threw a 48-yard bomb incomplete to Martavis Bryant on third-and-3 against the Seahawks (+45 ALEX).
For the third year in a row, we'll point out a very disappointing rookie in the NFC. This time it's Josh Rosen in Arizona. He didn't have much help this year, but Rosen was 34th in ALEX (-1.0), 32nd in conversion rate, and last in DVOA. With Kliff Kingsbury coming over from Texas Tech to the NFL, the duo will try to recreate the improvement we've seen from McVay-Goff and Nagy-Trubisky the last two years.
Finally, ALEX is presented in splits by distance: short (1 or 2 yards to go), medium (3 to 6 yards), and long (7-plus yards). The conversion rates are also shown with a ranking.
|2018 ALEX: Distance Splits on Third Down|
Short: You know the "Short" section is small-sample stuff when Allen was the only quarterback to convert 100 percent of his plays. Rosen had the worst conversion rate (25.0 percent) in short situations. Rodgers (+7.9) loves to throw deep on third-and-short, but six quarterbacks still had a higher ALEX than he did there. It's just that 14 quarterbacks had a higher conversion rate too.
Medium: Rodgers shined more on the medium passes, where he had the highest ALEX (+8.5) by more than 2.5 yards over the next closest quarterback (Derek Carr was +5.9). Rodgers had the seventh-best conversion rate on those throws, while Ryan Fitzpatrick was No. 1. Interestingly enough, Dalton (+1.5) and Driskel (+1.1) had the lowest ALEX on medium passes, but Dalton was second in conversion rate compared to 34th for Driskel. Allen was dead-last in conversion rate on these throws, connecting a third of the time.
Long: This might be my favorite stat from this 2018 study. On third-and-long throws, Mahomes (+4.4) was first in ALEX and fourth in conversion rate while Smith (-4.6) was dead last in ALEX and 21st in conversion rate for Washington. That speaks a lot to why the Chiefs traded up for Mahomes and made the quarterback change, and they have a historic offense to show for it already. This was a better split for Brady as he finished third in ALEX and sixth in conversion rate. Winston actually led all passers in conversion rate while Fitzpatrick was 34th for Tampa Bay. Tannehill was last in conversion rate. The state of Florida could use some quarterback shake-ups, but maybe not in Tampa Bay just yet.