2017 ALEX: Midseason Report
by Scott Kacsmar
Our first look at ALEX this season found that through Week 4, quarterbacks were throwing their average third-down pass just 0.7 yards beyond the first-down marker. That would be the lowest average since 2006. The average ALEX from 2006 to 2016 was +1.3. Through Week 10, not much has changed. The league-wide ALEX on third down is +0.8, so we're still on pace for the first season on record to drop under +1.0.
For those new to this stat, ALEX stands for Air Less EXpected, and 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 3-yard pass on third-and-12, then that would be -9 ALEX. Any pass thrown beyond the first-down marker would earn a positive ALEX figure. Most quarterbacks are wise enough to attack the sticks on fourth down, which is why ALEX is best applied on third downs to gauge the aggressiveness of a quarterback. As you will see, we have identified some quarterbacks with a strong year-to-year tendency for aggressive or ultra-conservative play. ALEX data is always subject to change once corrections are made in the NFL's play-by-play stats, as well as the removal of certain incompletions such as passes thrown away or batted at the line.
The following table looks at the 38 quarterbacks this season with at least 25 third-down passes through Week 10. I kept the minimum low so we can look at current starters such as Brock Osweiler, Tom Savage, and Brett Hundley. The quarterbacks are ranked from highest to lowest ALEX. Also included are their conversion rate, DVOA, and average need yards for a first down. Finally, Short% is the percentage of third-down attempts that were thrown short of the sticks (negative ALEX). ALEX and Short% generally have strong correlation (near minus-0.85), but Short% is a good way to account for outliers.
The pride of the 2015 quarterback draft class sits at the top of the table, but Jameis Winston is currently nursing a shoulder injury while Marcus Mariota just threw a career-high four interceptions in Pittsburgh last night. Needless to say, the junior year hasn't gone as planned for these two. Even though both have averaged fewer than 7.5 need yards on third down, their conversion rates and DVOA are subpar. Mariota's +3.5 ALEX would be the lowest average to lead the league in our database since 2006.
[ad placeholder 4]
At the bottom, we observe some statistical support for the idea that John Fox's Bears have tried to hide quarterbacks Mike Glennon (-1.5 ALEX) and rookie Mitchell Trubisky (-2.6 ALEX) this year. Trubisky's numbers are most interesting, because he has the largest difference in his rank in conversion rate (28th) and DVOA (third). There are a few noteworthy splits for Trubisky.
- On third down with 1 to 4 yards to go, Trubisky is 4-of-12 for 35 yards with an unimpressive -20 DYAR and -29.3% DVOA. Average success rate is 57.2 percent in this situation in 2017.
- On third down with 6 to 10 yards to go, Trubisky is 9-of-15 on conversions with the highest DVOA (198.4%) in the league. The only five quarterbacks ahead of him in DYAR all have 30 to 38 attempts compared to Trubisky's 15 passes.
- On third down with 13-plus yards to go, Trubisky has completed 8-of-9 passes, but for only one first down. His 33 DYAR ranks ninth, so his failed completions are still generating some positive value.
We'll see how he finishes the year, but right now, Trubisky's -2.6 ALEX would tie Jared Goff for the lowest by a rookie since 2006. Speaking of Goff, he has one of this year's largest splits between rank in ALEX (30th) and conversion rate or DVOA (fourth in both). Some of this can be traced to one play, which was the "ALEX Failure of the Year" for a defense. The Giants allowed Robert Woods to take a bubble screen on third-and-33 all the way to the house for a 52-yard touchdown in Week 9. That play was -34 ALEX, and there is nothing quite like it in our database going back to 2006. Without that play Goff's DVOA would still be 64.3% on 0.0 ALEX, so this still speaks to Sean McVay's success with designing an offense that is dominating with YAC and play-action passing right now.
Also near the bottom in ALEX, you'll find a very surprising veteran name. Drew Brees ranks 36th (-1.3 ALEX), but is still seventh in both conversion rate and DVOA. He's actually throwing the third-shortest passes (6.2 air yards) of any quarterback in all situations this year, but the Saints are making it work so far. Alvin Kamara has been a fine addition as a receiving back. Brees has ranked 22nd or lower in ALEX four times before with New Orleans, but this would be the first season he was ever below 0.0.
Carson Wentz and Tom Brady lead all quarterbacks in conversion rate and DVOA. Wentz was just 27th in ALEX last year (0.0), but has climbed to fifth this year (+2.5) with a far more aggressive approach that is paying dividends so far. Brady has ranked in the top eight just as often as he has in the low 20s in ALEX before, but finds himself a little above average at 15th right now (+1.3 ALEX).
Aaron Rodgers (+3.2 ALEX) ranks third, and has ranked in the top six in every season since 2008. Hopefully we'll get to add more data to him later this year, but that probably depends on how Green Bay performs without him. So far, Brett Hundley ranks seventh in ALEX (+2.3), but hasn't had nearly the same type of success as Rodgers. That's not surprising in any way, and Rodgers' overall efficiency also helped set him up with the shortest distance to go on third downs (6.5 need yards). Similar things are happening in Houston, where Tom Savage has been very aggressive (+3.1 ALEX), but has the second-worst conversion rate (31.0 percent) in the league. Deshaun Watson will be missed greatly the rest of the year.
Speaking of rookies, DeShone Kizer is in that 2016 Jared Goff territory of ranking dead last in many categories -- but ALEX is not one of them. When it comes to third downs, he has the worst conversion rate and worst DVOA, but also has the highest average need yards (9.0).
[ad placeholder 3]
I usually mention Alex Smith earlier than this, given that we named the stat fter him. But he's not nearly as interesting now as he was after Week 4, when he was dead last again at -4.7 ALEX. Sure enough, in a prime-time game in Houston in Week 5, Smith came out firing beyond the sticks all night with +6.2 ALEX. It was a great success too, with Smith converting 8-of-12 third downs that night. His ALEX has been at least +1.9 in every game since, though his success rate has fallen to 34.2 percent and the Chiefs have lost three of the last four games.
Then we have the case of Tyrod Taylor, whom Buffalo benched this week. Taylor actually ranked 10th in conversion rate (42.4 percent) even though he had the fourth-highest need yards (8.7). His ALEX dropped to -0.1 this season, ranked 29th. We've seen this continued change in Taylor's game. In his first season with Buffalo, his ALEX was +4.5, which ranked second in the NFL. It dropped to +2.3 in 2016, but still good enough for seventh in the league, and he was better at converting plays. Taylor's conversion rate was actually a career best this year, but alas he is benched after the Bills allowed 81 points in two losses. A big problem this year was the massive changes at wide receiver. Taylor was targeting running back LeSean McCoy more than anyone on third down, and those 16 targets produced a -10.9 ALEX. In 2016, Taylor's go-to targets on third down were deep threats Marquise Goodwin (+6.2 ALEX) and Sammy Watkins (+7.1 ALEX). Both of those receivers, as well as Robert Woods (+1.4 ALEX), are in the NFC West this year.
Finally, the average need yards are certainly a factor in how successful and perhaps how aggressive quarterbacks are going to be on third down. Like in past years, I have provided a "heat map" to show how the quarterbacks stack up in the different third-down splits: short (1 or 2 yards to go), medium (3 to 6 yards) and long (7-plus yards). The colors indicate where a player is well above average (darker green) versus below average (darker red) based on standard deviations. Those conversion rates are also shown with a ranking.
This highlights the way Mariota and Winston have struggled at each depth level, while Brees and the Chicago quarterbacks have had some unusual success in medium and long situations despite their low ALEX numbers. We also see some alarming splits from Savage, Kizer, and Hundley. Andy Dalton has managed to do well in third-and-long, but poor in shorter situations this year. Don't get worried about his 0.0 conversion rate on third-and-short -- that was just 0-for-3 passing.
Brady ranks first on third-and-long, while Wentz's success has been on short and medium attempts. Even Savage and Glennon are 100.0 percent on a few short throws, but it's the medium range where Wentz has excelled with 19 conversions in 26 attempts.