Writers of Pro Football Prospectus 2008

Most Recent FO Features

BortlesBla14.jpg

» The Week In Quotes: December 9, 2016

This week: a bad coach gets paid, then insulted; a bad quarterback gets optimistic; another bad quarterbcak gets a cunning plan; a bad play gets Matt Ryan irked; a bad play gets burned; and Jets and Raiders fans get drunk.

04 Jan 2016

ALEX: Season Review

by Scott Kacsmar

Today we will look at more of a season review of ALEX instead of focusing on Week 17, when Jameis Winston followed up a minus-39 ALEX pass (the lowest of the 2015 season) with a minus-30 ALEX pass to end Tampa Bay's 38-10 loss in Carolina. That will definitely affect the numbers for Carolina's defense, which we will look at for the first time this season. Any subsequent ALEX studies will come in the offseason, and do remember these numbers are subject to change after we go through the game charting to remove certain plays.

Ben Roethlisberger finished the season with a plus-6.83 ALEX on third down, the second-highest average since 2006 behind only Michael Vick's plus-6.97 for the 2006 Falcons. Those are the only two seasons (minimum 50 passes) above 5.0. This season, there were four completions with an ALEX of at least 40, and Roethlisberger had three of them.

We do have a new low benchmark with Alex Smith's minus-3.4 ALEX beating out 2009 Trent Edwards (minus-2.8). Smith and Blaine Gabbert now account for five of the 10 lowest ALEX seasons since 2006. Nine of those 10 seasons have ranked 26th or worse in conversion rate. Edwards lost another "record" to Gabbert. In 2009, Edwards threw short of the sticks on 62.3 percent of his third-down throws. Gabbert just finished the season at 65.0 percent.

Defenses: 2015 ALEX Rankings Through Week 17

This is the first time we are looking at defensive numbers. It would not be hard to get the same data for 2006-2014, but to this point I have not gone through that to see if there are any defenses that do a good job of consistently forcing quarterbacks to check down on third down. Dick LeBeau's strategy of "tackle the catch" would certainly resonate with how the Steelers approached things, and defensive coordinator Keith Butler is really no different in Pittsburgh, but here are what the numbers look like for 2015. 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.

Rk Team ALEX CONV% Rk DVOA Rk Avg. Need Rk Short% Rk Passes
1 CHI -0.6 43.8% 27 53.1% 29 7.8 18 48.2% 3 137
2 STL -0.6 34.2% 3 0.5% 6 7.9 12 52.3% 1 155
3 SEA 0.2 36.7% 10 -4.5% 2 7.9 10 51.1% 2 139
4 CIN 0.4 43.4% 25 29.9% 20 8.0 8 45.1% 9 173
5 TEN 0.4 45.9% 30 50.7% 28 7.8 15 42.5% 15 146
6 ATL 0.4 40.9% 18 33.4% 21 7.9 11 42.3% 16 137
7 JAC 0.4 47.9% 31 94.2% 32 8.2 3 46.1% 5 167
8 CAR 0.6 39.2% 15 12.3% 12 8.0 9 38.1% 25 176
9 WAS 0.6 38.5% 14 25.6% 19 8.3 2 39.9% 22 148
10 MIA 0.7 44.6% 28 62.7% 31 7.7 21 45.3% 8 148
11 BUF 0.8 37.3% 12 8.8% 9 7.7 22 39.9% 21 153
12 MIN 0.9 37.0% 11 25.5% 18 7.9 13 43.8% 11 146
13 TB 0.9 50.7% 32 46.6% 25 7.2 29 39.4% 24 142
14 NO 1.0 43.0% 24 56.1% 30 6.8 30 40.4% 19 114
15 DEN 1.0 36.5% 8 -1.7% 4 8.2 5 44.1% 10 170
16 SD 1.1 36.7% 9 21.8% 15 7.8 14 45.3% 7 139
Rk Team ALEX CONV% Rk DVOA Rk Avg. Need Rk Short% Rk Passes
17 DAL 1.1 36.1% 6 13.1% 13 7.8 16 43.8% 12 144
18 CLE 1.2 36.3% 7 10.0% 11 7.3 28 42.5% 14 146
19 OAK 1.2 40.1% 17 9.2% 10 7.4 27 41.4% 18 157
20 GB 1.3 35.2% 5 -10.2% 1 8.2 4 45.7% 6 162
21 ARI 1.3 37.7% 13 3.1% 8 7.5 26 37.7% 26 159
22 PIT 1.3 41.2% 20 0.2% 5 7.6 24 40.0% 20 170
23 BAL 1.4 42.9% 23 50.5% 27 7.8 19 46.8% 4 154
24 DET 1.5 43.5% 26 49.2% 26 7.6 23 36.2% 29 138
25 NYJ 1.6 32.4% 2 -3.2% 3 7.8 20 43.4% 13 173
26 IND 1.7 42.2% 21 22.2% 16 7.8 17 37.0% 28 173
27 HOU 1.9 29.2% 1 2.9% 7 8.3 1 39.6% 23 154
28 PHI 2.2 42.8% 22 37.5% 22 7.5 25 41.6% 17 173
29 NYG 2.7 44.9% 29 37.9% 23 6.7 32 35.4% 31 178
30 KC 2.9 34.6% 4 14.8% 14 8.1 6 34.6% 32 159
31 NE 3.1 39.4% 16 43.4% 24 8.1 7 37.1% 27 170
32 SF 4.6 41.0% 19 25.0% 17 6.8 31 35.8% 30 134

Well, Blaine Gabbert may not have thrown many passes past the sticks, but his opponents sure did. With an average ALEX of plus-4.6, the 49ers were more than three standard deviations above the average of plus-1.3. That is pretty interesting, considering that the now-fired Jim Tomsula had been in charge of a revamped defensive roster that opponents attacked all year. The 49ers did have the second-shortest distance to defend, but the correlation here between ALEX and Avg. Need was minus-0.35. My first thought was the 49ers had three games against Roethlisberger and Carson Palmer, so that had to move the needle. As it turns out, Roethlisberger's ALEX in that stellar Week 2 performance was plus-19.2, to go along with a DVOA of 271.7%. Palmer's ALEX was 5.4 over two games, but he was one of seven quarterbacks to have at least 5.0 ALEX against the 49ers this season. The lowest game was Teddy Bridgewater (minus-1.0) in that Week 1 San Francisco win that feels like eons ago, but Bridgewater also has a growing reputation for being a low-ALEX quarterback.

In the future we could probably work in some type of opponent adjustment for quarterbacks and the defenses they play, but most defenses fall in a pretty small range. Only the Rams and Bears had negative ALEX, and both were minus-0.6. There was no Alex Smith equivalent on defense this season.

There was very little correlation (minus-0.09) between ALEX and conversion rate, which is quite different from this year's quarterback results, where the correlation was 0.46. Of course we are used to results that show offense, especially something the quarterback controls, being more consistent than defense. The Rams, led by Aaron Donald up front, had negative ALEX and the third-best conversion rate allowed. The Bears, with a cast of nobodies, had negative ALEX and ranked 27th in conversion rate.

Once the numbers are finalized we will look back at those old seasons to learn more about the defense's role in ALEX.

Quarterbacks: 2015 ALEX Rankings Through Week 17

The following table shows where each qualified quarterback (minimum 50 passes) ranks in ALEX on third down only. There are also rankings for DVOA, average need yards (ranked from highest to lowest) and conversion rate.

Rk Quarterback Team ALEX CONV% Rk DVOA Rk Passes Avg. Need Rk Short% Rk
1 Ben Roethlisberger PIT 6.8 47.7% 5 59.5% 8 109 6.6 37 27.5% 1
2 Carson Palmer ARI 4.2 53.2% 1 96.7% 2 141 7.1 32 27.7% 2
3 Tyrod Taylor BUF 4.1 40.0% 20 43.4% 15 120 8.1 11 38.3% 15
4 Ryan Fitzpatrick NYJ 3.1 41.0% 18 36.2% 18 156 7.5 25 35.9% 9
5 Aaron Rodgers GB 3.1 34.1% 29 24.7% 21 129 7.6 22 29.5% 3
6 Cam Newton CAR 3.0 39.4% 23 48.0% 10 127 7.9 15 36.2% 10
7 Jay Cutler CHI 2.9 45.4% 8 64.3% 6 141 7.1 34 37.6% 14
8 Blake Bortles JAC 2.6 39.0% 24 30.3% 20 146 7.7 21 32.2% 5
9 Andrew Luck IND 2.6 42.0% 14 42.4% 16 69 7.6 23 36.2% 11
10 Brian Hoyer HOU 2.5 42.9% 12 44.5% 12 105 7.7 19 35.2% 6
11 Andy Dalton CIN 2.1 44.7% 10 68.9% 5 94 7.8 16 37.2% 13
12 Brock Osweiler DEN 1.9 41.8% 17 17.1% 23 79 7.3 28 35.4% 8
13 Russell Wilson SEA 1.8 50.4% 3 90.6% 3 131 7.4 27 43.5% 24
14 Josh McCown CLE 1.8 51.3% 2 103.8% 1 78 7.2 30 41.0% 21
15 Joe Flacco BAL 1.8 34.0% 30 -14.5% 34 103 7.0 36 46.6% 28
16 Johnny Manziel CLE 1.6 41.9% 15 72.4% 4 62 8.3 5 38.7% 16
17 Peyton Manning DEN 1.5 32.3% 33 -29.9% 36 99 8.2 7 39.4% 17
18 Ryan Tannehill MIA 1.5 30.7% 35 -1.7% 33 163 8.6 2 46.0% 25
Rk Quarterback Team ALEX CONV% Rk DVOA Rk Passes Avg. Need Rk Short% Rk
19 Jameis Winston TB 1.2 41.9% 16 5.4% 30 148 8.4 4 31.1% 4
20 Tom Brady NE 1.2 43.8% 11 60.2% 7 144 7.4 26 47.2% 30
21 Drew Brees NO 1.1 46.4% 6 55.4% 9 168 8.0 12 40.5% 19
22 Marcus Mariota TEN 1.1 40.2% 19 12.5% 27 102 8.1 8 35.3% 7
23 Kirk Cousins WAS 0.8 45.3% 9 48.0% 11 159 7.5 24 47.2% 29
24 Colin Kaepernick SF 0.7 33.8% 31 11.6% 29 74 7.7 20 43.2% 23
25 Derek Carr OAK 0.7 38.7% 25 37.3% 17 163 8.1 10 39.9% 18
26 Matt Hasselbeck IND 0.6 40.0% 21 -15.5% 35 70 7.0 35 48.6% 31
27 Matt Ryan ATL 0.6 49.4% 4 44.2% 13 168 7.2 31 36.3% 12
28 Eli Manning NYG 0.6 38.7% 26 16.7% 25 163 7.1 33 42.9% 22
29 Philip Rivers SD 0.5 42.1% 13 35.3% 19 171 8.1 9 50.3% 33
30 Nick Foles STL 0.3 24.4% 36 -31.1% 37 119 8.3 6 49.6% 32
31 Matt Cassel DAL 0.2 31.7% 34 12.0% 28 60 7.8 17 51.7% 35
32 Ryan Mallett BAL 0.2 46.1% 7 16.2% 26 76 7.2 29 40.8% 20
33 Sam Bradford PHI -0.5 33.1% 32 16.7% 24 145 8.0 14 46.2% 27
34 Matthew Stafford DET -0.6 38.6% 27 3.5% 31 145 8.0 13 46.2% 26
35 Teddy Bridgewater MIN -0.8 39.4% 22 44.2% 14 132 7.7 18 51.5% 34
36 Blaine Gabbert SF -2.7 23.8% 37 -1.6% 32 80 8.9 1 65.0% 37
37 Alex Smith KC -3.4 35.2% 28 18.9% 22 122 8.4 3 59.8% 36

Next, ALEX is presented in splits by distance: short (1 to 3 yards), medium (4 to 7 yards) and long (8-plus yards). The colors indicate where a player is well above average (darker green) versus below average (darker red). Those conversion rates are also shown with a ranking.

Note: these numbers are subject to change at season's end. The data on 2006-2014 is the same as what we use for stats like receiving plus-minus and YAC+, which excludes passes that are thrown away, batted at the line or when the quarterback was hit in motion. The 2015 data currently includes all passes, but game charting will filter out those passes that were not truly aimed or intentional.

For those new to this metric, it is called 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 pass five yards behind the line of scrimmage on third-and-15, that would be minus-20 ALEX. The best application of ALEX is to look at third and fourth down when it's really crucial to get 100 percent of the need yards to extend the drive.

Posted by: Scott Kacsmar on 04 Jan 2016

5 comments, Last at 10 Feb 2016, 6:20pm by Scott Kacsmar

Comments

1
by ammek :: Tue, 01/05/2016 - 9:16am

It's remarkable that Rodgers ranks dead last in conversion percentage on third-and-short. And that excludes all the sacks he took. I think if we filtered out passes to running backs – on which the Packers were uncharacteristically successful this year – Rodgers would look even worse. Although so would Alex '–ALEX' Smith.

2
by cstoos :: Tue, 01/05/2016 - 9:29am

One of the first things I noticed on the defensive side of things was the lack of correlation with ALEX and Conversion % (which you later stated).

Interestingly to me, some of the best defenses in the league (HOU, NYJ, KC) were all near the bottom of the ALEX rankings. So, does defensive success on third down depend more on taking away the short pass?

Generally, the deeper the pass, the lower the completion percentage, so that works in the defenses favor. Additionally, the deeper the pass, the longer the QB has to hold on to the ball before the throw which likely generates more sacks. This is all conjecture, but I am willing to bet the numbers support it.

Additionally, I guess ALEX doesn't really punish defenses for giving up HUGE plays. A 9 yard throw for a 9 yard gain on 3rd and 8 counts the same as a 9 yard throw for a 90 yard TD on 3rd and 8, correct?

3
by Scott Kacsmar :: Tue, 01/05/2016 - 9:38am

My first thought is the passes that are intentionally thrown away could be increasing the ALEX for good defenses. Sometimes the QB really throws that thing far out of bounds, and I'm not sure what kind of air yards they're attaching to those plays. We'll filter those out soon, so that could bring the ALEX down. Obviously forcing a throwaway is a win for the defense, but with stats like ALEX, +/- and YAC+ we're really trying to focus on the results of aimed passes.

"Additionally, I guess ALEX doesn't really punish defenses for giving up HUGE plays. A 9 yard throw for a 9 yard gain on 3rd and 8 counts the same as a 9 yard throw for a 90 yard TD on 3rd and 8, correct?"

Right, DVOA dishes out the punishment there. YAC+ rewards the receiver. ALEX acknowledges the depth of the throw as a plus-pass.

4
by mleland :: Fri, 01/22/2016 - 5:25pm

Hi Scott, do you have this kind of data for previous seasons?

5
by Scott Kacsmar :: Wed, 02/10/2016 - 6:20pm

Yes, expect some offseason posts on that.