Metrics will Improve, But Will Arizona's Win %

  • Oops!
    Something went wrong.
    Please try again later.
·14 min read
In this article:
  • Oops!
    Something went wrong.
    Please try again later.


This content is not available due to your privacy preferences.
Update your settings here to see it.


It is hard to properly value the efficiency that a good running quarterback brings to an NFL offense. For example, consider the following:

Derrick Henry is a beast of a running back and had a career year last year, averaging over 125 rushing yards per game with over 2,000 total yards on the season… but…

...in the first three quarters, Kyler Murray’s rushing produced nearly four times more expected points than Derrick Henry’s rushing, on one-third of the carries. Kyler Murray averaged over ten times more expected points per rush than Henry.

Murray’s 0.41 EPA/att on 98 attempts dwarfed Henry’s 0.04 EPA/att over his 293 attempts in the first three quarters of games.

In last year’s book, I urged the Cardinals to use Murray’s legs even more in 2020.

They obliged, and Murray’s rushing EPA/att was tops in the NFL last season. It was a huge reason for Arizona’s early season success and their 5-2 record out of the gates.

It is vital that opposing defenses try harder to minimize Murray’s rushing upside (which is easier said than done), but it’s also vital that we account for it when analyzing the Cardinals’ rushing output.

Because if we leave it in the calculus, the Cardinals 2020 rushing offense ranked third in EPA/att.

But if we remove QB rushing, the Cardinals ranked below-average, 18th in the NFL, with just 4.1 YPC, 48% success, and -0.04 EPA/att from non-QB runs.

After Kenyan Drake delivered 5.2 YPC and 0.12 EPA/att in 2019, he regressed down to just 4.0 YPC and -0.05 EPA/att in 2020. Chase Edmonds wasn’t significantly better (-0.02 EPA/att). This rushing offense wouldn’t have contributed anything save for Murray’s ability on the ground.

It didn’t matter the run type, Murray was tremendous with them all.

On designed quarterback runs, including sneaks and draws, he gained +0.24 EPA/att with 6.1 YPC on 67 attempts. On scrambles, due to either defensive pressure, coverage, or open run lanes, Murray averaged +0.64 EPA/att with 8.2 YPC on 52 attempts.

The only runs which were unsuccessful were designed runs on third down (-0.53 EPA/att).

When you have a quarterback delivering such efficiency on the ground, and a set of running backs incapable of producing anything close to such efficiency, it becomes highly problematic when your quarterback gets hurt and cannot run effectively.

Murray dealt with a shoulder injury over the second half of the season, an injury he sustained when he landed on the shoulder on the first drive in Week 11 at Seattle. He battled through that game and kept playing, but after the season he said he initially hurt it against the Dolphins in Week 9 and then re-injured it in that Seattle game. Regardless of when the injury first occurred, what was clear was Kingsbury and Murray decided he should run the ball less often in that Week 11 game in Seattle.

Murray went from rushing attempts of 10, 14, 11, and 11 during the prior four games (three of four were wins) to five runs in each of the next three games, starting in Week 11 (Week 11 in Seattle, Week 12 in New England, and Week 13 vs the LA Rams).

Arizona lost all three games.

A run game which recorded positive EPA every single week since Week 1 (nine weeks of EPA above zero) with six wins in nine games, suddenly recorded negative EPA on the ground week after week after week.

It wasn’t just the first three weeks when Murray was initially injured. Arizona recorded below zero EPA for six of their final seven games, and lost five of them.

Although Murray later said he had to “play through the shoulder and took hits here and there and kept playing – it was fine,” it was clear he wasn’t able to run as often or as effectively as he did earlier in the season.

Pre-injury: 9.0 rushes/game, 0.50 EPA/att, 7.6 YPC, 62% success
Post-injury: 6.1 rushes/game, 0.20 EPA/att, 5.1 YPC, 60% success

The inability to run as often or as effectively had a substantial impact on Arizona’s overall ability to win games, because, despite acquiring DeAndre Hopkins, and being an “Air Raid” offense, Arizona ran the ball more than most teams in neutral situations. 2020 showed an increase in run rate over 2019, Kingsbury’s first season in the desert.

But poor rushing alone wasn’t what sent the Cardinals from a 6-3 start to a 2-5 finish. The shoulder injury predictably impacted Arizona’s passing attack as well.

In the first nine games, Murray had 14 completions over 20 yards. In the last seven games? Only six.

Kyler’s splits pre- and post-injury were clear:

Pre-injury: 0.05 EPA/att, 8.1 aDOT, 7.6 YPA, 3.7% sack rate
Post-injury: -0.06 EPA/att, 7.2 aDOT, 6.6 YPA, 5.8% sack rate

Pre-injury, his completions were thrown to an average depth 2.7 yards short of the sticks. This was with an average of 9.1 yards to go on each completion. Post-injury, his completions were thrown with an average of only 8.8 yards to go, a shorter distance. If he averaged the same air yards, he would complete his passes just 2.4 yards short of the sticks, forcing less onto the receiver’s plate post-catch. But with the injury, these completions were significantly shorter: thrown 4.0 yards short of the sticks. In summary, his average depth of completion dropped nearly two yards, from 6.4 pre-injury to 4.8 post-injury.

This certainly didn’t help the Cardinals’ ability to avoid third downs or at least third-and-longs.

Pre-injury, the Cardinals gained 233 first downs and went to third down only 112 times, the second best ratio in the NFL. Post-injury, their rate of third down avoidance dropped to 17th.

Pre-injury, the Cardinals averaged 6.8 yards-to-go on third down, 13th best in the NFL. Their conversion rate ranked an identical 13th best in the NFL.

Post-injury, the Cardinals averaged 7.1 yards-to-go on third down, eighth worst. And their third down conversion rate was fourth worst in the NFL.

Worse early down production led to far more third downs, longer third downs, far worse performance on third down… and predictably, more losses.

Another thing that happened was Murray took more sacks after getting injured. Over the first 10 weeks of the season, he was pressured on just 24% of his dropbacks, fourth lowest in the NFL. 14% of his pressures turned into sacks, 11th lowest in the NFL. When he wasn’t sacked on these pressures, he averaged 7.0 YPA, 10th best in the NFL.

But after his shoulder injury, Murray was pressured on 30.4% of his dropbacks (increase of over 6%) and 16.3% of his pressures turned into sacks. More pressures and more sacks. When the pressure didn’t result in a sack and he released a pass, he averaged only 5.3 YPA, a significant decline from 7.0 YPA pre-injury.

In total, it was a season that saw a huge reduction in sacks for Murray, as the Cardinals shifted from allowing 50 in 2019 to just 29 in 2020. As a team, they shifted from being -10 in sack margin to +19, a 29-sack swing. It was the second best swing of any team last year.

In last year’s book, I predicted the Cardinals would face the 10th most difficult schedule of run defense and that turned out to be the case. But this was not a substantial increase over 2019 when they aced the 11th most difficult schedule.

Their poor 2020 rushing primarily related to a regression of RB Kenyan Drake (particularly in 11 personnel where he slipped from 53% success and 4.3 YPC in 2019 to 44% success and 3.8 YPC in 2020) and the Kyler Murray injury.

Kingsbury’s offense has been great at avoiding runs into loaded boxes. Only 10% of their runs went into 8+ man boxes (fewest in the NFL) and only 33% went into 7-man boxes (sixth fewest). These were very similar rates to 2019, and are unlikely to see any improvement in 2021. A whopping 51% of Arizona’s runs went into light boxes of six or fewer defenders, the second highest rate in the NFL. That must continue in 2021. Kingsbury certainly hopes the addition of RB James Conner will improve rushing efficiency, but I’m skeptical.

Last year with the Steelers, in runs against standard or light boxes in the first three quarters, Connor’s 3.8 YPC was last of the Steelers’ three primary ball carriers, as was his -0.10 EPA/att. And as a reminder, he shared the backfield with Benny Snell and Anthony McFarland Jr.

It’s always interesting to study college coaches shifting to the NFL and their philosophy. We know in college, Kingsbury utilized substantial rates of 10 personnel, featuring 4-WRs, zero TEs, and 1-RB. As discussed in last year’s book, while Kingsbury used 10 personnel on over 60% of offensive plays his first month on the job (and fewer than 3-WRs less than 9% of offensive play calls), Kingsbury greatly reduced his dependence on 10 personnel immediately thereafter, after noticing lack of success and mounting WR injuries.

I praised him for this personnel flexibility and not being wed to using 10 personnel if the team wasn’t built to accommodate it… while wondering what Kingsbury’s strategy would be in 2020 after acquiring DeAndre Hopkins and being able to trot out Christian Kirk and Larry Fitzgerald at a minimum for 11 personnel, while rotating on another receiver to get to 10 personnel.

Kingsbury didn’t even attempt to utilize heavy rates of 4-WR sets at all to start the 2020 season. Instead, he opted to go with heavy rates of 2-TE sets in 12 personnel. Over the first three weeks of the 2020 season, Kingsbury’s “Air Raid” offense used the second-most 12 personnel of any team in the NFL.

They used 12 personnel on 30% of offensive snaps (the league average was 18%), reducing their 11 personnel down to 52% (average was 61%). Kingsbury used 10 personnel on only 13% of offensive plays, an incredible decline from his first three weeks in 2019 where they were at 61% 10 personnel.

For an “Air Raid” team built on spacing and most typically using no tight ends in 10 personnel, Kingsbury’s Cardinals finished the 2020 season ranked second in usage rate of 2-TE sets, using them on 31% of offensive plays.

I don’t view A.J. Green as a tremendous offseason addition (posted career lows in yards per reception, yards per target, touchdowns, receiving first downs, and catch rate last year after returning from missing the entire 2019 season) but I do think rookie WR Rondale Moore will provide more juice for this passing attack.

A change I fully believe in betting on is for Arizona to return to more 3- and 4-WR sets in 2021 and with it, a reduction in usage of 12 personnel.

Another thing the Cardinals should absolutely look to incorporate more in 2021 is pre-snap motion. Arizona’s efficiency improved from +0.01 EPA/att without pre-snap motion to +0.05 EPA/att when using it, and their YPA increased by 0.6. These improvements ranked 10th most and 13th most in the NFL last year. In addition, when Arizona ran the football after using pre-snap motion, their rushing efficiency improved more than any other team in the NFL.

And yet the 2020 Cardinals ranked dead last in pre-snap motion usage rate overall, before pass plays, and before run plays. Similar to Kingsbury adapting his Air Raid philosophy to use more tight ends when it showed it increased efficiency, he should do similar and use more pre-snap motion since it produces such dividends for the Cardinals.

On the other side of the ball last year, Arizona’s defense had the good fortune of facing the 20th easiest schedule of offenses, including the fifth worst collective group of passing offenses based on yards gained per pass attempt. Arizona played offenses from the terrible NFC East and AFC East (everyone but Buffalo was bad), plus the rebuilding Lions and Panthers.

Fortunately this season, they’ll still get the Lions and Panthers, but instead of playing the NFC East, the Cardinals defense must slow down the NFC North along with the Cowboys (featuring a healthy Dak Prescott), and the Browns. They also will face the Bears but not until Week 13, and it’s a game for Chicago which is three weeks after their bye and right after a mini-bye. Justin Fields is almost guaranteed to play against the Cardinals by Week 13, and he’s a far scarier prospect than Andy Dalton or Nick Foles.

But one interesting wrinkle has elevated the Cardinals slightly is they are scheduled to face both the Packers and Texans this season, both of which had Pro Bowl-level quarterbacks at minimum in Aaron Rodgers and Deshaun Watson. Both of which may not be facing the Cardinals this season, boosting the Arizona defensive outlook.

If both Rodgers and Watson don’t play against the Cardinals, these numbers aren’t as riveting, but as of now, the schedule shifts from playing the fifth easiest pass offenses in 2020 to the toughest pass offenses in 2021, the most difficult increase in schedule toughness of any defense this year.

For years, the Cardinals defense had been great at home. From 2012-2017, when playing at home the Cardinals held opposing offenses to nearly three fewer points than oddsmakers projected, by far the single best mark in the NFL for any team. When playing at home, the Cardinals held opposing offenses under the projected team total in 30 of 49 games (61%), a huge money making opportunity. But since 2018, the Cardinals have allowed opposing offenses to exceed their projected team total in 17 of 24 games (71%).

In Kliff Kingsbury’s two years, only three in 16 home games (19%) saw the Cardinals defense hold their opponents under their projected team total. That’s the second worst mark in the NFL behind the Lions.

To be fair, road offenses have been more productive than average over the last couple years, and we know no crowd noise impacted road offenses quite favorably in 2020. But the Cardinals defense has been much worse than average, after being the best defense in the NFL in this metric for over half a decade.

With the defense adding DE J.J. Watt and CB Malcolm Butler in free agency (replacing Patrick Peterson), drafting LB Zaven Collins 16th overall, and drafting defensive players with five of their first six picks, the front office knows that this defense needs to improve.

Statistically, the defense was much better last year than in 2019, improving from 27th against the pass in 2019 to 10th in 2020 and improving from 23rd in total defense in 2019 to 10th in 2020.

Kyler Murray has seen a remarkable start to his career. He is the only quarterback in NFL history to start his career with consecutive seasons of at least 3,500 passing yards, 20 passing TDs, and 60% completions. Murray is also the only quarterback in NFL history with consecutive seasons (regardless of years in the NFL) of at least 3,500 passing yards, 20 passing TDs, and 500 rushing yards.

The influx of talent, primarily with Rodney Hudson, J.J. Watt, the drafting of Rondale Moore, and a clear focus on improving defensive efficiency should make the Cardinals a better team in 2021.

Arizona plays the ninth toughest overall schedule of 2021 opponents, and an absolutely brutal increase in opposing offenses. With Kyler’s legs and more talent at the receiver position, the arrow will absolutely point up for this offense, particularly at the start of the season with three AFC South opponents in the first seven weeks.

Arizona is becoming a trendy team to back this offseason. At this time last year, Arizona was an underdog of 6+ points in five of eight road games. This year, in nine road games, the Cardinals aren’t underdogs of 6+ points in any of them. For Arizona to have a shot at taking a significant step forward in 2021, they must improve defensively, keep Murray healthy, and figure out how to win divisional games. Under Kingsbury, Arizona has won just three of 12 divisional games (one of which was in overtime). The 49ers and the Rams will be better at key positions this season and Kingsbury is a collective 1-7 in eight games against them over the last two years, losing by nearly double digits on average.

Stay tuned over the next eight weeks as we preview all 32 teams with daily articles and videos right here at the preview hub. For complete team chapters featuring dozens of visualizations and 462 pages, pick up a copy of Warren Sharp’s new ‘2021 Football Preview’ book.

PointsBet is our Official Sports Betting Partner, and we may receive compensation if you place a bet on PointsBet for the first time after clicking our links.