Happy 2021 NFL season!
NFL franchises use contextualized data to create competitive advantages. In order to realize an edge, teams need to employ the right data in the right way at the right time. This means distilling, interpreting and applying only the most influential data in a framework that accounts for their personnel, their opponents and their evolving game situations. My goal is to be your analytics department. Each week this season, I'm giving you a peek into which numbers flag in my models as the most impactful ... or the most misunderstood.
This column will follow the format of: one trend to monitor, one news story through the lens of analytics and a couple of my favorite, or least favorite, projections.
As always, let me know if your eye test is picking up on something interesting, or if there's a stat or trend you'd like me to look more deeply into. You can hit me up on Twitter @cfrelund. As with any great analytics department, the more collaborative this is, the more value we can create.
The lines below provided by Caesars are current as of 2 p.m. ET on Friday, Sept. 10 unless otherwise noted below.
Trend to watch: Play-action passing on first down
Here's a look at league-wide percentages of passing plays called on first down over the last five seasons (listed in order from 2016 to 2020): 49.9, 47.1, 50.1, 48.1 and 49.7. Now, here's the breakdown of play-action rates on first down during that same span: 30.5, 34.9, 35.7, 36.8 and 37.7. While passing on first down has been relatively steady over the last half-decade, the utilization of play-action on first down has increased. The trend makes sense with defenses having to honor both the pass and the run on first down, so spacing would likely be more favorable to the offense. But does the strategy work?
After having my math vetted by two third parties, the answer is that there is a demonstrated correlation to increased scoring. Teams that used play-action on at least 20 percent of first-down plays earned an average of 0.8 more first downs and touchdowns per game than those that did not. Teams that used play-action on at least 20 percent of first downs and also ranked in the top 10 in quarterback, running back and wide receiver units based on win share were the most successful, with 3.2 points added to their score per game compared to the other group. This aligns with the eye test, but even teams without as much top-tier talent saw an increase in points added when they used play-action on first down in 2020.
Teams with at least two units (QB, OL, passing offense, rushing offense) ranking 16th or better in win share all saw at least a 0.7 increase in points added per game when using play-action on first down. I could write a whole white paper here, but the bottom line was there is a correlation between play-action passes on first down and increased offensive success in 2020. Here's your math jargon for the week: correlation doesn't imply causation. In other words, I don't mean to suggest teams should use play-action on every first down in 2021, but rather, this was a driver of success in 2020, especially for teams with the right personnel. Now we'll be able to see if that trend continues or if defensive schemes adapt enough to lessen its impact in 2021.
Impact of Ravens' injuries on projected win totals
At the onset of training camp, the Ravens were forecasted to have an AFC North-leading 11.1 wins this season. After the loss of J.K. Dobbins and more roster developments throughout the league, the simulations I ran last week brought that number down to 10.8, which was just below the Browns (10.9). After Thursday's injuries to Gus Edwards and Marcus Peters, the Ravens' total is down to 9.6. The other AFC North teams took the biggest shares of the win totals redistribution, which propelled the Browns into a more commanding first-place forecast (now up to 11.2). The other ripple effects were a 3.2 percent increase in the Chiefs' chances to earn a first-round bye (Baltimore hosts Kansas City in Week 2), and the Bears, Chargers and Broncos -- who are all projected to miss the postseason as of now, but remain within striking distance -- each increased their playoff berth odds by at least 1.2 percent. It's worth noting that the Ravens still have the seventh-best forecasted win total in the AFC, so all isn't lost. It's just a much tougher road (playoff odds decreased to 53.3 from 62.4 percent).
One projection I like
Calvin Ridley to earn more than 86.5 receiving yards vs. the Eagles. The Eagles' defensive front ranks sixth-best as a unit per their win share entering the season. The group has demonstrated that it can generate pressure without relying on the blitz (fourth-best at doing this over the past two seasons, per computer vision). However, the secondary ranks 26th ahead of the season. Ridley's precise route running creates a reliable timing pattern for Matt Ryan, which drives Ridley's forecast of more than 86.5 receiving yards vs. the Eagles in 55.7 percent of simulations. Bonus: Falcons RB Mike Davis' rushing total at 44.5 is also low compared to my forecast of 61. It's not to say the Eagles' front is not going to be strong, but the play-calling fit and tendencies of Falcons head coach Arthur Smith, along with Davis' YAC potential, drives this over.
One projection I love
Josh Allen to earn more than 314 combined passing and rushing yards vs. the Steelers. Part of why I love this is because first-down passing (and play-action) forecasts to be a source of the Bills' success against Pittsburgh (and no team passed more often than the Bills on first down last season), bringing this whole article full circle. Last season, Allen averaged 284 yards through the air and 26.3 on the ground (totaling 310.3 per game), so why go above his average against the Steelers' talented defense, especially with T.J. Watt, who became the league's highest-paid defensive player on Thursday, playing? Considering this front's potential to generate pressure, it's probable that the Bills' offensive game plan includes a lot of pressure-diffusing tactics. That's an area where Allen had a lot of success last season, so his projection jumps to 324 total yards. No QB had a higher quick-passing completion rate than Allen's 84.7 percent in 2020, and no QB had a higher air-yards-per-attempt average when under pressure than Allen's 13.2 (both per Next Gen Stats). Lastly, Allen throws for more than 2.5 touchdowns in 42 percent of simulations. When a projection for touchdown passes is greater than two for a QB who also rushes, think of 35 percent as a good threshold.