NFL Power Rankings, MNF Advanced Review
An ugly injury for Kyler Murray and the process of developing Power Rankings
The adjusted scores quantify team play quality, with emphasis on stable metrics (success rate) and downplaying higher variance events (turnovers, special team, penalties, fumble luck, etc). Adjusted expected points added (EPA), in conjunction with opportunity-based metrics like total plays and drives, projects adjusted points. Adjusted scores have been tested against actual scores and offer slightly better predictive ability, though their primary benefit is explanatory.
All 2022 Adjusted Scores are available to subscribers via Google Sheet. Check them out. I welcome your feedback.
Kyler down, turnovers sink the Cardinals
The long-term story of the game is the Kyler Murray injury, which hasn’t been confirmed as an ACL tear at the time this is published, but doesn’t look good.
One of the first things that comes into my mind with a major injury is where a player sits in the contract cycle, and the silver lining is that Murray got a new contract in the offseason. Murray got $30 million this year as part of his new deal, has been paid a total of around $60 million in his career, and there’s another $37 million in full guarantees left in the deal.
Colt McCoy is one of the better backups in the league, but the Cardinals at 4-9 are out of the playoff picture, in the bottom-four of the NFC. The adjusted scores lean slightly to the Cardinals, who were up 13-7 in the second quarter despite a missed 50-yard field goal, and were looking to add to the lead before failing on a good decision to try and convert a 4th & 1 from the Patriots 32 yard line with less than two minutes left in the half. Then the bottom fell out for outlier bad plays for the Cardinals in the second half, including a fumble-6 on a DeAndre Hopkins reception (-8.1 EPA) and an interception inside the red zone (-4.4 EPA). Overall, the Cardinals were 1-5 on fourth down conversions, losing a combined 6.6 expected points.
The Patriots had a 40% offensive success rate, Mac Jones averaged only five air yards per attempt, and passed at nearly a 70% rate despite being more efficient on the ground (+0.22 EPA per play) than through the air (-0.11). As of today, the 7-6 Patriots are hanging onto the seventh seed in the AFC with tiebreakers over the Chargers and Jets, but they have a tough schedule remaining: @LV, CIN, MIA, and @BUF. The early odds have the Patriots as basically a pick against the Raiders in Week 15. I have the Patriots at only a 25% chance of making the playoffs, with the toughest schedule remaining of teams in contention.
The Unexpected Points NFL Power Rankings
It took me a week to update my data sources and code, but now I have my team strength numbers ready to go to project forward the rest of the season.
I’m walk through my process for building them in some detail, but don’t want to bore readers with methodological choices that ultimately end up making marginal differences in the team strengths and projections.
These numbers are built on expected EPA per play numbers that also drive the adjusted scores, meaning outlier plays, fumble luck and other high variance aspects of the NFL are discounted. I adjust those numbers further for team strength of schedule, the technical process laid out here by Alok Pattani, former ESPN Analytics Specialist who helped create some of their metrics, including QBR.
Armed with efficiencies and strength of schedule adjustments, the final tweaks are decaying past results (most recent game will be more representative of the future than Week 1) and using a combination of art and science to adjust for injuries, especially at the quarterback position. My injury adjusted are a combination of base rates for certain positions, past historical performance for the individual players and some ole’ fashioned opinion on players that we lack sufficient data to judge.
The Results
Below is a plot for visual learners showing the team and opponent adjusted scoring rates (similar to EPA on a points basis), with arbitrary tiering illustrated with the diagonal lines. The tiering gives 60% of the weighting to offense and 40% to defense, reflecting the higher stability of the former.
The Chiefs on in their own tier following some struggles for the Bills, who join the Eagles, Bengals, 49ers and Cowboys in a larger second tier. My numbers have the Chiefs as significantly better offensively, with Patrick Mahomes leading the NFL in efficiency, and a sustainable advantage in passing rate over expectation. The Chiefs have had one of the more difficult strengths of schedule on offense, but it hasn’t mattered.
You can compare my tiering to those available at RBSDM.com and posted by Ben Baldwin that use unadjusted seasonal EPA per play. At the top of the NFL, the biggest differences are my numbers adjusting down the Eagles for strength of schedule, and moving the Jets up for the opposite reason.
The third tier of NFL teams including mostly playoff contenders, with the Browns and Packers mixed in, two teams who have been unlucky generally this season according to my adjusted scores, and playing better of late for at least the Packers. The next tier includes most of the rest of the NFL, with the injury riddled Rams near the bottom and the Texans in a tier unto themselves. There is a funny symmetry with the Chiefs and Texans both having slightly below average defenses and sitting at opposite ends of total team strength.
I use these power rankings to derive the amount of points team’s are expected to win by against an average opponent on a neutral field (Chiefs at +7.9), which can inform our decisions on who will win upcoming games, or even cover the spread for those inclined to spice up the slate with some wagers.
I would not have guessed that the Chiefs would get adjusted ahead of the Eagles since the narrative is how 4 NFC East teams could make the playoffs and the AFC West is much worse than was expected. Interesting stuff.
There’s another 73.3M fully guaranteed on Kyler’s deal (38m in 2023 and 35.3 of 2024 salary)