NFL Week 14 Advanced Review
NFL Week 14 Advanced Review
Herbert v. Tua debate is over, a Purdy good performance, #OurJaguars making a run?
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.
Week 14 Takeaways
Tua vs Herbert debate is over (for now)
It was billed as a heavyweight matchup between two sequential picks in the 2020 NFL draft, but ended up a one-sided affair. Tua Tagovailoa struggled for the second straight week, posting negative EPA per play and his worst completion percentage over expected at -21.4%. This was the first week Tua didn’t have positive EPA on passes attempts when excluding interceptions, and that’s with +6.6 EPA added on a 60-yard touchdown to Tyreek Hill on 3rd & 5 (the most impactful of the game).
Herbert’s stats weren’t eye-popping stats (0.11 EPA per play, eighth highest of the week), but it’s difficult be efficient with an average depth of target of only five yards. Herbert needed his receivers to produce outsized yards-after-catch, and they significantly underperformed expectation based on their target locations. Watching the game (something I do every now and then), you could see how difficult it was for Herbert to generate any offense under constant pressure. It was an elite performance, no matter what the stats say. The Chargers were the biggest winners in playoff probability of the week, nearing 60% odds to get in, and would have been under 20% with a loss.
Tua’s unadjusted efficiency this season is still second behind only Patrick Mahomes, but back-to-back struggles in highly anticipated matchups has mitigated much of the positivity around his breakout campaign. The Bayesian forecast for Tua’s passing efficiency is still slightly higher than Herbert’s, but it’s now close after a big drop from Tua in recent weeks. Mentally adjusting for surroundings, it’s fair to give the advantage to Herbert going forward, and I’ll post later this week on how we can make adjustments more concrete while sit incorporating EPA into the process.
Brock Purdy is The Way, The Truth and The Light
Brock Purdy posted the fourth highest EPA per play of the week (+0.44) after having the 13th best number in Week 13 (+0.10). The trend tells us a week or two from now Purdy will be the most efficient quarterback in the NFL.
In all seriousness, it’s been a good start to Purdy’s career, without the need to get carried away. In back-to-back weeks, Purdy has maxed out the highest leverage and most unstable aspects of quarterback play, averaging significantly higher EPA per play on third downs last week (+0.29 EPA per play to -0.03), then posting a massive advantage under pressure versus from a clean pocket against the Bucs (+1.3 vs -0.4).
It’s important to perform well in the most difficult situations to be an elite quarterback. But over a large sample, quarterbacks have to be able to execute on early downs and from a clean pocket to consistently be above average. It’s too early to make any sweeping conclusions about who Purdy will be the rest of his career, or even the rest of this season.
That said, but we can’t bury our heads in the sand and refuse to weigh the evidence Purdy has produced. I feel better about the 49ers now than I did when Jimmy Garoppolo went down, but also agree with betting oddsmakers that an incremental positive shift in team team strength seems right (49ers’ Super Bowl odds moved to +1000 from +1200 over the weekend).
Eagles and 49ers separate from the Cowboys
Adjusted score differentials for the Eagles and 49ers weren’t as strong as the actuals, but both were in the double-digits. The Cowboys scraped by with a four-point victory against the Texans, who have the worst record in the NFL and are the lowest ranked team according to advanced metrics.
The Cowboys defense was strong, holding the Texans offense to 15th percentile efficiency. But Dak Prescott and the offense struggled mightily. The Cowboys had flat EPA per play, the team’s first below average output since Week 10, and only the second subpar effort since Prescott’s return in Week 7.
This was likely a blip on the road to the playoffs for the Cowboys, but enough to move them below the Bengals in in my power rankings, down to sixth and substantially below the Eagles (up to second) and 49ers (fourth). The Cowboys still have the No. 2 defense is opponent efficiency, but the 49ers are No. 1 on that side of the ball, and the Eagles are close by at No. 4.
Bills limping down the stretch
The Bills were the clear No. 1 team in my power rankings the first nine weeks of the season, but fell out of the top-two for the first time following this disappointing win against the Jets. The team’s adjusted score differential was +54 the first seven games of the season and +14 the last six games.
This was the least efficient game of the season for Josh Allen (-0.1 EPA per play), and Allen has ranked outside of the top-12 in efficiency four of the last six weeks. Allen is still a top-three quarterback for the season, but now there’s a lower efficiency differential between him and No. 10 than it would take to get to No. 1.
On a positive note, as long as the Bills continue to win, however ugly, they’ll hold onto the No. 1 seed and a playoff bye. The Bills finish the season at home against the Dolphins, at the Bears, at the Bengals and at home versus the Patriots. It’s a harder schedule than that of the Chiefs (@HOU, SEA, DEN, @LV), but my numbers give the Bills a slightly better than 50% chance of ending up on top of the AFC.
Forget 2023, the Jags have a shot at the playoffs this season
My adjusted scores had the game differential only two points in the Jaguars favor rather than 14, but it was still a good day for Trevor Lawrence & Co. The second-year quarterback was the fourth most efficient quarterback this week, most importantly completely avoiding interceptions and sacks. Lawrence has posted positive EPA every game since Week 9, with 10 touchdown passes and zero interceptions over that stretch.
The Jags are now only two games behind the Titans and have banked a win in their first head-to-head matchup. If they can win one more game than the Titans over the next three weeks, the two teams will face each other again in Week 18 to decide the AFC South. Both teams have a layup game against the Texans and a meeting with Cowboys. The remaining game is the Jets for the Jags and the Chargers for the Titans, equally strong teams according market-derived power rankings.
Titans weren’t good, weren’t that bad either
I know football guys hate it when nerds act like defense doesn’t have agency in the results of the opposing offense. That said, it’s hard to deny that the Titans offense was unlucky with turnovers in this game. They should have lost the game, but not in dominating fashion.
The Titans lost all three fumbles (Jaguars didn’t lose their fumble), totaling a massive 14.1 expected points lost. Add on top of that a 1st & 10 interception on a miscommunication (-5.8 EPA), and the negatives were too much to for the Titans to overcome. The Titans opened the door slightly for someone else to take the AFC South, but their their playoff probability is still roughly 80%.
Are the Seahawks turning into a pumpkin?
The Seahawks have lost three of their last four games, and I have them with less than a coin flip’s chance of making the playoffs. Some of their early season magic has faded, especially on offense and specifically for the odds-on comeback player of the year, Geno Smith. The first four weeks of the season, Smith was a top-five efficiency quarterback, but ranks 18th from Week 5 on. In fact, Smith’s total EPA was higher the first four weeks (30.9) than the following nine (29.7).
Anyone watching the RedZone channel still sees more than a couple incredible throws from Geno on a weekly basis. His high-end plays aren’t the problem; his negative plays are the issue.
The first four weeks of the season, Geno lost roughly five expected points per game to interceptions, sacks and fumbles. In the last nine games, that number has ballooned to 12 expected points lost on a weekly basis. No matter how good Geno has been outside of his mistakes, they caused enough downward pressure on his efficiency to make putting up elite numbers nearly impossible. It’s never this simple, but if Geno can’t clean things up, the Seahawks will probably miss the playoffs.
Browns done at 5-8, but 8-5 by adjusted scores (yeah!)
The return of Deshaun Watson hasn’t been the offensive boost the Browns were hoping for, especially with Jacoby Brissett playing with above average efficiency as a starter. Watson has averaged -0.15 EPA on 78 plays in two starts (only Baker Mayfield has been less efficient), despite generating a massive +11.7 EPA on scrambles and designed runs. Nothing is clicking with the Browns passing game right now.
You can’t score points if you can’t sustain drives, and the Browns have gone a combined 8-27 on third downs the since Watson’s arrival, losing a total of 10.9 expected points on those plays.
Kevin Stefanski has also called a strangely conservative game with the quarterback they traded away a boatload of picks for and signed to the biggest fully guaranteed contract in NFL history. The Browns passed at a rate 17% under expectation in Week 13 and 2.2% under this week, while never leading the Bengals and trailing the last 40 minutes of the game. On fourth down the last two weeks, Stefanski chose to kick four out of the six times there was at least 1% win probability to gain by going for it.
I get that Watson hadn’t played in nearly two years, but a team that couldn’t afford to lose the rest of the season needed an outlier outcome to make the playoffs. Stefanski should have been turning the risk dial all the way up, not making decisions as if he had a replacement-level quarterback.
Oddsmakers were right about the Vikings, but their pass offense was a positive
The Vikings status as underdog and then loser was the story of the game, but their offense showed some positives. Yes, the 10-2 Vikings lost by 11 points to the Lions, but the adjusted scores showed a more narrow difference in team strength. The Vikings lost out in high variance plays (-13 EPA in special teams and penalties), and didn’t recover either of their fumbles (-6.3 EPA).
This was, by far, Kirk Cousins’ most efficient game of the season, posting an EPA per play of 0.52 (No. 1 for the week) and completion percentage over expectation of +17.8% (again, No. 1), all on high volume. The Lions are the third worst pass defense in the NFL according to my schedule-adjusted numbers (only better than the Raiders and Bears), but beating up on bad defenses is something a good offense does.
The Vikings rushing game, on the other hand, was stuck in the mud. Dalvin Cook fumbled on when looking to step up a potential jump-pass, and on traditional rushing touches the Vikings lost 10.2 expected points with a 12% success rate.
As I explained last week, these results don’t materially affect a Vikings team locked into the Division title and with no real chance of getting the No. 1 seed and a bye. The last few weeks of the season will be about signs that the most important aspects of play improve for the team, and this week you have to like a top-notch performance for the passing game, the most stable and important factor in team strength.
The good, bad and ugly in 4th down decisions
I discussed the Raiders’ failure with fourth down decision in the TNF review. I didn’t expect to see #OurRavens with an even bigger loss of win probability. The Ravens punted twice on 4th & 1, once from the Steelers 46 yard line with a six-point lead. It’s understandable that John Harbaugh would have less faith in Tyler Huntley than Lamar Jackson, but a competent rushing quarterback should have made the short-distance tries no-brainers.
Guess what? Andy Reid gave up significant win probability refusing three strong go-for situations this week: Rinse, repeat.
The Cowboys, Vikings and Chargers were the only teams this week to go-for-it more often than not in advantageous situations. The Cowboys added 6.9% expected win probability on those decisions, but the result was lost EPA after failing to convert from the opponent’s 1-yard line. The Vikings also had some bad luck with good decisions, failing on 4th & 1 their first drive from their own 46-yard line. More often than not, Kevin O’Connell’s aggressive mentality will pay off down the road. Brandon Staley is trying to work his way out of the nerd doghouse with a week-high expected 7.5% win probability added on fourth down decisions. The Chargers covert one of two tries, netting an expected point in the process.