NFL Week 7: Giant Overreaction

Welcome back! Last week’s free play on the Niners was sweat-free for the vast majority of the game as they proved to the world that their defense is as advertised. The week finished 3-3 for +0.14u, moving the season to 22-17 (56.4%) for +6.79u (11.2% ROI). Before we move on to this week’s matchup, I wanted to highlight the college basketball announcement I made yesterday. The donation run we made last year that raised $25,000+ will be returning ($20+ donation to Doctor’s Without Borders gets you the last month of plays), while those interested in a full season package can donate $20+ to get a $100 discount code to bring the price from $299 to $199. That discount code can also be used on NFL, which would bring it from $329 to $229 for the rest of the season.

Arizona Cardinals (2-3-1) vs. New York Giants (2-4)

After six weeks, a lot of attention has been drawn towards the improvement the Cardinals’ offense has shown this year. They currently rank 17th in points scored after a year in which they finished dead last. Kingsbury truthers, however, have conveniently picked this week to highlight the improvement given that the Cardinals just finished playing the Bengals (31st in defensive DVOA) and the Falcons (29th in defensive DVOA). Prior to those two games, the Cardinals ranked 25th in points scored.

Granted, the Giants are not a defensive powerhouse (they rank 23rd in defensive DVOA). However, sometimes the deeper nuances of marginal improvement from team to team is what makes the largest differences at the professional level. This is especially true when the team’s performance is steadily improving in that regard, as the Giants have over the course of the season. In particular, let’s look at how their pass defense has improved over the first six weeks (Note: For DVOA, positive numbers represent allowing more scoring):

Granted, the Giants’ pass defense is still poor (28th in pass defense DVOA) but the gap between the Giants and the 30th ranked Bengals is the same size as the gap between the Giants and the 19th ranked Eagles, and the gap between the Giants and the 31st ranked Falcons is the same size as the gap between the Giants and the 7th ranked Broncos.

A large reason for that gap (despite the relative closeness in rankings) is the Giants’ ability to pressure the quarterback, as they rank 11th in adjusted sack rate. That doesn’t bode well for the Cardinals whose offensive line ranks 28th in adjusted sack rate allowed, especially given the defensive lines they’ve faced. Aside from the Panthers (3rd), the Cardinals have faced the Lions (25th), Ravens (26th), Seahawks (29th), Bengals (30th), and Falcons (32nd). Those five bottom-dwelling teams averaged 2.6 sacks per game against the Cardinals, whereas the Panthers sacked Kyler Murray 8 times in their affair.

Let’s take a look at the other side of this coin: the Giants’ offense vs. the Cardinals’ defense. The Cardinals pass defense has been very poor this year, ranking 29th and performing just a smidge better than the Bengals. They do get Patrick Peterson back this week from suspension, but the track record for first games back (even for elite talents) isn’t promising, so the improvement that he’ll bring and that everyone seems to be penciling in may be exaggerated for this week. Even if Peterson does play lights out, the talent the Giants get back this week (Saquon Barkley and Evan Engram) is far more impactful. In fact, this week easily marks the most offensive talent the Giants have had on the field so far this season, given that the Giants will have Daniel Jones, Saquon Barkley, Golden Tate, and Evan Engram together for the first time.

Engram’s return is particularly noteworthy given the Cardinals’ inability to guard tight ends, especially those on the talent level of Engram. Through the first six weeks, opposing tight ends have put up the following performances:

  • W1 T.J. Hockenson: 6 receptions, 131 yards, 1 TD
  • W2 Mark Andrews: 8 receptions, 112 yards, 1 TD
  • W3 Greg Olsen: 6 receptions, 75 yards, 2 TDs
  • W4 Will Dissly: 7 receptions, 57 yards, 2 TDs
  • W6 Austin Hooper: 8 catches, 117 yards, 1 TDs

Outside of the Bengals’ weak performance in Week 5, opposing tight ends have torched the Cardinals for an average of 7 catches, 98.4 yards, and 1.4 touchdowns.

All in all, my model sees this as a game between a firmly below average team going against a firmly just-above-bottom-of-the-barrel team. Then you add in the fact that the Giants are on extended rest coming off of a Thursday Night game and the Cardinals are a west coast team traveling to the east coast for an early game, and the model sees more than enough value to make NYG -3 -100 a play for 1.35 units.

NFL Week 6: California Love

Yes, I know – it has been a while. Three weeks ago my dog had to have his knee surgically repaired, and the recovery has been slower than anticipated with some unexpected hurdles along the way. But he is showing progress and is doing a lot better these days, which means I have time to sit down and put together one of these write-ups for you fine folks.

The model thus far is 19-14 (57.6%) for +6.66 units (12.5% ROI). I’m pretty pleased with those marks given that there have been 13 quarterback changes in the first five weeks and they are the largest and arguably the most difficult adjustment to make to a team’s evaluation. If you are interested in joining, the model plays have been prorated down to $349 for the rest of the season given that a month of plays is already in the books.

San Francisco 49ers (4-0) vs. Los Angeles Rams (3-2)

The matchup I’ll be looking at this week is one that I anticipate a massive divide on across handicappers, modelers, and bettors everywhere. The Rams and Sean McVay have been adored by the analytics community for quite some time now and for good reason. In 2018, they finished 2nd in offensive DVOA including finishes of 5th in passing DVOA and 1st in running DVOA. As a result the Rams only trailed the Chiefs last year in scoring, averaging 31 points per game. With Cooper Kupp returning from injury this year, there were expectations that this offense could generate similar (if not better) results in 2019. Kupp has been more than advertised, averaging 8.2 catches and 101 yards per game and has been responsible for more than half of all of Jared Goff’s touchdowns (4 of 7).

Despite that, a lot of data points suggest that this Rams offense has taken quite a step back, especially through the air. A lot of it has to do with the inefficiencies from the offensive line and Jared Goff. After finishing seventh last year in yards per attempt (8.1), Jared Goff now ranks 17th with 7.4 yards per attempt. Much of this comes from the fact that the Rams offensive line has graded poorly in pass blocking this year, allowing the 12th most QB hits per game after allowing just the ninth fewest in 2018. That obviously doesn’t bode well for a quarterback who only completed 41% of his throws while under pressure last year.

Those numbers alone should be concerning for the Rams, but it gets much uglier when you look at the defense they have to face on Sunday. For as much attention as the Patriots’ defense has received this year, the 49ers have been just behind them as a complete unit. On a per game basis, the 49ers currently rank:

  • 4th in points allowed
  • 2nd in yards allowed
  • 2nd in defensive DVOA
  • 1st in pass defense DVOA
  • 5th in rush defense DVOA
  • 5th in sacks
  • 7th in adjusted sack rate
  • 8th in 3rd down conversion rate
  • 2nd in interceptions
  • 3rd in penalties committed and penalty yards

Credit needs to be given to the 49ers’ front office for the amount of defensive talent they have accrued through the draft. In the last five drafts, the 49ers have hit on the following Day 1 and Day 2 picks on defense: Nick Bosa, Fred Warner, Solomon Thomas, Ahkello Witherspoon, DeForest Buckner, Arik Armstead, and Jacquiski Tartt. That is quite the haul, and having a Kyle Shanahan offense to practice against for the past two seasons has undoubtedly played a huge role in their development.

To make matters worse for the Rams, Brandin Cooks is currently in the concussion protocol and Todd Gurley might be limited in some significant capacity. Cooks brings a deep threat element to that offense that will be hard for Josh Reynolds to replicate, which would make the jobs of Cooper Kupp, Robert Woods, and Gerald Everett (who just came off of a game in which he logged career highs in targets, receptions, and yards) much more difficult. Gurley’s impact is a bit more arguable, depending on where you stand in the “Do running backs matter?” debate.

But the Rams aren’t the only ones with noteworthy injuries coming into this week’s game. Starting right tackle Mike McGlinchey will be out for the next 4-6 weeks, which hurts a bit more than usual given that starting left tackle Joe Staley is already out and with the 49ers leading the league in rushing rate (57%). Fullback Kyle Juszcyk is also out, but the 49ers are arguably the deepest and most versatile team at the position and will surely find a way to distribute his touches.

Being deep and versatile best describes the 49ers as a whole so far this season. Sure, they have the capability just like any team performing well to start the season to come falling back down to earth. As for now, they currently rank 10th in pass offense and 7th in rush defense, and rank 2nd in pass defense and 5th in rush defense. Altogether, the 49ers are the second strongest team in my model as of this writing. That was certainly not something I expected heading into this season. In fact, the 49ers have outperformed my priors so much that their expected wins vs. league average has increased by 2.67 wins through just five weeks. The next best improvement has been New England, who has improved by 1.65 wins so far this season. To provide more context, there were only four teams through Week 5 last year that saw an improvement of one win or more, and only one team improved by two or more: the Kansas City Chiefs.

The Rams on the other hand have regressed a bit more than I had expected, finding themselves as the tenth best team in the model and decreasing their expected wins vs. league average by 0.67 wins. As for this week’s matchup, the model has this game made to be LAR -0.2 after accounting for game specific factors, rest discrepancies (Rams coming off TNF, 49ers coming off MNF), injuries, etc. That makes SF +3 +104 good for a 1.4 unit play.

NFL Week 3: Backups Galore

Last week’s free play on the Jaguars played out like a treat. The Jaguars’ defense, fresh off of a game where they allowed 40 points, held the Texans’ offense to just 13 points. That was right around the average for the Jaguars following games in which they allowed 21+ points (11.1) which I covered in last week’s write-up. This allowed the Jaguars’ offense, led by backup Garnder Minshew, to do the bare minimum to cover the 9 point spread. Speaking of backups, last week’s model plays went 5-3 for +4.72 units with three of the plays being on games in which backup quarterbacks were unexpectedly needed (we went 1-2 on those games). To keep with the theme, we’ll be diving into a game that involves a backup.

New Orleans Saints (1-1) vs. Seattle Seahawks (2-0)

Simply put, there just isn’t a categorical edge I can find for the Saints when looking at this matchup. First let’s look at how the Seahawks’ offense stacks up against the Seahawks’ defense. The Seahawks currently rank seventh in the model passing offense and will be going against a Saints’ pass defense that ranks in the bottom third of the league. Seattle also ranks just outside the top third in rushing offense and will be going against a Saints’ rush defense that ranks 27th in yards allowed per game, 28th in yards allowed per attempt, and 25th in rushing first downs allowed per game.

On top of that, the Seahawks should have no problem putting together long drives given that the Saints’ defense is already allowing the fifth most yards per drive in the league (only edging out the Giants, Raiders, Dolphins, and Redskins) and ranks fifth worst in generating three-and-outs. The Seahawks should be able to put together a few drives that land them inside the red zone, where they are just one of three teams to have scored a touchdown on all of their trips there. But even when the Seahawks fail to push the ball into Saints territory, they will have All-Pro punter Michael Dickson to help pin the Saints deep in their own territory.

As for how the Saints offense might fair, the model expects the Saints’ offense without Drew Brees to be in the bottom third when it comes to both passing and running the ball. The Seahawks’ pass defense only ranks around league average in the model, but their run defense ranks tenth. So things may look bleak on offense for the Saints if Sean Payton were to opt for a more run-heavy, possession-focused type approach in lieu of Brees’ absence. If not, I don’t expect the Saints to find much repeated success through the air whether it’s Teddy Bridgewater or Taysom Hill behind center on Sunday. I expect the former to be a better quarterback but a harder piece to either fit into the Saints’ offense or to adequately adapt the offense to in a week, whereas the latter is a better fit to the existing offense (obviously given that he already had a role prior to Brees’ injury) but is less talented and very likely to see diminishing returns the more he is used.

Given everything highlighted above, it’s not hard to see why there would be value on the Seahawks being favored by four points. As most know, the Seahawks possess one of the better home field advantages in the NFL. Even when considering that the average homefield advantage in the NFL has seen an approximate 20% decline in the past two decades, it would be hard to argue that the Seahawks’ homefield advantage is any lower than the equivalent of three points. So intuitively, if you subtract out those three points we are left with a line that suggests that the Seahawks are just a point better than the Drew Brees-less Saints on a neutral field.

That should be a hard buy for anyone, and my model agrees. Prior to the season, my model had the Saints as the fourth best team at 9.64 wins and the Seahawks as the seventh best with 9.38 wins (Note: model wins are expressed in terms of a team playing 16 games against a league average team). While Seattle has managed to hold steady in the model through two weeks of play, the Saints are a much different story. As a result of Drew Brees absence and the Saints’ poor defensive play to start the season, the Saints have been downgraded to 6.27 model wins, good for 27th in the league. The result is a 68.2% probability that the Seahawks would beat the Saints on a neutral field, which is already higher than the implied probability of the Seahawks moneyline for the game. After accounting for game-specific factors the model makes this game Seahawks -8.4, which makes SEA -4 a 2.2 unit model play.