There is considerable optimism among New York Giants fans for a successful 2023 season that builds on the surprising progress made in 2022 under new head coach Brian Daboll and his staff. After all, the Giants seemed to fill three of their biggest needs in the draft, at cornerback, center, and wide receiver, while bolstering the linebacker position, interior defensive line, and receiver group in free agency and trading for an elite (when healthy) tight end.
Of course other teams didn’t spend the off-season just twiddling their thumbs. Some seemed to improve a lot, others not so much, and yet others are tearing things down and rebuilding. “Expert” projections of how teams will do in 2023 come down to a few factors:
- How a given team’s success or failure in 2022 is viewed. In the eyes of some analysts and NFL people, the Giants did it with smoke and mirrors last season: Luck broke their way in several games (Tennessee missed field goal, Christian Kirk barely stopped at the 1-yard line on final play in Jacksonville, Darnay Holmes end zone interference non-call in Washington). Opponents began to catch up with what Daboll and Co. were doing in midseason and the Giants had more problems after that. Thus a predicted “regression to the mean” in 2023.
- How a team’s offseason went. This is all in the eye of the beholder. As Ed Valentine noted, Bill Barnwell of ESPN thought the Giants had a terrible off-season. On the other hand, Pro Football Focus gave the Giants an A grade.
- How difficult a team’s schedule is compared to the previous season. This is something we can actually put hard numbers on. The question is: Are those numbers meaningful?
Strength of schedule
The simplest thing one can do is to just use the previous season’s records of the opponents on a given team’s schedule to see who can be expected to fare better or worse this season. Tankathon does this before the season based on last year’s records and then updates it during the season with records for the current season (which determine draft order when two or more teams finish with the same record). Here is the preseason strength of schedule (SOS) ranking, in decreasing order, for the top 16 teams:
More than anything else, SOS depends on the division a team plays in and what division from the other conference they play. The Giants play in what looks like one of the toughest divisions in the NFL. They also have the misfortune of playing all the AFC East teams this year, any or all of whom could make the playoffs and several of which could be Super Bowl contenders. As a result, their SOS is ninth most difficult in the NFL. Yikes. Let’s hope Daboll doesn’t have to send out the other kind of SOS by mid-season for a sinking Giants’ ship.
But how predictive is the strength of schedule based on previous season records of current year opponents? Here are the strength of schedule rankings before the 2022 season began (opponent winning “percentage” next to each team):
The Giants supposedly had the fourth easiest SOS (.465) in 2022, which is why some people thought they might have a chance to at least be mediocre despite their talent deficiencies. That is also why some people continue to predict a regression this year, i.e., going from the 29th to the ninth-most difficult schedule equals fewer wins in 2023.
But the narrative that the Giants had an easy schedule last year is false. Now that the 2022 season has actually been played, we know that Philadelphia, Minnesota, Jacksonville, Seattle, and Detroit were all much better teams than their pre-season projections. A few 2022 Giants’ opponents were worse, too (Tennessee, Green Bay, Indianapolis). But on balance the Giants’ opponents were better than the NFL average (.526 winning percentage) rather than among the worst as predicted using 2021 records.
The Giants weren’t alone. The Rams were predicted to have the toughest 2022 schedule in the NFL (.567). No wonder their record was so bad, right? Actually, their opponents wound up with a .517 winning percentage, not a tremendous challenge for a Super Bowl champion. Neither number meant anything because the Rams themselves were a lot worse in 2022, due to injuries and player departures necessitated by the way they put together their Super Bowl roster. Meanwhile Dallas was tied for the easiest schedule pre-season (.462). Their opponents actually finished with a .507 winning percentage, middle of the pack.
Here is the relationship between the previous year’s winning percentage of current year opponents and actual current year wins over the 2016-2017 NFL seasons from a study by Sharp Football Analysis:
There is a very slight downward trend that is statistically meaningless (R-squared = .0019, p-value = 0.73 for those who want to know). The same plot for a longer period (2010-2017) actually shows an equally meaningless very slight upward trend.
- The Giants’ 2022 season was probably not a mirage. They were actually a pretty good team. They didn’t just feast on the dregs of the NFL.
- Pre-season SOS is a not a meaningful indicator of what a team will face once the season begins.
So why do we keep on hearing that the Giants have a tough schedule in 2023?
Market projected win totals
It should be no surprise that how teams did last year doesn’t have all that much bearing on how they will do this year given that teams change from year to year. That’s one purpose of the off-season. So a potentially more plausible approach to strength of schedule rankings is to try to combine the previous year’s information on opponent strength with more current information on how teams may have or have not improved since then. There are several ways to do this.
One of the most popular is to use the Vegas betting markets’ estimates of projected team win totals. Here is that estimate (easiest schedule at left) for 2022 from Sharp Football Analysis:
Well, that didn’t work out too well. In this metric the Giants should have had the easiest schedule in the league last year, not the slightly harder than average schedule they actually wound up having. It did a little better on the Cowboys, predicting the 10th easiest schedule rather than the easiest. Kansas City, with by far the toughest 2022 schedule according to the pre-season betting markets in the supposedly fearsome AFC West, wound up playing six games against the Denver and Las Vegas dumpster fires and the injury-depleted Chargers, plus under-performing Arizona, Indianapolis, Tampa Bay, Tennessee, and the Rams.
One problem with Vegas forecasted win totals is that although sports books do considerable research, they don’t actually try to forecast winners and losers of games and seasons. They forecast how people will bet on games, and they adjust their spreads to reflect how betting is going in the lead-up to a season. And they are not immune to false narratives. Ben Wolbransky of The Thirty Third Team looked at how actual team win totals compared to Vegas forecasted totals over two decades of NFL games:
There’s obviously a positive correlation, and the mean average error is 2.2 wins. So information about how teams have improved or regressed over the off-season does have some effect. But there is just as obviously a tremendous amount of scatter, with individual forecasts being off by as many as 8.5 wins.
Conclusion: The betting public (and professional analysts) can be swayed by developments that don’t matter as much as they think they will, or matter in the wrong direction. Look at the 2022 Denver Broncos, middle of the pack in SOS with or without Vegas forecast influence and with a brand new future Hall of Fame quarterback and a successful offensive coordinator as new head coach joining a team with a stingy defense and lots of talent at wide receiver. They were seen as a possible Super Bowl contender. That ship sank as soon as it left the harbor. It turns out that it matters which Hall of Fame quarterback and successful offensive coordinator you bring in, but bettors, as well as pundits, aren’t as talented as we might imagine at identifying these things. There’s also a considerable amount of randomness, since things like major injuries to key players can’t be anticipated (e.g., 2022 Ravens, Lamar Jackson).
Elo ratings are based on the concepts originally used by physics professor Arpad Elo to rank chess players . As applied by Pro Football Focus to NFL strength of schedule, PFF ELO “takes into account a team’s record from last season, their opponents’ records, home and away games, injuries and roster changes” to assess which teams have easier vs. harder schedules for the upcoming season. This seems like a reasonable attempt to merge each team’s starting point with information on how they changed during the off-season.
In PFF ELO, PFF player grades are the basis for relative rankings, i.e., teams with higher player PFF grades are favored to defeat teams with lower PFF grades. Details about how different players and positions are weighted are behind the curtain, though. Are they wizards or more like the Wizard of Oz? Here’s what they said about the Giants in 2022:
The Giants being a trendy longshot was certainly prescient. But for largely incorrect reasons:
- They wound up not having the easiest schedule in the NFL, as predicted, but rather a schedule of middling difficulty.
- Their most difficult stretch was not weeks 3-6, as predicted (Dallas, Chicago, Green Bay, Baltimore), but weeks 11-14 (Detroit, Dallas, Washington, Philadelphia)
- Neither the Cowboys (4-1) nor the Eagles (5-0) stumbled out of the gate, yet the Giants still made the playoffs, partly because of one thing that did come true - Daniel Jones exceeding relatively low expectations.
So taking the concept with the large grain of salt it deserves, here are the 2022 vs. 2023 PFF ELO rankings:
In 2022, PFF ELO “learned” as the season went on, as Elo systems are designed to do (by increasing a team’s Elo score when it beats a team with a higher Elo score, etc.). So although the Giants were ranked as having the easiest schedule before the season began, we can see that their schedule ranked somewhat closer to middle of the pack by season’s end (as given by position along the x-axis).
That said, PFF agrees with everyone else that the Giants have a tough road to hoe in 2023, with the fifth hardest schedule in the league, behind only New England, Miami, the Chargers and Buffalo. Hence the expectation that the Giants are set up for a fall in 2023, which is what the upper left quadrant of the diagram implies. (PFF’s prediction for the Giants is 8.4 wins.)
The most difficult stretch to navigate? Weeks 3-8 look like quite the challenge: @SF, SEA, @MIA, @BUF, WAS, NYJ. But the only likelihood is that these games will not play out the way we imagine.
Cynthia Frelund has her own model that uses projected 53-man rosters and conducts 100,000 simulations of each game that “account for many contextualized data points that have been proven to correlate to wins and losses based on historical football games that were actually played; personnel is the most important, but other key factors include play-caller tendencies and in-game situations.” Like others, her model has a dim view of the Giants’ 2023 prospects:
She also identifies that early season stretch as the most difficult part of the schedule. Frelund’s accuracy in picking all NFL games was 63.6% in 2022. But her predicted 6.5 win total for the Giants was far off (as was everyone else’s, of course).
Believe in the coaching staff
You’ll notice one thing about all the strength of schedule and win prediction estimates that are out there: They’re mostly about the players, and those players’ prior performance. That’s understandable, because even though we like to project our favorite players “breaking out” in the coming season, there’s no easy way to actually predict that.
Much less thought (or none) is given to the coaches. It’s probably fair to say that most coaching staffs in the NFL are close to neutral - they neither add much nor subtract much from their team’s performance. They are caretakers, and so they don’t much affect their team’s chances of winning.
But there are a few exceptions. On one end of the spectrum, intelligent football people such as Nathaniel Hackett and Kliff Kingsbury proved not up to task of head coaching, at least at this point in their careers.
Then you had Brian Daboll and his staff. Is there a single person who thinks that the Giants could have gone 9-7-1 and won a playoff game if Joe Judge or Pat Shurmur or Ben McAdoo had been given the same roster? PFF briefly alludes to coaching in its assessment of the 2023 Giants:
It’s more than just good coaching, though. There are a number of good coaches and staffs in the NFL, and the Giants will face some of them this season. There’s something different about Daboll, though, that was touched on in a recent episode of The Athletic Football Show. In response to a question about what exactly Brian Daboll and his staff did last year and whether they can replicate it in 2023, hosts Robert Mays and Nate Tice made a key point about the Giants not running an innovative new offense, as for example Sean McVay did with the Rams to create a scheme different from what opponents were used to seeing and then daring them to catch up to it:
“Asking if he can replicate something special...I think is mis-representing why they were good... I don’t necessarily expect the Giants offense to look the same this year as it did last year. I hope it doesn’t... They just looked at the players they had and said, what is the best offense that we can run with these players? The 2022 New York Giants offense looks nothing like the 2021 Kansas City Chiefs offense or the 2021 Buffalo Bills offense...That’s why I’m so bullish on it... Because it’s not about, like, replicating one schematic tweak here or there, or one quirk that allows you to stay one step ahead of other teams...It’s all core concepts. So what does that version of it look like this year? I think that being able to just say, not, this is my system, but these are my players, this is how best to utilize them? That is much more flexible, and you can kind of graft that onto different versions of whatever your team is moving forward. Those are the types of coaches I have the most faith in.”
Mays and Tice are optimistic about the Giants for that reason. It didn’t take long for newcomer Darren Waller to notice the difference from other coaching staffs, as Ed Valentine reported:
Waller said last week that in building a playbook predicated on players, Giants coaches are seeking player input.
“They value our opinions here. As a player, I feel like a lot of places I’ve gone, you’re told to do things a certain way, and you do those things. But here, they ask a lot of questions,” Waller said. “They want to know what you’re thinking, what do you like to do more. So, to offer input is a really cool thing because coach and players have got to be in partnership. We’re all together. We shouldn’t be clashing with each other. We’re all going the same direction.”
None of that means that the 2023 Giants can improve on their surprise 2022 performance despite what looks like a tougher schedule. Injuries to key players, 50-50 plays that go the wrong way, and a dozen other things can lead to a season playing out differently than was envisioned. Giants fans need look no further back than 2017. A surprise 2016 Giants team with a supposed offensive genius as head coach that got “smoked” in the playoffs (to use a Daboll term) collapsed in 2017 to 3-13. Vegas’ forecast for the 2017 Giants was 8.5 wins, and the experts surveyed for this article predicted a record anywhere from 8-8 to 12-4.
But perhaps more than the perceived quality of opponents, the quality and flexibility of the coaches on the Giants’ sideline is the best predictor for the season ahead. Are the Giants more or less “roster-proof” with Daboll and his staff at the helm? The Pittsburgh Steelers have pretty much been that way under Mike Tomlin, who has never had a losing season in his 16 years there, has been to two Super Bowls, and has won one.
Would you sign up for that kind of career from Brian Daboll, Giants fans?