I’m Gone To Carolina In My Mind (Are the Panthers for Real?)

107
I’m Gone To Carolina In My Mind (Are the Panthers for Real?)

When I wrote about the NFC South in my preseason projections, the model did not provide a positive outlook for the Panthers. Yet the team has rushed out to a record of 3-0. This article takes an in-depth look at what the Panthers have accomplished if it is sustainable, and where the model went right and wrong. The aim is to improve our positions involving the Panthers as the season progresses. 

Pythagorean Theorem

Carolina’s over/under win total was 7.5 games. The model went slightly under at 7.18 wins. Having started the season 3-0, the Panthers need to win just five of their remaining fourteen games. This equates to just over 35% of their remaining games.

The Pythagorean Theorem provides a quick and reliable way to make some basic predictions about the balance of the Panthers’ season. The formula is the same basic one you learned in middle school geometry: 

We will adapt it slightly to work for the NFL. Specifically, we will assign the following parameters:

a = points socred

b = points allowed

c = expected win percentage

Then we will write the formula as follows:

 (a^2)/((a^2)+(b^2)) = c^2 

When we run the formula for Carolina, we are left with an expected win percentage of .841. Since the Panthers have won three games and have 14 games remaining, we can achieve an expected win total from the following formula:

(.841*14) + 3 = 14.77 games. 

I do not believe anyone expects the Panthers to win 14.77 games this season. This highlights the danger of small sample size and just how easy the Panthers schedule has been to date. Nevertheless, this is what their current numbers suggest. Realistically, the Panthers, by most models, are one win over expectation through three games.  

Strength of Schedule

The first thing the model noted about the Panthers was that they had the easiest schedule in the NFC South. Each team in the NFL plays three games against opponents that the rest of their division does not play. The model refers to these games collectively as the team’s nonstandard opponents. Carolina had the easiest nonstandard opponents in the NFC South.

TeamOpponent OneOpponent TwoOpponent Three
Tampa Bay Bucs@ RamsBears@ Colts
Atlanta Falcons@ Jaguars@ 49ersLions
New Orleans SaintsPackers@ Seattle@ Tennessee
Carolina Panthers@ TexansVikings@ Cardinals

The projected win percentage for the Panthers’ opponents was slightly worse than the Falcons at 41.18%.

The Panthers have already played the Texans, the weakest team on their list of nonstandard opponents. They have also played the Jets, the weakest team from the AFC East slate of opponents that the NFC South faces this season. The Panthers have faced and beat Zach Wilson, Jameis Winston, and Davis Mills. That list of quarterbacks includes two rookies who were starting for the first time in their careers. 

To date, the Panthers have played the fifth easiest schedule of any team in the league. Their remaining schedule currently ranks as the 10th most difficult. The highlights include the following opponents: Tampa Bay (twice), Dallas, Arizona, and Buffalo. So they will be trading the quarterbacks above, for Brady (twice), Prescott, Murray, and Josh Allen. Of their remaining fourteen games, eight are on the road, including three of their last four and four of their last six games.  

What Has Gone Right

In a word, scheduling. 

But it is more than that. So far, Darnold has gone right. My model has Darnold ranked as the 10th best quarterback. I surveyed other publicly available models and found that they all generally agree and have Darnold ranked between ninth and twelfth. 

Let’s start with Darnold’s career statistics: 

As we can see from the graphic above, Darnold is having a career year across the board. His completion percentage is up over 6% from his previous career high. He is averaging 1.4 more yards per attempt than he has previously. His touchdowns are on about the same pace, but his interceptions have plummeted. He currently possesses a 3 to 1 ratio of touchdowns to interceptions. His previous best was 1.46 to 1. 

The model is not a huge fan of the traditional touchdown to interception ratio for several reasons. It prefers to judge quarterbacks on the percentage of elite-level throws (whether they are caught or not) to throws that should have been intercepted, whether they were or not. What follows is those stats throughout Darnold’s career:

YearElite Throws/Turnover Throws
2018 – 20191 to 1.25
2019 – 20201 to 1.3
2020 – 20211 to 2
2021 – 20222 to 1

This constitutes shocking growth by Darnold. This is Wentz and Allen-esque, and naturally raises the question of whether or not it is sustainable. 

Those that believe that the numbers are sustainable will make the same argument many have made, that we simply have to throw out Darnold’s numbers with the Jets because the coaching fell below even a base level of competence, and he had no weapons in New York. This argument concludes that we are finally seeing the quarterback Darnold can be in the NFL with competent coaching and adequate weapons. 

I somewhat agree with that argument. Darnold had received significantly subpar coaching in New York. Rhule is a monster upgrade in the coaching department. Still, Darnold had some weapons in New York, including one of the same weapons in Carolina, Robby Anderson. 

Here, I turn to the principle of Occam’s razor. The principle is often misstated as the “simplest explanation is usually the best.” That is not entirely correct. I prefer to think of it as when multiple explanations for a given phenomenon exist, the one with the fewest assumptions is more probable. In this case, it requires fewer assumptions to attribute Darnold’s increased efficiency to the opponents he has faced.   

Ultimately, both are likely true. Darnold is playing better because he has better coaching and weapons. And because he has played an easy schedule which included facing two rookie quarterbacks in their first start. So his play should be viewed as semi-sustainable. He will be better than he was in New York, but not as good as he has been through the first three games.  

The other aspect that is going right is indeed the coaching. This is especially true on defense. Rhule’s defense has produced more unblocked rushers to the quarterback than any other team in the league. This is not the result of chance. Rhule is an exceptional coach, and his players are getting free rushes precisely because of the schemes Rhule is employing. Rhule is in contention early for coach of the year. The coaching is sustainable. 

What is Next

In a word, regression.

After their win on Thursday night, most reputable models raised their season win expectations for the Panthers to 9.5 -10 win range. My NFL model remains slightly shallower than those models and has them at 9.16.

We cannot discount the ease of their schedule to date. We also cannot ignore the McCaffery injury. He is out for several weeks with a hamstring injury sustained last Thursday. On the plus side, Christian is getting laid for the first in months, so the over/under there has swung to the over as well:

It is challenging to quantify McCaffery’s importance to this offense adequately. He makes up 18% of the Carolina passing game and 63.4% of the rushing attack. More than that, he indirectly contributes to the success of the rest of the team, even without the ball. This is a massive loss for the Panthers – a team already short on offensive weapons. 

Darnold will regress. Perhaps not back to his NY Jets numbers, but off of the highs he saw in the first three games. The Panthers will regress with him.  

Much of the smart money has been on the Panthers all season. Sitting at 3-0, the public will start to pile on. The model will look to spot fade them when the opportunities present themselves. The Panthers are not as good as their record suggests, and that always creates opportunities for bettors.

About the author:

+ posts

I write about data and sports. I created my first model in 1997 using nothing more than Excel. Currently, I have data-driven models for the NFL, NBA, and World Cup Soccer.

Mathematics is the music of reason.
— James Joseph Sylvester, English mathematician

Leave a Reply