Is Kenny Pickett a Franchise Quarterback?

As a Steelers fan, I wanted to find out if the Steelers made the right or wrong decision drafting Kenny Pickett. I will continue to analyze this decision over the next two to three years as the opinion may change over time. I took tabular quarterback data to provide insight on whether Kenny Pickett is a legitimate draft pick that will become a franchise quarterback, or, after only one year of data, can we see that the Steelers made the wrong decision.

I decided to find out what the data says to help legitimize any gut feelings I may have. I decided to take quarterbacks that have won 2 or more Superbowls from 2000 onward. These quarterbacks are Tom Brady, Peyton and Eli Manning, and Ben Roethlisberger. These are only quarterbacks who have won 2 Superbowls since 2000. I took these successful quarterbacks and added in unsuccessful quarterbacks like Matt Leinhart, Brady Quinn, Vince Young, Sam Bradford, and Jamarcus Russell. In my opinion, these were the “can’t miss” quarterbacks when they were originally signed. Again, this is just personal preference, and these quarterbacks are the ones that I chose. Matt Leinhart had accolades out of college and was a legit franchise quarterback when drafted. Brady Quinn had all the physical features of a quarterback and a solid college career. Jamarcus Russell was the ultimate physically gifted quarterback when coming out of college.

Lastly, I chose Brock Purdy, Jalen Hurts, Joe Burrow, and Justin Herbert as other quarterbacks that are in the first 4 years of their careers. I chose not to do Patrick Mahomes (6th year) or Josh Allen (5th year) because of this particular filter.

I wanted to understand what the important stats are for winning football games, because this is the best attribute to understanding if you have a franchise quarterback who can carry the team to victories. I decided to utilize my data science experience for this data collection process to determine what are important stats for game-winning quarterbacks. Finally, I can compare Kenny Pickett to the best and worst quarterbacks in the dataset.

Statistically Significant stats to Winning Football Games:

What are significant attributes to a quarterback winning the game or not? Using a confidence interval of 99, here are the stats that show as extremely important to winning football games:

YDS : 0.002997273283384088
Passing Average : 6.301510058519433e-08
TD : 1.0783590009070264e-06
INT : 8.293465016191021e-11
SCK : 0.006626087156617834
SCKY : 0.003781831597707964
RATE : 8.580941309584172e-13
ATT.1 : 5.392816749993909e-06
Rushing Average : 0.008061007489830765
LOST : 0.007764692121825656
Total Turnovers : 7.92566627028684e-12
Turnovers + Sacks : 2.999139977304595e-07
% covering responsibilities : 0.005901444990044792

Predictive Power Score stats to Winning Football Games:

Predictive Power Score or PPS is a score that is asymmetric, agnostic, and helps identify linear or non-linear relationships between two columns of a particular dataset. The value spectrum of PPS lies between 0 (no predictive power) and 1 (highest predictive power).

Through PPS we can determine how useful a variable would be in predicting the values of another variable in a given dataset. Since it normalizes the data, it is very reliable. Predictive Power Score provides these top features for winning football games as a quarterback:

Generally, a PPS score near 1 (e.g 0.8) is considered as good and this tells us that a given column A is very likely to predict the values of column B.

Correlation stats to Winning Football Games:

The most highly correlated stats for winning football games are highlighted below. I personally chose to use 20% correlation in both directions as my cutoff for significance. Passing Average yds per attempt is positively correlated, Interception is negatively correlated, QB rating is positively correlated, Total Turnovers + Sacks is highly negatively correlated, and Total Turnovers is negatively correlated. However, we will remove the Turnovers + Sacks because it is highly correlated with the Total Turnover column, which we will want to remove.

Opinion on Kenny Pickett:

Anyway, enough stats! Let’s look at how I interpreted these stats to decide whether or not Kenny Pickett is a good quarterback. How does he look against the best and worst quarterbacks in this era?

Game Management Score with the Average game management score is little under 30. GMS is calculated based on your Responsibility minus your turnover+ sack value.

Kenny Pickett was the fifth best quarterback out of these results for first year quarterback game management score. Responsibility is your passing and rushing attempts per game.

Passing Rating was the highest correlated feature

Kenny Pickett was the sixth best quarterback out of these quarterbacks in his first year as a pro.

For Total Turnovers, Kenny finished in the top 4 with the least amount. Does Jamarcus as a top quarterback for turnovers make sense? Yes, it does because he barely played in his first season giving him little chance to turn the ball over. He played two games and had 6 turnovers and 12 turnovers + sacks. I view sacks as a team idea, but quarterbacks are able to avoid some of these by throwing the ball away or scrambling out of the pocket. Quarterbacks can have a hot route to get the ball out of their hands, so it is as much of a team stat as it is a quarterback accountability stat.

Lastly, the final stat to review that was of high importance was the interception stat. Kenny came in at fifth again. Kenny consistently improved in his first year. Kenny came into his own by the end of the season. Despite the offensive play calls or other complaints, the Steelers fans believed it was a job well done by Kenny.

I show this final visualization because I want you the reader to see that Kenny Pickett had to have a good game in order for the Steelers to win.

In their first year, no quarterback has won multiple games with a passing rating under ~80 except for Roethlisberger. Kenny was not able to rely on his defense like Ben was able to in his rookie season. Did Ben have some help growing into the Hall of Fame quarterback he is today? Absolutely! Ben had it much easier comparatively speaking. Roethlisberger performed worse than Pickett did in his first season when we compare their stats for winning games and don’t include outside factors like defensive help.

The Big Opinion

As a Steelers fan, Kenny may not have been the best rookie quarterback this season. However, overall, he is clearly showing potential to be a franchise quarterback. If he continues to grow into his quarterback role, it will only make the team better.

The only part of Kenny Pickett that concerns me is his similarity to Matt Leinhart’s first NFL season. Both Leinhart and Pickett were proven college quarterbacks and had similar first seasons. The main difference being that Matt did not take care of the ball as well as Pickett in his first year. Pickett really had a few bad games at the beginning of his season and then started to take care of the football at a very impressive rate. Pickett turned the ball over one time in 260 attempts for his last nine games! Leinhart had 9 turnovers in 276 attempts for his last eight games! He did not play in the final game. Historically, Pickett improved more than any other quarterback in this dataset. In Roethlisberger’s rookie season, he had 7 turnovers in 202 attempts in his final eight games.

Machine Learning Model on Year 1 Quarterbacks on Hall of Fame Probability

I decided to take my dataset that I have created for Quarterbacks in their first 4 years and provide a dependent variable of whether these quarterbacks will be hall of famers or not. My hall of fame quarterbacks are Ben, Tom, Peyton, and Eli.

My non- hall of fame quarterbacks are Jamarcus, Vince, Sam, Matt, and Brady.

I took their average game management score, interception per game average, Passing YDs average, Total Turnover average, and Turnovers + Sacks average to give a probability of these young quarterbacks getting into the hall of fame utilizing just the data from their first NFL season. I used a basic model with a Logistic Regression (67%) for easy explanation. The results are below:

Trevor Lawrence has the best chance of being a hall of fame quarterback.

In the visual above, it shows the quarterbacks’ average stats in their first year and their hall of fame probability. In this machine learning model, it is highly likely that Kenny Pickett will not become a hall of fame quarterback. The model gives him a 47% chance of becoming a hall of famer.

Kenny showed solid improvement and will be, at the very least, a franchise quarterback. The Steelers have drafted a solid quarterback and, if he shows the same care for the football as he did in his last nine games, it will be a great 15 years and potentially a hall of fame career. Reach out in the comments if you want more Steelers content or have ideas for other analytic content! BCDC is a UFC website, but I enjoy doing analytics on any sport. If you want other quarterbacks included in this list and another review, then please again reach out in the comments!

Share this…
Share on facebook
Share on twitter

Leave a Reply

Your email address will not be published. Required fields are marked *