Advanced Analytics Reveal NFL Discrepancies in Team Strength and Awards

Wharton Moneyball · December 24, 2025 · Listen to Original Episode →

Original Title:

TL;DR

Advanced analytics, particularly DVOA, identify Seattle and the Rams as historically elite teams, suggesting a significant discrepancy with other advanced metrics that place them lower, implying a potential overestimation of their Super Bowl chances by conventional models.
The NFL's increasing reliance on advanced analytics for awards voting is shifting narrative-driven selections towards data-backed performance, potentially altering historical MVP and All-Pro outcomes by emphasizing statistical superiority over perceived momentum.
Generative AI models, when prompted with specific parameters like Bayesian inference and betting market data, can produce sophisticated probabilistic forecasts for sports outcomes, demonstrating their utility in quantifying uncertainty and simulating future events.
The collection and normalization of historical play-by-play data, extending back to 1977, enable deeper comparative analysis of team performance across eras, revealing historically strong teams that missed the playoffs and challenging conventional historical rankings.
The NFL's extreme schedule variations this season, coupled with teams exhibiting unusual performance in one-score games, contribute to a high degree of statistical variance, making win-loss records less predictive of underlying team strength.
The growing influence of analytics in awards voting suggests a future where narrative may hold less sway, potentially preventing players with statistically weaker performances from winning major awards despite team success.
The development of advanced charting categories, such as performance against specific coverages or run concepts, allows for granular analysis of team strengths and weaknesses, moving beyond aggregated statistics to understand situational effectiveness.

Deep Dive

Advanced analytics, particularly DVOA, reveal significant discrepancies between perceived team strength and actual NFL season performance, suggesting that conventional win-loss records are insufficient for accurately assessing contender status. These metrics challenge established narratives and highlight how data-driven insights can redefine our understanding of team capabilities, playoff probabilities, and even award voting.

The current NFL season is marked by unusual performance deviations, where teams like Carolina, New England, and Chicago appear to be overperforming their underlying statistical profiles, while Detroit, Kansas City, and Indianapolis are playing better than their records indicate. Aaron Schatz's DVOA ratings, which incorporate special teams and down-weight turnovers due to their lower predictive power, position Seattle and the Rams as historically elite teams, suggesting they have a nearly 50% combined chance to win the Super Bowl. This contrasts sharply with other advanced analytics models, such as EPA-based systems, which place their combined Super Bowl probability significantly lower, around 27%. This divergence underscores the impact of methodological choices in analytics, demonstrating how different approaches can lead to vastly different conclusions about team strength and championship potential. The analysis also reveals extreme schedule variations due to conference and divisional strength, with the NFC West identified as the strongest division and the NFC South as the weakest.

The growing influence of analytics is reshaping how awards are determined, moving beyond subjective evaluation to incorporate quantitative measures. While narrative still plays a role, the increasing availability and understanding of advanced statistics among voters suggests a future where data may more heavily dictate outcomes, potentially diminishing the impact of anecdotal evidence or "gut feelings." Furthermore, the expansion of historical play-by-play data, now available back to 1978, allows for deeper historical comparisons and a more robust understanding of long-term trends in team performance and player evaluation. The application of generative AI to sports forecasting, as demonstrated with NBA win totals, college football playoff probabilities, and NFL Super Bowl odds, showcases a new frontier in predictive analytics, leveraging Bayesian reasoning and simulation-based models to quantify uncertainty and generate probabilistic forecasts. This technological advancement promises to further democratize sophisticated analytical tools, enabling more nuanced predictions and a deeper comprehension of the inherent variability in sports.

Action Items

Audit DVOA calculation: Incorporate special teams data for 10 historical seasons to assess impact on team strength assessment.
Create play-scoring normalization: Define consistent success point thresholds for down-and-distance across all historical play-by-play data (1978-present).
Analyze awards voting bias: Track 5-10 historical MVP/All-Pro voting patterns to identify correlation between narrative and statistical performance.
Develop quarterback historical trust metric: Quantify long-term QB performance beyond current season for 3-5 high-variance teams.
Measure conference strength discrepancy: Calculate average DVOA for top 5 teams in each conference for 3 past seasons.

Key Quotes

"A lot of teams where their record doesn't really match what the underlying statistics suggest about how good they are right? Like Carolina, New England, Chicago don't seem quite as good as their records; Detroit, Kansas City, Indianapolis have played better than their records."

Aaron Schatz explains that a team's win-loss record can be misleading, and underlying statistical performance, as measured by advanced metrics, often provides a more accurate picture of a team's true strength. Schatz highlights specific teams whose records do not align with their statistical performance, indicating a potential disconnect between perception and reality.

"The reason for that is almost everybody else who does advanced analytics uses some form of EPA, expected points added, with some sort of adjustments, and I don't. I use my own stat which is DVOA, and because of some of the ways that DVOA works and EPA works, Seattle and the Rams come out a lot higher in DVOA than they do in EPA."

Aaron Schatz differentiates his DVOA (Defense-adjusted Value Over Average) metric from EPA (Expected Points Added), explaining that this methodological difference leads to significantly higher valuations for teams like the Seattle Seahawks and Los Angeles Rams in his analysis compared to other advanced analytics models. Schatz notes that these discrepancies arise from how each metric accounts for factors like turnovers and specific play outcomes.

"My numbers would say yes, it's a surprise. It's surprising, and I don't know whether maybe I have them a little too high and no one else would have this, but yeah, I mean, everybody has right now has Seattle and the Rams at their top two teams."

Aaron Schatz expresses surprise at the market's consensus, which aligns with his own DVOA ratings, placing the Seattle Seahawks and Los Angeles Rams as the top two teams. Schatz acknowledges that his own assessment might be higher than other analysts, but notes that the general agreement on these two teams' prominence is notable.

"The reason to have them as the favorite to come out of the AFC is Josh Allen. It's the idea that you have more historical trust in Josh Allen than you do in Drake Maye or Bo Nix or Trevor Lawrence."

Aaron Schatz suggests that betting markets favor the Buffalo Bills to win the AFC not due to their current season-long performance metrics, but rather because of the perceived historical reliability and trust placed in their quarterback, Josh Allen, over other emerging quarterbacks. Schatz implies that this "historical trust" influences betting odds more than current statistical performance.

"The search for historical play by play data and the upcoming edition of the 1977 season. So is it fair for me to say that what you and I, I completely agree with this by the way, that you've built most of your analytical career for the NFL on having this play by play data as opposed to just aggregated statistics?"

Eric Bradlow confirms with Aaron Schatz that Schatz's extensive NFL analytics career is fundamentally built upon detailed play-by-play data, rather than simpler aggregated statistics. Schatz agrees, highlighting the importance of this granular data, which he has been instrumental in collecting and making accessible, extending back to the 1970s.

"I was the first analytics person to be on the panel starting in 2021, but they've done a lot of changing of the panel over the last few years. So now you have Mina Kimes and you have Sam Monson, who used to be with PFF, and you have Doug Farrar, who's more of a film guy but understands analytics."

Aaron Schatz describes his experience as an early analytics advocate on the Associated Press awards voting panel, noting the panel's evolution to include more individuals with analytical backgrounds or an understanding of analytics. Schatz points to specific individuals now on the panel who bring a data-informed perspective, even if their primary focus is film study.

Resources

External Resources

Books

"Moneyball: The Art of Winning an Unfair Game" by Michael Lewis - Mentioned as a foundational text for analytical approaches to sports.

Articles & Papers

"DVOA" (Football Outsiders) - Discussed as an advanced statistic for measuring team efficiency.
"EPA" (Expected Points Added) - Referenced as a common advanced statistic for measuring team efficiency.
"FPI" (Football Power Index) - Mentioned as an example of an advanced analytics metric used by ESPN.

People

Aaron Schatz - Chief Analytics Officer at FTN Fantasy and founder of Football Outsiders, discussed for his work on DVOA and analytics in sports.
Eric Bradlow - Professor of Marketing, Statistics, and Data Science at the Wharton School and host of Wharton Moneyball.
Kate Massie - Mentioned as a regular contributor to Wharton Moneyball.
Shane Jensen - Mentioned as a regular contributor to Wharton Moneyball.
Audi Weiner - Mentioned as a regular contributor to Wharton Moneyball.
Michael Jordan - Referenced in a discussion about historical basketball players.
LeBron James - Referenced in a discussion about historical basketball players.
Teddy Roosevelt - Mentioned in a historical comparison of sports.
Patrick Mahomes - Referenced in a discussion about current NFL quarterbacks.
Tom Brady - Referenced in a discussion about current NFL quarterbacks.
Sam Darnold - Mentioned as the current quarterback for the Seahawks.
Josh Allen - Referenced as a quarterback for the Bills, discussed in relation to betting odds and MVP consideration.
Drake Maye - Mentioned as a potential NFL quarterback.
Bo Nix - Mentioned as a potential NFL quarterback.
Trevor Lawrence - Mentioned as a potential NFL quarterback.
Jalen Hurts - Mentioned as a potential NFL quarterback.
Cam Newton - Referenced in a discussion about past MVP voting.
Jeremy Snyder - Mentioned as a collaborator in collecting historical play-by-play data.
Mina Kimes - Mentioned as an analytics-oriented member of the Associated Press voting panel.
Sam Monson - Mentioned as a former PFF analyst and member of the Associated Press voting panel.
Doug Farrar - Mentioned as a film analyst and member of the Associated Press voting panel.
Dan Orlovsky - Referenced as a former quarterback and ESPN analyst who uses advanced statistics.
Lamar Jackson - Discussed in relation to his 2023 MVP vote.
Jannik Sinner - Mentioned as a competitor to Carlos Alcaraz in tennis.
Carlos Alcaraz - Mentioned as a top professional tennis player who recently changed coaches.
Juan Carlos Ferrero - Former coach of Carlos Alcaraz.
Paul Annacone - Mentioned as a potential future guest on Wharton Moneyball to discuss tennis.

Organizations & Institutions

FTN Fantasy - Mentioned as the employer of Aaron Schatz and a provider of advanced sports analytics.
Football Outsiders - Mentioned as the organization founded by Aaron Schatz, known for DVOA.
ESPN - Referenced as a media outlet that provides advanced statistics and employs analysts who use them.
Associated Press (AP) - Mentioned as the organization responsible for voting on NFL awards and All-Pro teams.
Vegas - Referenced as a source for betting odds and market probabilities in sports.
Pro Football Reference - Mentioned as a website that now hosts historical play-by-play data.
Bet365 - Referenced as a source for betting odds used in probability calculations.
DraftKings - Referenced as a source for betting odds used in probability calculations.
Wharton School - Mentioned as the academic institution where Eric Bradlow is a professor.
Wharton Moneyball - The name of the podcast.
Wharton Podcast Network - The network on which the podcast is distributed.
Oklahoma City Thunder - Mentioned in the context of generative AI predicting win totals.
Golden State Warriors - Referenced for holding the record for most regular season wins.
New England Patriots - Mentioned as a historically good team.
Washington Redskins - Mentioned as historically good teams.
Chicago Bears - Mentioned in relation to team performance versus record.
Detroit Lions - Mentioned in relation to team performance versus record.
Kansas City Chiefs - Mentioned in relation to team performance versus record.
Indianapolis Colts - Mentioned in relation to team performance versus record.
Carolina Panthers - Mentioned in relation to team performance versus record.
Seattle Seahawks - Discussed extensively in relation to advanced analytics and team strength.
Los Angeles Rams - Discussed extensively in relation to advanced analytics and team strength.
Jacksonville Jaguars - Mentioned in relation to team strength and conference performance.
Houston Texans - Mentioned in relation to team strength and conference performance.
Buffalo Bills - Mentioned in relation to betting odds and AFC contenders.
Denver Broncos - Mentioned in relation to betting odds.
New York Jets - Mentioned in relation to betting odds.
Philadelphia Eagles - Mentioned in relation to betting odds.
San Francisco 49ers - Mentioned in relation to team strength and conference performance.
Tampa Bay Buccaneers - Discussed in relation to their division performance and control of their destiny.
Atlanta Falcons - Mentioned in relation to the Buccaneers' recent losses.
New Orleans Saints - Mentioned in relation to the Buccaneers' recent losses.
Miami Dolphins - Mentioned in relation to the Buccaneers' upcoming game.
Indiana University - Mentioned in the context of generative AI predicting college football playoff teams.
Ohio State University (OSU) - Mentioned in the context of generative AI predicting college football playoff teams.
University of Georgia - Mentioned in the context of generative AI predicting college football playoff teams.
University of Oregon - Mentioned in the context of generative AI predicting college football playoff teams.
Texas Tech University - Mentioned in the context of generative AI predicting college football playoff teams.

Tools & Software

Generative AI - Discussed as a tool for forecasting and analysis, specifically used by Eric Bradlow.
ChatGPT 5.2 - Mentioned as a specific generative AI model used for analysis.

Websites & Online Resources

acast.com/privacy - Linked for privacy information related to podcast hosting.
ftnfantasy.com - Mentioned as the website for FTN Fantasy.

Other Resources

DVOA (Defense-adjusted Value Over Average) - An advanced statistic used to measure team efficiency.
Play-by-play analytics - Discussed as a method for analyzing game events.
Bayesian reasoning - Mentioned as a statistical approach used in forecasting models.
Simulation-based analytics - Referenced as a method for forecasting outcomes.
Expected Points Added (EPA) - A metric used to evaluate the efficiency of plays.
NFC West - Discussed as the strongest division in the NFL.
NFC South - Discussed as the weakest division in the NFL.
AFC South - Mentioned as a surprisingly strong division.
Play-by-play data - Discussed as a crucial resource for historical NFL analysis.
Beta binomial predictive model - A statistical model used for forecasting.
Moneyline odds - Used to convert betting odds into implied probabilities.
Betting vig - The commission or fee charged by bookmakers, which needs to be accounted for in probability calculations.
Advanced analytics in awards voting - Discussed as a growing influence on how players are evaluated for awards.
Charting data - Detailed statistical data collected on specific aspects of play.
Man coverage vs. Zone coverage - Defensive schemes analyzed using charting data.
Run concepts - Offensive strategies analyzed using charting data.
Basketball charting - Mentioned as a new area of analytics being developed by FTN.