This website is for a project called “ego Analytics” (evenly gifted organizations). It is based on the idea that theoretically every NFL organization is provided with the same general resources, like draft capital and cap space, and how these organizations have used those resources over time have determined their relative levels of success.

The leadership of an NFL organization can be broken down into two parts: the coaching staff (Head Coach and Offensive/Defensive Coordinators) and the front office (General Manager and other Executives). Their relationships to the players and each other defines the organization.

The name is additionally a reference to how many front office decisions are ego-based, such as holding out for a new contract. While most sports analytics models focus on a bottom-up approach, ego Analytics uses a top-down method, primarily evaluating the leadership of the NFL organization instead of the performance of the players. In other words, sports analytics typically focuses solely on what happens on the field, but ego Analytics includes what happens off the field. Just as players and coaches actions are dictated by on-the-field rules, these off-the-field actions by NFL front offices are guided by rules. Understanding these rules for, and the resulting actions of, front offices can help predict what should happen on the field.

This project is designed to do just that – to generate a set of metrics for coaching staffs, front offices, and the organizations overall. Using those metrics, one can generate a Win Total prediction for each team. Then one could bet Win Total Futures using those predictions. Due to a high level of aggregation and more level playing field between the sportsbooks and bettors, Win Total Futures are higher value bets.

The results in the table below are based on 1 unit for every bet. The table includes IMP (Impact on the Model’s Performance), which represents the external conditions that affect the results of the model, along with the variables that go into IMP (Stability, Power Utilization, and Upset Luck). The range is roughly 60-100, and the long-term average should be around 82.

Menu Options

Specific Season Homepages: overview of the metrics for each NFL organization and the betting results

How It Works: overview of the project and the metrics that make up the economic model

Model Evaluation: describes various methods to evaluate the accuracy and effectiveness of the model

Organizational Scores: in-depth view of those metrics which includes the lower-level metrics upon which those higher-level metrics are generated

Organizational Graphs: visualizations of the Scores over time for each of the NFL organizations

Heat Check: a metric to estimate how likely a Head Coach or General Manager is to be fired

Gambling Performance: in-depth view of the Win Total Futures (Over/Under # of Wins) performances for each season

Miscellaneous Information: definitions, sources, and additional content using the metrics of this project

Legend for the Table Below

The green section shows the results of the model via Win Total Futures betting

The blue section shows the elements external to the model and their effects on the performance of the model

The purple section shows the return of the S&P500 from Open on the day before the start of the regular season to Close on the day after the end of the regular season

The red section shows the return of the Nasdaq Composite from Open on the day before the start of the regular season to Close on the day after the end of the regular season

The orange section shows the return of the Dow Jones Industrial Average from Open on the day before the start of the regular season to Close on the day after the end of the regular season

The yellow section shows the results of the “Top 20” bets for the model in each season

These bets are based on the 20 largest differences between ego prediction and Win Total line (see specific years)

This style of betting has its advantages and pitfalls in that they should be the highest value bets, but in bad years, the profit can be significantly lower due to risk consolidation