Using statistical analysis, we analyzed NFL team and player data collected from multiple sources to investigate 4 interesting questions about performance and play factors. Please see the report and presentation files for a more detailed explanation of the project and results.
There is no strategic reason why home field advantage should exists for NFL games where in sports like Baseball, which are turn based, the home team gets the final chance to score runs. We investigate if teams win more games at home and verify a statistical significance to a home field advantage.
Conclusion: For most teams, there is a significant home field advantage when we looked at 5 year intervals of data over 41 years.
Athlete pay is generally a combination of their skill and the revenue they bring in for fans coming to see them perform. We investigate, specifically, if there is a statistically significant relationship between player compensation and their performance.
Conclusion: We found a strong case against there being a significant impact of ompensation on player performance.
Throughout the last few decades, the NFL has updated their rules and some changes were designed to have major impacts on the game. We investigate if 2 key rule changes have had any significant impact on gameplay. Most specifically, we investigated the encroachment rule (a defensive penalty for "off-sides") implemented in 1995 and the rule pertaining to roughing the passer, unnecessary roughness and horse-collar tackling which was implemented in 2005.
Conclusion: We found a significant increase in penalty yards per game following both rule changes but the most significant rule was the encroachment rule.
Using multivariate linear regression techniques, we filtered through all team features to fund the most appropriate model for predicting win percentages of NFL teams. This was an attempt to test the validity of a linear relationship between team data and win percentages.