NFL simulations have become a "thing", perhaps trendy, perhaps chic, a fringe exercise creeping into the NFL's mainstream. There are vocal critics and constructive critiques, but proponents and opponents alike know one thing to be true: they're not perfect. They can, however, provide interesting context. Maybe even for the Philadelphia Eagles going into Dallas.
BGN's simulations are based on a simple enough concept: efficiency. It's a model that relies on an element that Chip Kelly uses as a gauge for success: total plays. Generally and historically speaking, an efficient team can often be a successful team. But efficiency itself is not a universal truth. Teams are efficient in different ways and high efficiency alone does not a good football team make. There are just too many factors involved.
BLEEDING GREEN NATION INTERACTIVE SIMULATION
Football Outsiders' DVOA and ESPN's new "Football Power Index" (FPI) try to consolidate as many of these factors as possible into one metric in a meaningful way. DVOA is a bit better in this regard (I can't take too seriously a system that currently ranks the Eagles as the third best team in the league). But there is a something to be said for simplicity and elegance. After all, the bottom line is the score, right?
Take 538's Elo ratings. Theirs is adapted from a rating system used for chess (apt), named after the Physics professor who created it, Arpad Elo. When applied to football, it works something like this: all franchises are assigned the same base score (1500) at the point they were founded. As teams play each other, their "Elo" scores increase or decrease based on two elements: the final score of the game and the relative Elo score of the opponent. This ensures the rating of a team that blows out a lesser-rated opponent will not increase as much as the rating of a team that blows out a higher-rated opponent. At season's end, a team's rating transfers to the beginning of the next season with a slight regression towards the mean. Simple and elegant.
The highest Elo-rated Eagles team in franchise history was the 1950 championship team (Elo rating, 1759). The worst was a 1936 squad that went 1-11 (1246). More recently, the Eagles team that Andy Reid inherited had an Elo rating of 1299. His team peaked in Week 15 of the 2004 season after beating the Dallas Cowboys 12-7, improving their record to 13-1. Of course that was the season the Eagles lost to the New England Patriots in Super Bowl XXXIX.
PHILADELPHIA EAGLES: ELO RATINGS THROUGH HISTORY
The rating for Chip Kelly's Eagles climbed steadily from the start of the 2013 season (1400) and peaked last season (1621), after a(nother) 33-10 win over Dallas. Since then the Eagles' Elo rating has dropped to 1542 and the team is ranked 10th in the league. This week, the Eagles again face Dallas (1504) and, according to Elo, are a slight underdog thanks to a 65 Elo-point bump for home-field advantage. Elo gives the Eagles a 46% chance of winning.
Coincidentally, BGN's simulation favors the Cowboys and gives the Eagles a 46% win probability, but only if Vegas' line were a push. Since the Eagles are three point favorites, the pendulum swings in their favor, giving them a 54% chance of annihilating the Cowboys.
Coincidence or not, games against the Dallas Cowboys have served as spring boards for past Eagles teams. Sure hands willing, Sunday could make for another leap forward.