Falcons vs. Eagles 2015: Let's Get Simulated

Back when Chip Kelly was still a head coaching candidate, I read as much as I could about the man.  The most revealing document I found was one you most likely read as well, Kelly’s white paper, "Efficient Use of Practice Time".  In it, he addressed more than just strategies for maximizing practice; he explained his philosophy, the type of player he likes, the spread offense and it’s tactical advantages.  But what caught my attention most was this little nugget, only 5,826 words deep:

We base the success formula for offense on the total number of plays. Take those plays minus the dropped balls, offensive penalties, and negative yardage plays, and divide by the total number of plays. If the answer is 80 percent or better, you win the game.

Chip didn’t say it explicitly, but what he described is a formula for offensive efficiency, and it correlates highly to winning games.  When put to the test using individual game data from 2009-2012, home teams with a higher offensive efficiency than away teams win 89% of the time; away teams with a higher offensive efficiency win 80% of the time (a difference which I have come to think illustrates home field advantage).

I have since modified Chip’s formula to create a defensive efficiency metric and used it to simulate games, pitting each team’s offense against the other’s defense, while also incorporating the impact of turnovers.  The result is less a prediction and more of a forecast that highlights the statistical efficiency conditions under which each team can win. It’s also the greatest thing since PFF converted grades to scaled ratings and removed its premium statistics for consumer use.

The Simulation

The simulation itself is run by you.  Select the number of simulation runs and adjust the turnover margin to see how game outcomes change.  Since this is the first week of the season, this particular simulation is based on last year’s stats.  As the season goes on, I expect the simulation results to be more meaningful.  In the meantime… mess around with it!

Philadelphia Offense versus Atlanta Defense

Last year, Nick Foles, Mark Sanchez, and Matt Barkley commanded the Eagles offense with 70% efficiency.  The offense ran 1,127 plays and committed 36 turnovers and 47 penalties.  Given last year’s stats alone (granted, anticipation is extremely high for this year’s offense), this is better than Atlanta’s defense which operated at 65% efficiency, on the field for 1,038 plays and securing 28 turnovers, six less than the Eagles committed.  Nevertheless, Sam Bradford has provided the team with enough reason to believe efficiency will improve.

Atlanta Offense versus Philadelphia Defense

In terms of efficiency, Matt Ryan’s offense was not too dissimilar from the Eagles’, if not slightly more efficient (70.8%).  The Falcon’s offense ran less plays (1,035) and committed less turnovers (23), but committed more penalties (49).  The Eagles’ defense saw the field more often (1,113 plays)  but, amazingly, allowed 15 less pass completions than the Falcons defense.

Point Spreads and Bayes Odds

Point spreads can tell us a bit more than how strong one team is favored over another.  They also can tell us the probability that a particular team can win.  We know from history that teams favored by 1.5 pts win 55% of the time, six point favorites win 69% of the time, and two touchdown favorites win 86.5% of the time.

As I write this, the Eagles are 3.5 point favorites, which means, based on this alone, they have a 60% chance of beating the Falcons.  When we combine this knowledge with the results of the simulation using Bayes Theorem, we can generate updated odds.

What It All Means

When the turnover margin is zero, the simulation favors the Eagles about 56% of the time.  When taking into account the point spread, this gives the Eagles a 66% chance of winning the game.

But turnovers, and the extent to which teams score as a result of them, can kill.  If the Eagles’ turnover margin is +1, then their odds increase to 81%.  When they lose the turnover margin by one, then they still have a 51% chance of winning, thanks in part to their status as favorites.

Bottom line… The Eagles got this.