Although there still exists an aura of mystery behind Chip Kelly’s sports science initiative in Philadelphia, we do know a few things. We know about the personalized smoothies. We know about the extended periods of rest. We know about the hockey puck-like sensors on player’s backs. We know about Sports Science Coordinator Shaun Huls’ military background. But that’s really about it. Huls doesn’t speak to the media and Kelly likes to protect the secretive nature of his sports science program. As Eagles players return to work after the bye and head into the home stretch of a season that now contains talk of playoffs, have Kelly and Huls succeeded in cultivating healthier, more efficient football players? Does it translate to team success?
They are tough questions to answer. There is of course anecdotal evidence of success. Against Washington, Nick Foles connected with Brent Celek on a tight end screen and Celek proceeded to sprint 43 yards with very un-Celek-like speed, slicing through Redskins defenders with Desean Jackson aplomb (where the hell has that been, Brent?). Can this be attributed to sports science? And this season LeSean McCoy established new human records for fastest recoveries from broken ankles and ripped hamstrings (sports science works!). Granted, these examples are (slightly) circumstantial, but there may be some factual evidence as well.
In order to determine (as best we can) the relative health of a football team, we need to look at the weekly injury reports. Granted, this isn’t ideal; each NFL team and coach most likely submits injuries in a manner that supports their own strategies. In the short-term, injury reports are inherently hazy. But I think in the long term, they hold at least some truth. Let’s look at the IR first. Through eleven weeks, the average NFL team has six players on Injured Reserve. In this regard, the Eagles are… average. Jeremy Maclin, Phillip Hunt, Jason Phillips, Curtis Marsh, Arrelious Benn, and Joe Kruger are all on the IR. Kansas City and Detroit lead the NFL with the least number of players on IR, amazingly with just one each (Tampa Bay has the most, with 12). But many of the Eagles’ severe injuries occurred before the season started. Only two players (Benn, Kruger) were placed on the IR after meaningful games began. Nine other teams have placed the same number or fewer on the IR, including poor performing Jacksonville, Tennessee, and Baltimore.
What about the rest of the injury report? What about all those probables, questionables, and doubtfuls? Again, it’s tough to quantify because each team can report injuries differently. For example, through eleven weeks, the Houston Texans have 140 reported instances of listing players as "Probable"; the New York Jets are right behind them with 136. In contrast, the Baltimore Ravens have just 17 instances, the Eagles have 43, and the New England Patriots have 35 (the NFL average is 50). And to give you an idea about how NFL coaches "play" with the injury report, we can look at how many players are listed as "Questionable" (players with a 50% chance of playing). The average NFL team has 22 instances of players listed as questionable. The Pittsburgh Steelers have the least (6), the Eagles have the 9th least (15), and the Patriots have by far the most (64). Bill Belichick keeps you guessing.
But in order to get a better picture of overall team health, I created a Health Score from the information in the weekly injury reports. I multiplied the number of players in each injury report category (unduplicated) by the likelihood that they would not play as defined by the report (players listed as probable technically have a 25% chance of not playing, but I arbitrarily used 10%, since most times those players do play). I subtracted those results from 61 (the 53 man roster plus 8 team practice squad) then divided that by 61 to come up with each team’s rate of health. The formula looks like this:
Team Health Score = (61 – (# Players Probable x 0.10) – (# Players Questionable x 0.50) – (# Players Doubtful x 0.75) - # Players Out/IR) / 61
Using these criteria, the average NFL team through eleven weeks is about 68% healthy, with the most healthy team being the Washington Redskins (81.7%) and the least healthy are the Cincinnati Bengals and Seattle Seahawks (58.5%). The sports science-driven Eagles rank sixth (73%).
|2||Kansas City Chiefs||78.61%|
|3||San Francisco 49ers||75.98%|
|13||San Diego Chargers||69.67%|
|15||Tampa Bay Buccaneers||69.02%|
|16||New York Jets||68.85%|
|21||Saint Louis Rams||65.66%|
|24||Green Bay Packers||63.61%|
|25||New York Giants||63.11%|
|28||New England Patriots||59.84%|
|30||New Orleans Saints||58.69%|
So the Philadelphia Eagles are among the best in the NFL in players on IR and Health Score, but do these accomplishments translate into success? Again, tough to say. There is a weak, but still significant, relationship between the number of players a team has placed on the IR and win percentage (r-squared = -0.19). But sometimes win percentages just show that teams are simply really good (or really bad). Do injuries impact whether bad teams improve or good teams regress? Comparing players on IR to a team’s DVOA difference from last year (see Football Outsiders), there is a stronger relationship (r-squared = -0.29). So it’s possible then, that among other factors, 29% of a team’s improvement from one year to the next can be attributed to the lack of players on IR. The fact that the Eagles have only placed two players on IR since the start of the season could be one of the (many) reasons why the Eagles have two more wins than last season with five games to play.
Health Score correlations are interesting. There is virtually no relationship (r-squared = .003) between a team’s Health Score and win percentage. There are too many factors (namely, talent and coaching) that contribute to numbers of wins. Good teams have good enough depth and coaching that allow them to overcome injuries (Seattle, New Orleans) and bad teams with high Health Scores are still bad (Washington). Even when removing outliers, the relationship improves but is still very, very small (r-squared = 0.016). In comparison, the relationship between Health Scores and DVOA improvement is a little stronger, but still weak (r-squared = 0.029). Only when removing outliers does the relationship improve (r-squared = 0.16). However, the effect of team Health Scores on DVOA Improvement is ten times stronger than the effect of team Health Scores on win percentage.
Ultimately, it appears a team’s relative health doesn’t have a huge impact on the number of games a team wins in a particular season (11 weeks into a season, at least); depth of talent and coaching impact wins and losses more. But team health can have more of an impact on how a team improves from one season to the next. From the Philadelphia Eagles’ perspective, the contributions made by Sports Science Coordinator Shaun Huls seem to be working. It doesn't hurt that Michael Vick replacement Nick Foles is breaking more NFL records than Shady McCoy breaks ankles. The Eagles possess one of the healthier teams in the NFL and are also one of the most improved. Hopefully they will continue to be. If they are, it will be exciting to see how the Eagles’ healthier, more efficient football players crush the Cowboys in Dallas Week 17.
*Author Note (11/26/2013): I used IR dates provided by CBS Sports, which were wrong for Joe Kruger and Arrelious Benn. They were both placed on IR before the season started. Without validating the dates for all NFL teams, it's possible the Eagles are the only team in the NFL that hasn't placed a player on IR after the season started. Thanks Zach Berman: