Don't Trust The Hype: Analyzing Hype And Performance
As fans of professional football, we are good for at least two things: money and hype. Money is what the League cares about most. It's also what the team owners, players, coaching staff, TV networks, advertisers, fantasy sports sites, and everyone else care about most. After money, we fans make another important contribution: hype. In a general sense, hype encompasses our various means of support for our teams and the media's analysis of those teams (after all, media members are glorified fans anyway).
But does hype actually correlate to anything meaningful, like wins or individual performance? Using the most hyped position in sports, quarterback, let's examine the significance of hype.
First, we have to establish how hype can be measured. Because there are no advanced metrics for hype, we'll create one; let's call it HAQ (hype about quarterback). Using a composite of several preseason power and fantasy rankings, we should arrive at something that approximates how much hype a quarterback has gotten before the season.
Since rankings represent higher value with lower numbers (i.e. #1 is the best), the lower a quarterback's HAQ, the more he is hyped. To adjust for the most powerful hype vectors (ESPN), rankings from top media outlets were weighted more than miscellaneous rankings. Obviously, this ad hoc metric paints in very broad strokes, but it should tell us something about relative hype levels. Here are the top 10 QBs by HAQ for 2015:
The most telling comparison will demonstrate how, if at all, HAQ correlates with wins and some other individual performance metric. The most comprehensive individual performance metric for quarterback's is Football Outsiders' Defense-adjusted Yards Above Replacement (DYAR). This gives the value of the quarterback's performance compared to replacement level, adjusted for situation and opponent, and then translates it into yardage. Let's also include winning percentage for 2015. Here's the table and accompanying plots:
|Aaron Rodgers||1.4 (HAQ)||385 (DYAR)||100% (Winning Percentage)|
What does the data tell us?
HAQ does not correlate with DYAR. As the first plot shows, there is not a strong correlation between hype as measured by HAQ and QB performance as measured by DYAR. Here's an easy way to visualize this lack of correlation: try to imagine a trend line on plot #1. Unfortunately, this is close to impossible because there is no trend.
Contrast plot #1 with plot #2, which shows us that DYAR and winning percentage are strongly correlated (with the obvious exception of Peyton Manning's data point). The trend here is clear, and this correlation suggests that DYAR is not a bogus statistic. Therefore, the non-relationship between HAQ and DYAR indicates that hype doesn't mean anything (not necessarily a surprise).
There are a couple of specific points on these plots that merit extra attention. Luck and Wilson were adored by the hype machine (HAQs of 1.8 and 4), yet their seasons have been rough, according to DYAR. Other stats confirm their struggles: QBR's of 30.9 and 59.8, winning percentages of 33%. They also have fallen far in the most current QB rankings, which demonstrates just how fickle hype is to begin with.
As the final nails in hype's coffin, note the performances of Dalton and Palmer. Their HAQs are 22 and 18, respectively. Despite this marked lack of hype, their DYARs are at the top of the League (556 and 553, respectively). Just as Luck and Wilson were drastically overhyped for 2015, Dalton and Palmer were under-hyped.
The bottom line: enjoy the hype, but don't trust it.