It is the week of Super Bowl Sunday and I live with a Patriots fan. For the last two weeks all serious conversation in our house has revolved around some aspect of the upcoming game. Unless you have been living under a rock (or inside a book), you can probably guess that most of our conversations center around why a set of footballs used by the Patriots during the AFC Championship game were found to be under the minimum psi level specified by the NFL. Were the Patriots cheating by manually deflating footballs? Or is there a “natural” explanation for the deflation?
Heaving a big psi. (Photo by Flickr user Frankieleon, https://www.flickr.com/photos/armydre2008/)
The interesting question from an STS perspective, and the hinge which cheating allegations revolve around, is whether or not the atmospheric conditions at the AFC championship game could have caused a football to deflate what the NFL has called “a significant amount.” The question is a thorny one because it is entirely unclear who counts as an expert on football deflation, where one might turn to find an expert opinion, or even what criteria might be appropriate in determining who is, or is not, an expert on football deflation. Worse, how might one find a deflation expert who does not have a rooting interest for or against the Patriots at this late date? In short, who may enunciate the truths of football deflation?
Patriots head coach, and noted gridiron alchemist, Bill Belichick was the first to turn to science for an explanation. Like a modern day Boyle, he held a press conference in which he detailed an experiment conducted at the Patriots facility which he claimed demonstrated that natural conditions caused “significant” football deflation at the AFC Championship game. His explanation was detailed and involved a special method of preparing the football for play (that is, getting the correct feel for the quarterback) that can change the psi level without manual deflation.
Belichick’s experiment caused an immediate reaction. Science celebrities Bill Nye, the Science Guy (a well-known Seahawks fan and mechanical engineer) and Neil deGrasse Tyson (a theoretical physicist and perhaps a Giants, or Jets, fan) both weighed in on Belichick’s experiment, the former taking to morning television to rebut Belichick and the latter to twitter to voice his doubts. The next day support for Belichick’s experiment appeared in the form of three Boston area professors (at least one of whom is a Bills fan). Not one to miss a bandwagon, the NFL is currently consulting with the physics department at Columbia University (could they be Giants or Jets fans?) about the role of atmospheric conditions on football deflation.
While many mined theory for an explanation, others tried to replicate Belichick’s experiment. The experimenters at HeadSmart Labs claimed to replicate Belichick’s claims about atmospheric conditions causing deflation. Meanwhile, a series of posts by Chad Orzel on his football deflation experiments are summed up here. The HeadSmart Labs experiment included inflating the football in a 75 degree room, soaking the footballs in water, which they claimed made them expand, then moving the footballs to a 50 degree room prior to measuring psi. Doing this, they claim psi was lowered by .9 to 1.8 psi per football, which is in line with the deflation claimed at the AFC Championship game. Orzel, for his part, notes that the series of experiments he conducted, which did not show significant deflation, were performed using a pressure sensor that measures absolute pressure, not gauge pressure. This is a critical difference when atmospheric conditions form part of the argument. No word on the HeadSmart Labs gauge, but a safe guess is that their gauge did not use absolute pressure.
Throughout this crisis of knowledge, Belichick has proved to be a savvy experimenter with a feel for the difficulties of the experimental method and the role of contingency in knowledge production (theoreticians beware!). Take note of this quote from his press conference on the localization of experimental technique and material, and the difficulty of replication:
When you measure a football, there are a number of different issues that come up. Number one, gauges. There are multiple types of gauges. The accuracy of one gauge relative to another, there’s variance there. We’re talking about air pressure. There’s some variance there. Clearly all footballs are different. So, footballs that come out of a similar pack, a similar box, a similar preparation, each football has its own unique, individual characteristics because it’s not a man-made piece of equipment. It’s an animal skin, it’s a bladder, it’s stitching, it’s laces. Each one has its own unique characteristics. Whatever you do with that football, if you do the same thing with another one, it might be close, but there’s a variance between each individual football.
If this is the state of the atmospheric conditions argument, the argument for human intervention has been moving forward as well. As of this writing the NFL has reviewed video footage showing the movement of the footballs in question prior to kickoff. The investigation has now centered on the 90 seconds a Patriots employee spent in a locked bathroom with the footballs later found to be deflated. Here, it has been argued, in a space free of video cameras the employee had time, opportunity, and motive to alter the footballs. Where the physicist has been the default expert for the atmospheric argument, the detective has emerged as the default expert on human motivation.
Not wanting to be left out, over the last few days, data scientists have weighed in on a statistical analysis claiming the Patriots fumble rate over the last few seasons (very low by NFL standards) cannot be explained by random fluctuation. Hence, as the dim logic of data science tells us, there must be foul play!
American football, as the saying goes, is a game of inches. Increasingly, it is also a game of expert witness.
1 Comment
I hastened to check what the data scientists were saying: it looks like there’s only one data scientist making that claim and the other data scientists have all banded together to refute him; they also indulge in some old-fashioned name-calling, labeling him a “tout.”
1 Trackback