No one can prove one way or the other whether a member of the Patriots organization intentionally deflated the approved footballs before the January 18 AFC championship, unless they have access to body cam videos or other similar direct evidence. However, using a principle known as Occam’s razor, I can prove that it is highly unlikely that anyone did.
Occam’s razor, in a nutshell, says that in the absence of other direct evidence, among a choice of hypotheses the one that makes the fewest assumptions is the most likely. What we are up against here, incidentally, is a conspicuous dearth of hypotheses on the part of the accusers. The only one that I am aware of with any specificity among the accusers is that Mike McNally went into a bathroom and deflated all 24 balls during a 90 second window of time. If anyone wants to come up with another scenario and send it to me by private message I will consider it. (Note that I have comments turned off on all my blogs for spam reasons, so you can’t write them out here.) I will try to imagine one or two others just for balance. This is about logic and science, and debating the possibilities to reach a reasonable consensus and not conceding to agnosticism (“we’ll never know for sure”) or cynicism (“everyone is merely out for themselves”) both of which I find to be corrosive not just to a sport I have always loved, but to society generally.
My argument begins with Tom Brady’s press conference of January 22. The context is not critical to my argument, but it is somewhat important and worth going over. Up until he stood in front of what can only be described as a generally very hostile press corps, and took questions for 30 minutes (10 minutes beyond what he had originally said he would), no one anywhere had explained any of the details of the procedure for game ball preparation. When he did so, he was forthright in answering all the questions, and he appeared to me to be honest in his body language. Obviously I may have been fooled (I watched the whole thing three times), but I would wager that a majority of non-prejudiced observers would agree with me.
What he said about game ball preparation was (compiled from answers to different questions):
I have a process that I go through before every game where I go in and I pick the footballs that I want to use for the game. Our equipment guys do a great job of breaking the balls in. They have a process that they go through. When I pick those balls out, at that point to me they’re perfect. I don’t want anyone touching the balls after that. I don’t want anyone rubbing them, putting any air in them, taking any air out. To me those balls are perfect and that’s what I expect when I show up on the field. That happened obviously on Sunday night. It was the same process that I always go through. I didn’t think anything of it…. I think that there’s a process that everybody goes through breaking in footballs. It’s probably a lot like a baseball mitt when you’re a kid. I try to explain that to my friends a lot. When you use it and that’s your equipment, the football is something that I handle on every play. I want to be very familiar with the equipment that I’m using, just like my cleats, just like my helmet, just like my pads. You go through that process of breaking the balls in and getting comfortable with them. Of course I choose the balls that I want to use for the game and that’s what I expect to go out on the playing field with…. Once I approve the ball, like I said, that’s the ball that I expect out there on the field. It wasn’t even a thought, inkling of a concern of mine that they were any different. I just assumed that they were exactly the same: first half, second half…. I like them at the way that I like them, which is at 12.5. To me, that’s a perfect grip for the football. I think that particular term, deflated or inflated, whatever norm you’re using, you could probably use…. [O]nce I’m out on the field, I’m playing. I have no thought of the football at that point. I’m thinking about the defense, I’m thinking about the execution of the play and what I need to do. I’m not thinking about how the football feels. I grip the football…. I go in there and I choose however many balls are necessary for the games. Sometimes it’s 12, 16, 18, 24. This last particular game was 24. I felt them. They were perfect. I wouldn’t want anyone touching those. I would zip those things up and lock them away until I got out on the field and an opportunity to play with them. That’s what I thought I was doing…. Like I said, I don’t put any thought into the footballs after I choose them. When you’re out there playing in front of 70,000 people, like a home crowd, you don’t think about [it]. You’re just reacting to the game. I don’t certainly think about the football. I just assume it’s the same one I approved in the pregame…. We break them in in practice, certainly sometimes. Yeah, we definitely do that. It’s different from game to game. Some days one ball may feel good; the next day it may not. It depends on maybe how, I don’t know, the humidity in the air or how old the ball was. There are a lot of variables with obviously Mother Nature and the balls. Whatever feels good that day, those are the ones I would typically choose…. I grab it, I feel the lace, I feel the leather, I feel the tack on the ball. That’s really what you go for. It’s not like I ever squeeze the football. I just grip the football. I think there’s maybe a little bit of a difference of how I do that.
I apologize for the long excerpt, especially since it is repetitive, but this gives the reader a chance to see what Brady’s thinking was like in his pre-game preparations, and how the ball selection process fit into everything else he needed to do. Bear in mind also that this was early on in the controversy, and not in response to the Wells report (in fact much to the surprise of the press corps that day, the league office had not even contacted Brady at all yet).
He mentions elsewhere that the task of choosing the footballs was done about five hours before game time. According to the Wells report at approximately 3:45 the game officials examined and approved the balls Brady selected, and at 6:45 they were on the field and ready for play. (Note that the report gives additional details about what Brady’s actions in the selection process, beyond what he explained in the press conference, but which presumably came via Brady’s testimony to Wells.) There is nothing in the Wells report that implies anything about the balls being altered once they were on the field, and so therefore if they were altered after they were approved, such alteration took place in that three hour window.
It is important for the reader to have a grip (ha ha) on the events so far in order to follow what I am about to say, so I will review:
- Some time prior to 3:45 Brady selected 24 balls based on a general “feel” for the balls that included how the inflation affected the grip.
- He states that he did not want the balls to be altered in any way after that point, and that he trusted that the balls he would be given in the game were the same balls he selected.
- At approximately 3:45 the officials approved the balls.
- At 6:45 they were ready to be brought into play on the field.
At this point in my logic I now assert that one of three mutually exclusive event sequences occurred in the 3:45 – 6:45 window:
- The balls were not altered (by human intervention, anyway).
- The balls were altered (deflated), but Brady never had a chance to inspect them a second time (for example, to make sure that they weren’t deflated too much, or that something else happened to them in the process).
- The balls were deflated and Brady did have a chance to re-inspect them.
If event sequence 1 is true, we are done — neither Brady nor anyone else did anything, and he is therefore being unjustifiably punished.
Event sequence 3 is probably not true, and no one has ever mentioned in the Wells report or anywhere else about Brady re-inspecting the balls after alteration. Among other things it does not survive Occam’s razor — it is way too complicated as a scenario. (Did they sneak Brady in a back room to check them, for something that is of microscopic importance?) However, as we look in detail at event sequence 2, readers may be tempted to go back and say that he did re-inspect them and so potential scenarios need to proposed how and when he did that. Those hypothetical scenarios need to be fit into the known facts of the Wells report. In the end while we need to account for it, in absence of any of these scenarios I will categorically reject event sequence 3.
Therefore we are left with looking at event sequence 2: the balls were altered by taking some amount of air out of them but in an “open loop” fashion — Brady never re-inspected them and had to trust that whatever the process was, the process was going to result in balls that were up to his standards.
The issue that arises at this point is that whatever the overall ball deflation “program” was, which may have included poking of a needle (quantifiably or not) into the balls on game day, possibly re-measuring the pressure on game day after they were poked to make sure it was exactly the right amount, experimenting with air pressures during practice, deciding whether quantification of the process is important and if so what are the parameters, deciding on how and when the alteration should take place on game day in the context of all the other preparation that needs to go on — whatever that process/program was or is, it cannot be completely open loop. It is axiomatic that these processes are aimed towards satisfying Tom Brady and how he grips the ball. Therefore he needs to be involved in these experiments and decisions, he needs to be the final judge and approver of all the procedures, and he needs to trust that it all will work on game day.
How elaborate do his accusers think this is? Anecdotally the Patriots are known to be nothing if not thorough about every detail, so it is not implausible that they had some type of elaborate set of experiments. However, an elaborate set of tests is what does not survive Occam’s razor. Would they not want to show or determine that it was worth the effort by doing accuracy tests with Brady? Or would they just spritz out a little bit of air until Brady was happy? Or what exactly? If Brady was telling even an inkling of truth about the importance of the ball selection within the context of all his dozens of other concerns, such as knowing the types of defenses he is going to be up against on any given game day, being concerned about working with his offensive line, being concerned with his own physical preparation, meeting with the team, meeting with the press — anyone with even a small amount of knowledge of the complexities of football knows I am just scratching the surface here of what concerns all NFL players, especially the quarterback — if Brady was telling the truth, then any such experiments and tests would be a vast waste of time compared to what he claimed: that he simply selects the balls according to official protocol, and expects those balls to be the ones he gets to use. That would be a lot easier and soothing to the mind.
As I write this I can feel certain people wincing at the fact that I am going over these excruciating details (assuming they have gotten this far, and that I haven’t bored them to sleep or put them into a state of nausea), but that pain is exactly what I feel! In going over and over possibilities in my mind, I get dizzy trying to imagine what could have happened, let alone what other people think might have happened (which few have bothered to say), and that in itself is evidence that none of this survives Occam’s razor. It is just too complicated.
Of course what unfortunately happens at this point in the debate is people shrug their shoulders and (prejudicially, to be perfectly blunt) say “something happened” — they don’t know what or can’t imagine what, but “something happened.” Something did happen, of course — the balls were re-measured and re-inflated at half time, but there is a simple explanation for that anyway, which does survive Occam’s razor, as I will be elaborating on momentarily. Nevertheless, with respect to event sequence 2 the challenge is to the accusers and the burden is on them to propose how this all happened, or logic dictates that it is not likely that it ever did.
So what did happen? In a sense it is not what happened on that day per se, but what is all the history that culminated in the half time activities of the officials in that AFC championship game. The central factor is that there is a tremendous amount of general misunderstanding in the non-technical world about what the meaning of a tolerance in a specification is. Heck, there is a good deal of misunderstanding about that among well-trained engineers! In the case at hand, the specification on the inflation of a regulation NFL football is that it needs to be 13 +/- 0.5 pounds per square inch (and not “pounds” by the way — it is not “weight” that is being specified). The reason, historically, for such a wide tolerance comes from the fact that inflation gauges are pretty finicky — they all are, even to this day. For most purposes it doesn’t matter, partly because pressure changes due to the environment overwhelm any specified tolerance. You have your tires inflated more or less good enough, because it doesn’t matter too much as long as they are close. If you inflate them when it is 90 degrees, later that winter they will deflate a little bit, but how close you got the first time doesn’t really matter too much (although I recommend re-checking periodically). Air pressure gauges are at the low end of the spectrum of precision quality among all the various types of measuring instruments in the world.
While the specification is 13 +/- 0.5 PSI, what has come to be assumed is that you can inflate the ball somewhere between 12.5 and 13.5, and Brady has said openly (and it is not “illegal” for him to do so) that he likes it closer to 12.5. The problem is that tolerances such as these are not intended to allow a range, but to allow for random error, of which there are two sources: the gauge, and the skill of the people using the gauge. What the procedure should be is to inflate balls to 13.0 pounds always, with the tolerance given to allow for sloppiness in the measurement or errors in the gauge.
If you do allow an interpretation of the specification as being a range, then the more precise way to state the specification is this: the ball shall be between 12.5 +/- 0.5 PSI and 13.5 +/- 0.5 PSI. It is pretty obvious that is not what was intended historically, but allowing individual quarterbacks to go low (as Brady likes) or high (as Aaron Rodgers likes) means that my verbiage is the way the specification needs to be stated, at least in order for it to be consistent with reality.
What that means further is that in practice if a ball is on the low end, and you did have some type of an imaginary perfect gauge to measure it, the ball might actually be 12.0 PSI even though it passed approval at 12.5. In evaluating the circumstances of what happened during that AFC championship game, this is what needs to be assumed. No one knew then or knows now any more than that an approved ball, if it could be measured by an imaginary perfect gauge by an imaginary perfectly trained official, might be anywhere between 12.0 PSI and 14.0 PSI in reality.
Now of course Boyle’s law takes over, and provides a simple explanation: a ball at 12.0 PSI in an 80 degree room will be 11.3 PSI at 50 degrees, which allowing for further imprecision in the measurement technique (not to mention the bent pin on the gauge) is pretty much what they measured. (Remember also that three of the four Colts balls were below spec!) Furthermore nowhere in the Wells report is it mentioned, and to my knowledge no journalist has brought this up, that pressure loss is not merely a matter of air temperature but heat loss. Temperature is a measure of heat, but heat can transferred out of the interior of the ball via evaporation in addition to air convection. The only way to get a handle on all of this tolerance stacking, and to come up with a proper procedure, is to do controlled experiments ahead of time. All of these tolerance interpretations and physical analyses survive Occam’s razor because no imagined assumptions need to be made at all. There are a number of assumptions, but they are all well known facts about physics and the abilities to make physical measurements.
Tom Brady is being forced to take the fall because the NFL leadership has no intuition for science.
I have demonstrated that is is highly unlikely the Patriots altered the balls, and also that there is a reasonable explanation for why they were deemed to be low at half time and topped off. I have challenged the accusers to be more specific in their accusations, or else the accusations are out of bounds in the realm of logic and science. (I recognize that people “believe” in all kinds of unicorns, of course.)
Roger Goodell and the league office are who we are judging, so their conclusions must be rejected and we in the public at large need to be the arbiters. How important is deflategate in the grand scheme of things? Am I expending a lot of energy on something that ultimately is trivial? I don’t think so or I wouldn’t do it, and I will hopefully have more to say about that in a future essay. I will also, in future essays, address a few other dangling issues, but most of all (eventually) show how Roger Goodell is taking a classic car built by Halas and Lambeau and getting it stuck in the mud. Fortunately the tide is starting to turn on him now among thoughtful journalists, and maybe I can play a minor role to help push the car out of that mud.