Last Sunday night was off to a good start, at least in the context of what's appropriate in these disorientating days of pandemics, coups, and the Buffalo Bills winning the AFC East - I ordered eight chicken wings, but they gave me thirteen. Longtime wing connoisseurs will know that sometimes more wings means less wings, in the sense that a higher number for its own sake is meaningless when each piece is its own size, but a spread of thirteen to eight makes this a remote possibility - for eight to be greater than thirteen, you'd need one of those eight to be a quarter chicken. Anyway, my wings meant Sunday night was off to a flying start, and I sensed it was a good omen for the upcoming finale of the NFL season, where Philadelphia and Washington would determine the fate of the NFC East.
My positive mood lasted until around the fourth quarter, when Doug Pederson woke up Nate Sudfeld and sent the Eagles backup into the game for starting quarterback Jalen Hurts. The rest is, at least for helmet football fans, history, and already in the running for 2021's lowlight reel - the Philadelphia Eagles turned a competitive contest into a televised JV-Varsity scrimmage, chickening out for a quarter that ended with Washington winning the division and clinching a playoff berth by process of self-elimination; the losing side comprised not just of the Eagles, but anyone who had tuned in for a good game. Reader, if you need a sense of how bad it was in the moment, please note that NBC's commentary team, who are paid to sell the game to viewers, spent most of the fourth quarter in thinly-veiled disgust - neither Al Michaels or Cris Collinsworth went so far as to say the magic word, but Pederson was tanking (1). The odd thing about the commentary team's restraint was that they weren't so shy about criticizing Pederson just a few minutes earlier regarding a play-calling decision. It was on Hurts's last play before being replaced, and it led to an incomplete pass on 4th and goal from the four-yard line. The reaction was immediate - Collinsworth pointed out that going for it was supported by analytics, with the decision linked to a 5% increase in a team's odds of winning, to which Michaels replied along the lines of "I want to have a look at that math."
I didn't realize it at the time, but perhaps the reaction to the Sudfeld substitution - and its implicit commencement of tanking - should have had the same tone. The unrealistic aspect of my thought is that although I see tanking as the purest manifestation of analytics in sports, most people who follow these games don't see it the same way - analytics, being new, must therefore include only new ways to justify silly risks; tanking has been embedded in the realm of acceptable tactics for so long that experienced viewers miss the obvious. Tanking has all the hallmarks of good analytical thinking - it improves the chances of eventual success without guaranteeing it, leaving the matter of connecting action to outcome in the hands of execution; the doubters often center their objections around this intermediate step, and question whether the right factors are being considered in the calculation. This was evident in the way NBC's commentators questioned the fourth-down decision, essentially suggesting that the odds of scoring on the play were far worse than indicated by the analytics, but you'll never hear this kind of dispute when a team is tanking; protests usually center themselves around a set of high-minded principles, including "the integrity of the competition" or "the right way to play the game", but they never demand a closer look at the math.
I suspect one reason the math of tanking is accepted by most observers is because the concept could not be simplified further simpler - the earlier you pick, the better chance you have of selecting a superior player, so of course you want to do all you can to pick at the top of the draft. To restate it with the phrasing I used above, tanking is the action that improves potential outcomes in future seasons, with drafting being the execution step that connects the two. This is so easy to understand that I think most people accept the premise without asking further questions; if analytics were taught in K-12 education, tanking would be the introductory concept, covered between naptime and recess. An informal examination of draft results suggests the anecdotal evidence is weak but consistent with the practice - in 2017, the Bears selected Mitchell Trubisky ahead of eventual superstars Patrick Mahomes and Deshaun Watson, while in the following season future MVP Lamar Jackson was the fifth quarterback selected. Those examples confirm picking early doesn't guarantee getting the best player, but the teams at the top of the draft indisputably had the best chance - the narrative is not a failure of analytics, but a failure of execution.
However, in the paraphrased spirit of Al Michaels - I'd like to have a look at that math - I suspect the tanking equation is missing an important variable, which I'll label with the broad umbrella term of "culture". In essence, what I'd like the analytics of tanking to consider is the effect tanking has within the organization, and whether its negative cultural effects offset any advantages gained from securing a higher draft position. This was on my mind last Sunday night as I texted some fellow helmet football fans that Doug Pederson should be fired immediately, as in on the field when the clock struck 0:00 - I had thought his decision would have a crippling effect on next year's team, and in my mind a new coach gave them the best chance of returning to winning ways.
This kind of winged postgame reaction tends to soften with time, but I've only become more convinced over the past week. There were a lot of players on the field who were giving it their all based on a belief about competition familiar to anyone who has played sports at a high level - Doug Pederson had publicly ridiculed those beliefs in front of a national audience. Some of these players were on the last day of a contract, risking potentially career-altering injury by playing, but were willing to do so because the desire to win superseded selfish considerations - Doug Pederson made a mockery of their team ethos. These players had also just gone through an extraordinary NFL season defined by the COVID-19 pandemic, taking on risks both to themselves and their families that are inherent to any work environment in this moment, all of it for the sake of entertaining me and you by completing a televised helmet football season - Doug Pederson decided to cash in on their professionalism and treat the game like it was part of the preseason, which the NFL had cancelled this year on the grounds of exhibition games being an unnecessary risk.
I can't imagine what kind of culture a coach thinks he's building when he demonstrates that the sacrifices players make to win games are a secondary concern to the analytics of tanking. I'm sure there is a spreadsheet somewhere that proves the premise of drafting as high as possible - and I bet the math on it is indisputable - but I wonder how much the spreadsheet will matter next season when he needs to convince these players to lay it on the line. What if the situation with COVID-19 remains unchanged and the NFL once again gives players the option to opt-out of the season? I can't imagine last Sunday will make these players more interested in risking their safety. How will the team respond if they are once again knocked out of playoff contention? My hunch is that the locker room will splinter as it always does when a team becomes a collection of individual contractors, with each one pursuing selfish interests ahead of the collective good because of this evidence that the coach isn't going to offer his best effort to win meaningless games. I wonder what's going to happen next season when Jalen Hurts falters at the end of a close game - will the players remember when the coach benched him instead of giving him a chance to earn valuable experience? I don't think a single free agent turned off his TV last Sunday and said to his family - I want to play for that guy, he'll make a mockery of me and my hard work.
There's no way to know about the true effect of these things - and of course I would know even less in this specific case, given that I've never been in an NFL organization in any capacity - but I think these cultural questions will need to be considered next year if the Eagles have another poor season. There is something about winning in the NFL that, just like any other form of success, seems habitual to me - good teams seem to learn how to win over a period of multiple seasons while losing teams seem to return to the top of the draft, over and over, despite the alleged advantage of picking early. Like any habit, the habit of winning is developed by doing it, which means NFL teams should treat each of their sixteen games like a priceless opportunity to reinforce the behavior while building a culture of winning. I don't think I'd see this kind of thinking represented in the tanking math. I think I'd see what so many others suspect when they question analytics - technical precision, but very little nuance as it relates to the specifics of the decision and the people involved in the situation.
This mentality in analytics is the biggest problem I have with the practice, and why I fear it will fall short of its vast potential - no matter what the calculator says, you can't forget that the way people react to a decision is a massive consideration, and you can't make this reality go away by treating people like interchangeable pieces on a gameboard. You can't, in other words, do any meaningful analytics without factoring in the way people will respond to the decision. Until this shortcoming is corrected, the recommendations of analytics will be no more than expressions of yet another biased perspective, based in a fictional world where intangible qualities such as mutual respect, trust, and confidence - in other words, the key elements of a winning culture - have no place among the decision-making criteria.
Footnotes
1) By tanking, I mean a team deliberately losing games to improve its draft position in the following year. For those new to this topic, the major professional sports in America reward the worst teams in the following season with the first pick of new players. There isn't an explicit rule against the practice, but for the most part tanking teams are regarded with derision, ridicule, and scorn as they accumulate losses while their fans wait in an embarrassed purgatory and count the days until the draft. Some like to cite this as an example of "socialism" in American sports, which isn't strictly true - let's just agree that it's definitely not capitalism.
