“For 150 years, ‘clubhouse chemistry’ has been impossible to quantify,” the article Chemistry 162 in the latest ESPN The Magazine begins. If you’ve read anything on sports analytics, you know what this means: the impossible to quantify has finally been quantified, and you, the reader, are one of the lucky people to receive this wisdom from on high.
The thing is, when something has been impossible to quantify for 150 years, there’s usually a reason. My natural suspicions of anyone claiming to have neatly solved a 150-year-old problem were confirmed when I read the phrase “proprietary team-chemistry regression model,” the high-finance jargon Jeff Phillips used to describe the system developed by himself and group dynamics experts Katerina Bezrukova of UC-Santa Clara and Chester Spell of Rutgers. The solution to chemistry is here. We’re not going to tell you what it is, but you should totally believe us anyway.
Although none of the math involved is shown — a universal sign that a metric can be dismissed — Phillips was kind enough to give a rundown of the three aspects of chemistry his metric was supposedly measuring.
“Demographic factor: The impact from diversity, measured by age, tenure with the team, nationality, race and position. Teams with the highest scores have several overlapping groups based on shared traits and experiences.
Isolation factor: The impact from players who are isolated because of a lack of subgroups from these shared demographic traits. Too much diversity can, in fact, produce clubhouse isolation for players who don’t have teammates with similar backgrounds or experiences.
Ego factor: The impact from individuals’ differences in performance and monetary status. Too few All-Stars and highly paid players signal a lack of leadership; too many, however, creates conflict. The ideal level falls in the middle.”
The article provides no other details on how the various numbers — Tampa Bay gains two wins via chemistry! The Cardinals lose 2.6! — come together. As Phil Birnbaum illustrated at his blog there are myriad issues with the numbers as presented — standard issues of mistaking correlation for causation, of confusing which factor is the cause and which is effect, and the like.
But the issues here run deeper than shoddy math. If treating “demographic traits” as something to be gamed and stating “too much diversity can, in fact, produce clubhouse isolation” weren’t a red flag, observe these quotes from the article:
“The Royals’ Isolation score, on the other hand, is 4.7 wins higher, boosted by a pitching staff that includes eight Americans among the team’s 10 core pitchers.”
“The Dodgers are the NL West favorites despite a 25th-ranked chemistry score that will cost them two wins. Their Ego factor, based on several pricey acquisitions in the past two years, costs them one win alone. And while the roster is diverse, the club suffers from many isolated players, including SP Hyun-Jin Ryu (South Korea) and reliever Kenley Jansen (Curacao), costing the team another win.”
The suggestions here are dehumanizing. Teams should be careful to promote diversity — but not too much! Teams need to balance the positive effects of adding a brilliant player from a new country with the fact — and it is presented as fact, here — that it will be difficult for him to mesh with his new club. When we begin attaching numbers to these concepts purely based on census check boxes — and not the wildly varying personalities we can see from people within those demographics — it becomes creepy at best and racist at worst.
But since the idea is presented with a level of analytical gravitas, ESPN apparently believes it is worth presenting. Since the popularization of Moneyball in particular, analytics have been subsumed into the massive sports marketing machine, to the point where at times we see obvious plays to an audience that ESPN assumes will eat up anything with a seemingly objective win value attached to it. Between the combination of a proprietary black box metric, the effectively absent explanation of the process, and the presentation of players not as people but as collections of attributes, this is one of the most blatant cases of marketing to an analytics crowd I’ve ever seen.
And it gets creepier: Phillips, the article’s author, is a member of “The Parthenon Group.” The Parthenon Group’s website states, “Our mission is to be the strategic advisor of choice for CEOs and business leaders worldwide.”
The issue of chemistry is not one limited to the baseball field or other athletic endeavors. As anybody who has worked in an office or other group environment can attest to, both productivity and the general mood are improved when everybody gets along well. The application of a chemistry metric for CEOs is then obvious. If employers could figure out how well people work together without getting to know them as people — something you just can’t do through a resume, or in the short time of an interview — hiring practices could be streamlined, and all that money wasted paying chemistry-destroying employees could be saved.
CEOs and executives would love for it to be this easy, for them to be able to hire and fire based on check boxes on an application or entries in a database. We know the idea that the Dodgers would be better off without Kenley Jansen or Hyun-Jin Ryu due to their unique backgrounds is ridiculous. They’re humans. And like all humans, they are far too unique to be fully represented by the country they hail from, too complex for a formula to understand how they will mix together with others.
There is no need to worry about this particular article. According to his LinkedIn page, Phillips is a former employee of ESPN and Chemistry 162 was likely little more than the result of a favor done for a friend in high places. My concern, rather, is that this fantasy of measurable chemistry off basic attributes is such an appealing one for people who occupy positions of hiring power, whether in sports or business or any conceivable labor field.
Team chemistry is undoubtedly important, and to ignore its effects on our teams and our lives would be silly and wrong. But the approach of Phillips and ESPN here treats humans as nothing more than cogs in a machine. It treats race and identity not as part of a larger human fabric but as items to be plugged into a regression analysis. This is an alarming misuse of analytics, and one ESPN should be embarrassed to endorse through its publication.