A problem facing cricket analysis
An explanation for why cricket players and cricket analysts frequently clash
Robin Hanson of Overcoming Bias recently wrote about a topic that has grabbed my attention. Several fields have "fighters" and "fight analysts" but the relative attitudes towards these groups varies from field to field. On the one hand are financial analysts, whose non-partisan analyses are appreciated as much as the accomplishments of financial traders; on the other hand are sports, in particular cricket, where our fandom of elite players perhaps predisposes us towards reading analysts with suspicion. I would encourage you to read the full article but the crux of the explanatory thread is contained in the following extract:
So why do we celebrate fighters relative to fight analysts so much more in some areas than in others? The theory that occurs to me is that in the more we really like certain kinds of fighters, and the more unrealistic the images they tend to present of their actions and motives, then the more we dislike the analysts who are likely to dispute such images. On the other hand, if ... we can mostly accept the descriptions of fights that neutral abstract analysts are likely to come up with, then we are more willing to respect and celebrate such analysts.
The striking originality of this idea and the light it shines on the "culture wars" in cricket between players and analysts are worth pontificating on. I would love to do so in this article, but my main interest lies in taking this idea forward. So here's my first proposition: it is a problem unique to sport that the fighters and fight analysts of our field are as psychologically distant as they possibly could be. Analysts are generally of a certain breed - critical, conscientious, open to experience, and politically leftwing. This is a matter of selection similar to what results in most academicians being left-leaning: one has to be open to new experience to produce new knowledge.
Whereas on the other hand, athletes, at least in men's sports, are extraverted, agreeable, cautiously choose experience, and are politically rightwing. This too is a result of selection as frequent motivators for sporting entry include patriotism and traditional ideals of manhood. In other words, cricketers and cricket analysts are perfect complements on the Big Five personality traits. This could be great news, for players and analysts could combine their unique gifts and predilections to produce a well-rounded theory of cricket. But what I increasingly see happening is that cricketers and cricket analysts fail to see eye to eye with each other, on a variety of topics. Why is this?
Which brings me to my second proposition. I think that one of the reasons for the noisy conflict between cricket players and cricket analysts is that most of the good cricket analysis that is being carried out today unfortunately happens in online subcultures. Why is this unfortunate? Well, although x.com may be partly responsible for the birth and existence of such analyses, the point of social media is to produce positive assortative matching between its users. Positive assortative matching occurs when there is a correlation in sorting between the users of a platform that would not occur if they were sorted randomly. This is why a good course instructor would not let her students form their own assignment groups as assortative matching will ensure that the studious kids flock together. In the context of social media, as you probably know, it is how echo chambers, often filled with highly smart people, but those who think in a similar way, are created.
The rest of my argument is best formulated with at least an iota of rigor. One way of thinking about the confidence [in making a claim] is as the probability that one would observe evidence in favor of one's claim. Suppose p is the probability that an analyst would observe evidence in favor of her claim in the real world or on the cricket field. It is reasonable to think that in a positively assorted community on social media, where she is likely to be met with more favorable evidence, her confidence level would be p* where p* > p. (To make things less abstract, I am thinking of "evidence" in a very loose manner: positive feedback from one's mutuals on Twitter can also constitute favorable evidence in this framework.)
This would explain how it seems that a vast amount of cricket analysis that is produced online is sold with a greater dose of confidence than warranted. Of course, I am not contending that this is the only explanation. When it comes to broadcast analysis or analysis on a popular cricket webpage like ESPNcricinfo, the bigger reason is probably the important place held by narrative in sport, and if there is anything that stands in the way of telling a good story, it is skepticism about what you are trying to say. But I do think that analysts, nudged by the feedback mechanisms we are all today slaves of, sometimes fail to consider alternate causal channels and present their conclusions with unjustifiable confidence, leading to ridicule and antagonism from cricketers.
And I do know that suspicion of analysis predates social media. It is easy to forget in today's world where nerd culture has become subsumed into pop culture that we once lived in a culture that was actively discouraging of any form of systemization of the unsystematic. But the same thing can be caused by multiple causal channels and, perhaps because of my recent interest in the ills of social media use, this is the mechanism that appeals the most to me.
Either way, I think it is tragic that influenced by this and several other phenomena cricket analysis is fast becoming either a niche subculture of social media or an occasional curiosity on broadcast channels and on ESPNcricinfo. This is problematic because it will lead to poorer quality of analysis as the range of analysis will be constrained by the scope for narrative and the likelihood of clicks. Why perform a detailed analysis of the heterogenous effects of wind vectors on the T20 World Cup in the West Indies when you can reel off a thousand-word story on how Virat Kohli overcame his slump based on five statistics that will clearly bring in more advertising revenue. The other avenue for analysis to thrive in is through private communities of analysts who aid each other. And while this is promising, it is also only a second-best option. Remember that the history of our world's development is the history of our sharing of knowledge.
All this has led me to wonder whether the confluence between analysts and teams in cricket is destined to be a cyclical event that pops up in response to external shocks. Like the wave in 2020-21 which was clearly the child of the lockdown, or the current mini wave which has been triggered by an unexpected rise in data access. This may appear to be an overly pessimistic conclusion, and I am not a pessimistic person in general, but it does appear plausible to me. Unless our best analysts regain that elusive virtue that we all could use in the modern age, humility.