Monday, May 21, 2018

Unresolvable questions and the search for understanding

Researchers want answers, and ideally, they want good, solid answers that will stand up to examination and challenge. Unfortunately, some questions do not have any fixed answer. The absence of clear answers, however, does not preclude important learning, and researchers can benefit from being able to delve into such uncertainties in search of many interesting ideas, even if no specific answer can be found. 

In the late 1950s, a philosopher named W.B. Gallie demonstrated that there were some things could never be fully defined—he called them “essentially contested concepts.” In the late 1960s, design theorist Horst Rittel argued for a class of “wicked” problems (which included but was not limited to design problems) whose members had no definitive formulation, among other characteristics.  For both Gallie and Rittel, a crucial factor was the social element: different people view things differently. Gallie relied on a sporting example derived from cricket (I believe—I’m working from memory) to demonstrate how different views about the sport made it impossible to define the “best.”.  Gallie’s example is particularly salient for me because I enjoy the sporting fan’s common questions regarding which players are best and what teams should do—one of my favorite authors is Bill James, the baseball analytics guru—and yet, as a philosopher and researcher, the more I look at such questions, the more complexity there is to see. And ultimately, given that I accept the ideas of Gallie and Rittel, I see these questions as unanswerable.  Despite believing these questions are unanswerable, I still see the debates that they produce as interesting and often informative.

Not to mention that I am somewhat interested in basketball and the NBA playoffs, and a lot of current discussion revolves around an unanswerable question that is, nonetheless interesting, and perhaps even informative, to explore. That question is the question of LeBron James. How great is he? Is he the greatest ever, the GOAT (Greatest Of All Time)? Or is he “just” top five? You don’t have to read much to find people discussing LeBron’s “legacy.”

Before LeBron, Michael Jordan was generally regarded as the greatest of all time (GOAT). Some argued for Bill Russell with his 11 championship rings and multiple MVP awards, or, perhaps, for a few others, but Jordan was the most common choice as GOAT. LeBron, however, has been doing amazing stuff that no one else in basketball can do, and his accomplishments are piling up. He is currently having another spectacular postseason, at least in terms of individual performances, although his team is facing a 1-2 deficit to Boston in the conference finals.  It is these performances that spark the debate: “look at that performance,” says one side, “he’s the best ever.” The other side says “well, it doesn’t mean much if he loses. Jordan won all six times he went to the finals!” 

This is the simplistic version of the argument, of course, because on closer examination this simple argument will reveal complexity that cannot be eliminated—complexity of the sort that contributed to the claims of Gallie and Rittel that some things cannot be completely defined. The simplicity of saying “Jordan won all six times in the finals; LeBron only won three and lost five,” might be fine for chatting at the bar during a game, but it certainly isn’t enough for serious research.

Firstly, we note that we can’t just reduce the argument to who has more rings, because by the “rings” standard, neither Jordan nor LeBron is all that close to the top of the list. By the rings standard, Bill Russell is the greatest, followed by a bunch of his teammates and Robert Horry. And, with all due respect to Robert Horry and Russell’s Celtics teammates, they are not all-time greats—no one is suggesting that Sam Jones, with 10 rings, is the second-greatest player ever. Trying to reduce the debate to a single dimension distorts the question: a player’s performance is much more complex than that single dimension.  This problem of multidimensionality makes it difficult to evaluate many things—what is intelligence? what is creativity? what is a “good employee”? how do we evaluate students’ learning?

Once an issue is understood to be multidimensional, it becomes increasingly difficult to make any certain decision. Beyond championship rings, there are many statistics that allow a comparison between two basketball players—points scored, rebounds, assists, etc., etc.  But the more dimensions added to the evaluation, the greater the likelihood that there will be contradictory indications. If one player had better stats in every possible category than any other player, there wouldn’t be difficulties. But that’s not the case, and that leads to complexity and uncertainty: how do you choose to weight different dimensional in an overall evaluation? If Jordan has more points but LeBron has more rebounds, who is greater? What’s more important for evaluating greatness?

Focusing on won-loss records can give an example of this problem of evaluation. MJ won more rings than LeBron, and that matters. But LeBron advanced to the finals more times, and that’s worth something, too: after all, if LeBron is criticized for losing in the Finals, shouldn’t MJ be criticized for losing in the Conference Finals? How do we compare those different achievements? Or, at the other end of the playoffs, we can see that MJ lost in the first round three times, while LeBron has never lost in the first round. If MJ were clearly greater, shouldn’t he have a better record in the first round? There are lots of stats and sometimes Jordan’s are more impressive (30.1 pts/gm vs. “only” 27.2 for LeBron), sometimes LeBron’s are (7.4 rebounds/game, 7.2 assists/game vs. 6.9 and 5.3 for Jordan).

The search for detail in comparing the two may not lead to any conclusive answer about which is better, but it can help us see the question more richly, and this can inform us about basketball and about processes of evaluation.  And, in a way, what we get out of the examination is potentially more valuable than an answer: it doesn’t really matter who is “the greatest”—whether we say that Jordan is the greatest or LeBron is the greatest or Russell, Kareem, Bird, Magic, Wilt or whoever. It doesn’t really matter who gets called the greatest, or who really is the greatest (if it makes sense to reduce such complexity to such a simple question). But, although the question itself cannot be answered, what is learned in the process of trying to answer that unanswerable question can give us insight into the more general process of player evaluation, and that has practical value to basketball organizations or to fantasy players.

Some questions that have no answer are still worth asking and examining.


No comments: