Anyone who remotely keeps up with professional basketball has heard the following, whether it’s from pundits or other fans, and the familiar refrain goes something like this: “Player X (insert name here) does a lot of things that doesn’t show up on the stat sheet/box score, but he adds value to the team.”
The very fact that basketball watchers have been hearing this refrain for years on end is indicative of one thing: namely, that professional basketball has not achieved reflective equilibrium on how to measure player values.
“The method of reflective equilibrium consists in working back and forth among our considered judgments (some say our “intuitions”) about particular instances or cases, the principles or rules that we believe govern them, and the theoretical considerations that we believe bear on accepting these considered judgments, principles, or rules, revising any of these elements wherever necessary in order to achieve an acceptable coherence among them. The method succeeds and we achieve reflective equilibrium when we arrive at an acceptable coherence among these beliefs. An acceptable coherence requires that our beliefs not only be consistent with each other (a weak requirement), but that some of these beliefs provide support or provide a best explanation for others. Moreover, in the process we may not only modify priori beliefs but add new beliefs as well. In practical contexts, this deliberation may help us come to a conclusion about what we ought to do when we had not at all been sure earlier. We arrive at an optimal equilibrium when the component judgments, principles, and theories are ones we are un-inclined to revise any further because together they have the highest degree of acceptability or credibility for us.”
Defined thus, reflective equilibrium is both a process and a result. The process comes from the fact that we need to reconcile our claims about particular instances with general principles, and the result is the state in which claims about particulars and the principles are coherent with each other.
How does this apply to the context of measuring a player’s value in professional basketball? There are, in my opinion, two levels of gauging player values. First, on a more intuitive level, there is just simply the act of watching how a player plays in games. If you watch enough basketball, if you keep up with certain teams, if you follow certain players, eventually you come to gain an intuitive, if not articulated, sense of how valuable that player is to his team. However, there is a higher, more abstract level of measuring player values, namely: statistics. Currently, professional basketball has an official set of statistics that it measures for each player, and they are the items that make up the box score as we know it: minutes played, points, rebounds (sub-divided into offensive and defensive), assists, steals, blocks, personal fouls, turnovers, field goals, 3-point field goals, and free-throw percentages.
The reflective disequilibrium comes from the fact that one’s intuitive understanding of a player’s value is sometimes dramatically different from the statistical measures of a player’s value. Hence, that familiar refrain: “Player X does a lot of things that do not show up on the box score, but he adds a lot of values to the team.” That this is true is undeniable, as anyone who watches basketball somewhat seriously can tell you. So this suggests that thus far, we have no achieved coherence between two ways of gauging player value, even though on some level, most basketball fans acknowledge that both ways of measuring player values are true in a non-trivial sense.
An aside: for a really good example of this reflective disequilibrium, check out Michael Lewis’ (he of Moneyball fame) in-depth profile on Shane Battier, which, in my opinion, demonstrates in no uncertain terms the very real existence of this reflective disequilibrium.
If this reflective disequilibrium exists, how can we achieve equilibrium? I think the effort is already underway, and this is what I call the “analytic turn” in professional basketball. How does the analytical turn try to achieve equilibrium? It does so by coming up with new statistical measurements that better match our experienced, intuitive understanding of player values. While it would take too long to examine all the ways in which these new statistical measurements better capture our intuitive, if very experienced, understanding, certain things are easy to pick out. For example, one low-hanging fruit would be to measure charges taken: this is one long-neglected stat that fail to capture an important aspect of playing defense. A good defender will often voluntarily take the charge from an oncoming opponent, resulting in the opponent’s being hit with a foul and preventing him from scoring. This has a double positive effect: it stops a score, and it limits the opposing player by getting him into foul trouble. This is easily measurable, but yet it is not a part of the official box score that the league keeps.
This is just one little example out of many, and actors in the league, whether it’s statisticians or teams, are coming up with even more sophisticated measurements. As far as I can tell, the Holy Grail, so to speak, right now in the analytic movement in professional basketball, is to find the “Be-All, End-All” stat that accurately captures a player’s value. Some prominent examples include Dave Berry’s “Wins Produced Per Player,” John Hollinger’s Player Efficiency Rating, Adjusted Plus Minus, and BasketballProspectus‘ Wins Above Replacement Player (WARP). And this is not even to mention the various proprietary and confidental statistical measurements used by teams like the Rockets.
But, it’s just as important to note the fact that the reflective disequilibrium goes the other way too: sometimes, our understanding of a particular player’s value is based on official statistics that overrate their true value. I call this the “Fantasy Basketball Syndrome.” That is to say, these players would be great for your fantasy league, because they excel in all some of the official statistical categories, but in real life, they do not add value to their teams. That they exist is also undeniably true, but people tend to overrate them because they happen to excel in particular statistical categories.
So clearly, the disequilibirum exists on both ends, and the purpose of these new statistical measurements is to adjust both our intuitive understanding of player value with their quantative aspects. The ideal, at least in theory, is a set of statistical measurements that achieves coherence between our qualitative evaluations and our quantative measurements.
That this effort is underway is no doubt true, and some of these advancements are very interesting. However, the problem I see it, as it is with any emergent field, is the lack of consensus. Although many of the players in this movement share the same goal, they naturally have different, competing conceptions of how to go about it. I think it is high time that all the actors involved: fans, players, teams, statisticians, economists, and the league to start a commission of some sort to actively start developing a new system of statistics.
But will it happen? I have my doubts, because the current statistical measures officially kept by the NBA has entrenched a set of incentives. These statistics are part of contract negotiations. If you change the statistics, you inevitably change incentives from the various stakeholders involved, and in any large and entrenched institution like the NBA, change is unlikely, and if it does happen, it’s likely to happen because of a crisis.
Anyways, enough geeking out for the day.