clock menu more-arrow no yes

Filed under:

Offseason Thoughts: Advanced Analytics

New, comments

How has advanced analytics changed the way we talk about the NBA?

Houston Rockets Introduce Jeremy Lin Photo by Bob Levey/Getty Images

The battle between the eye-test and advanced analytics seems to be over, with the fight going in favor of the new-fangled statistical modeling and “Moreyball,” where we judge a player by their RPM, BPM, and other fancy acronyms without really understanding the math behind them. Teams favor the more efficient style of basketball, which is heavily focused on three-pointers.

The last vestiges of the old-school style are mocked, even if they can be a bit purposefully abrasive, such as when Charles Barkley claimed that the Warriors could never win a title because they were a “jump-shooting team.”

The Timberwolves are fascinating to examine from the lens of advanced analytics, or at least public ones, vs a more antiquated approach. Andrew Wiggins probably is the divisive player in the NBA when we compare his perceived value against what every single advanced statistic shows. Tom Thibodeau has re-made this team to play in a manner that most of the NBA has decided is unworthy.

It’s hard, in this day and age, to imagine what this debate used to be like. In a world where Basketball-Reference isn’t bookmarked on dedicated NBA fans’ web browsers, where there isn’t a veritable deluge of internet commentators and writers trying to come up with their own take, when people simply “played” basketball.

From the perspective of an advanced analytic skeptic, particularly those who played or play in the NBA, the movement must be incredulous to behold. Who exactly are these people who have never played a game of basketball at a competitive level that are saying what you are worth, how good you are? This, of course, often happened with reporters, but now we are telling players, no you don’t understand, you are bad at this thing because of math. Your expected rebounding percentage has an acceptable rate of 10 percent to 20 percent and you are falling woefully short.

How do we as writers really have anything to offer? We have not lived the actual experience of playing in the NBA. We, for the most part, do not attend practice. We do not see coaching outside of actual in-game experiences. We do not train over the summer. Our knowledge is inherently extraordinarily limited.

We are like, to borrow an example from philosophy, the man sitting in a room with a computer pretending that we know Chinese.

In this thought experiment, there is a someone in a closed room with a computer. On one end of the room, slips of paper with Chinese symbols come in. The man scans the symbols into the computer, which a program then understands and spits out another slip of paper with the appropriate Chinese symbols. The man then takes that new slip of paper and sends it out the other end of the room.

To the outside observer, whoever, or whatever, that is in that closed off room understands Chinese, as the outside observer is having an effective conversation with whoever is inside the room. But the man inside the room knows this cannot be further from the truth. The man does not understand Chinese, he is simply following directions.

This thought experiment is often used to represent the issue of machine learning as a simplistic Turing Test, as the computer itself does not truly understand Chinese either, it is simply following the directions of its programming.

However, in my opinion, the thought experiment also represents a Phenomenological problem. The man does not know Chinese because he has not lived having known Chinese. How do you know something, or claim to understand something, that you can never truly know?

However, this is not singular to writing about the NBA and basketball. All of life is like this if we take this argument to its logical extension. If we can never truly know anything outside of our own experiences, what do we have to offer that is not narrowly focused to our own lives?

What the advanced analytic movement (and the Internet in general) offers is a sort of flattening the requirements to have an intelligent and respected conversation about basketball and the NBA. I may not have played college basketball, but I can interpret and present statistics quite well. The requirements to enter the conversation dropped significantly and the prior set of rules was thrown out. That’s a good way to piss off people and generate a backlash about the “right” way to do things (see also, men’s rights activists or a certain person in power).

Like everything, the real answer is a push and pull somewhere in the middle between the two extremes. It’s good to be cognizant of the fact of how limited our scope of knowledge really is when we talk about the NBA, but it’s also important to be reminded that we are adding to the conversation rather than simply drowning ourselves in a din of uninformed hot takes backed by performative numbers.