Filed under:

# Why I Like Statistical Half-Truths

2008 Olympics Video Coverage at NBC Olympics.com!
On Sunday, former NBA coach Eric Musselman blogged about why he likes blogging:
The blog was almost like a re-birth. Thinking about basketball in different terms, reflecting on some of what I've seen and learned during my life and career, corresponding with other coaches, teaching, meeting new people -- it's been incredibly therapeutic for me.
I actually feel similarly about blogging about the WNBA, although I never coached or played pro basketball (and therefore have considerably less money in my bank account to feed myself with as I blog). And one way in which I’ve reflected on basketball in different terms is through the use of statistics.

Because the Olympic break is now upon us, I will probably be filling this blog with my Applied-APBRmetrics analyses in hopes of filling the void left by the break in WNBA games. I could just watch the Olympics, and I will, but it’s just not the same as watching teams compete over several months to be the best in a given league. I figured it would be nice to keep writing and lay out some things to look forward to once the WNBA action picks up again. I think it will be kind of fun to spend time crunching some numbers – rookie rankings, team rankings, MVP rankings, etc -- and then seeing how things play out in August. (Yes, I’m a nerd…and hopefully, some of you are tolerant enough to keep reading.)

So to justify my use of statistics, I look first to the words of others. Two people whose work and input I greatly appreciate, Kevin Pelton and David Sparks, have written about why they use statistics in their work. Their pieces resonated with me and inspired some additional thoughts that I thought I’d share before I make the plunge into the numbers over the break. And perhaps it will give you a little more insight into how I view sports fandom.

The Mentality of an APBRbitrarian

Pelton writes the following in his article, "Why I am an APBRmetrician":
[My history teacher] effectively managed to permanently convince me of the importance of providing evidence, more evidence and, when in doubt, still more evidence.

Since then, that mentality has permeated my thinking. When it comes to basketball, the best evidence available is usually statistics. So you think Player A is a good rebounder? Terrific, but do his statistics confirm that? Player B is a heady player? Good for him, but does his turnover rate or Roland Rating reflect that?

Not every qualitative basketball statement can be backed up with statistics, but I still find it important to support my claims, as Ms. Angersbach would have told me.
And in his recent manifesto at Hardwood Paroxysm, David Sparks writes:
Analysis is not the opposite beauty, methodological rigor is not the opposite of casual observation--rather each is a necessary part of a whole. For a fuller sense of enjoyment and understanding of any game, we look to statistics to confirm the impressions we have from just watching; just as sometimes, we look to the court to confirm what the numbers seem to be telling us. There is no right or wrong way to approach the appreciation or assessment of sports, and arguably, a perspective that ignored some aspect--be it gut reaction or regression analysis--would be substantially incomplete. All I know is that there is no dearth of subjective opinion available for your consumption, and all I can do is offer something a little different, and a little less arbitrary.
As Sparks alludes to, any account that ignored emotion, logic, or observation entirely "would be substantially incomplete." So although it might be more fun to set up a blog and just keep the caps lock on and say whatever comes to mind, that wouldn’t do much in the way of supporting the type dialogue that helps us understand things better. When dealing in a world in which the goal is to put a larger number on the scoreboard than the other team, we should probably find a way to balance our passion with numbers...and sometimes messy ones.

However, as a new fan, I find statistics extremely useful for a few additional reasons. In fact, I think WNBA statistics might be extremely valuable in attracting die-hard basketball fans (e.g. nerds) who enjoy the challenge of finding the magical formula(e) for various basketball phenomena, including winning championships.

First, when you’re new to a sport, there is just so much you have to soak in. To really enjoy it, you have to knowing the big stars, the teams around the league (and where they play), and eventually you learn the nuances of the game in terms of what wins championships and what falls short.

What statistics help me do is identify some of those key players or key elements of the game, and eventually see things in the action that I would otherwise miss. The stats have helped me to pick out players I never knew about (Deanna Nolan) and some I had just forgotten about since their college days (Lindsay Whalen). In general, I think statistics help me to focus my attention on specific aspects of a game when everything I’m looking at is new and different. The more advanced work that APBRmetricians do, can go a step further by actually explaining what’s important and what’s not.

I realize that not everyone is obsessed enough with basketball to dig deeply into the advanced stuff and I would hardly argue that they’re somehow "less enlightened". Of course you can learn to enhance your "fan vision" just by watching games and observing trends over time. The problem is that our minds have this nasty habit on fixating on things we like rather than approaching each game with an open mind. Even simple stats like points, rebounds, and assists can help us to judge a player’s performance more effectively and see the story of the game for what it was rather than what we want it to be.

Second, for a sport that leaves so many of its games unseen by fans who can’t show up at the arena, statistics are the only way to tell the story of how the season is progressing. And to really get into that, you have to do more than look at wins, losses, and terse recaps that don’t do the game justice. If I really enjoy a sport, I want to know the latest trends, what’s hot, and exciting new (and future) developments.

Although I believe stats enable quite a bit in trying to understand a new sport (for those who did not develop permanent math-phobia in high school), I think it’s important to know that they are only part of the story. Even statistics are not completely "objective" – the statistics that people choose to tell us inherently demonstrate one’s own subjectivities.

For example, assists don’t tell the story of turnovers, assist to turnover ratio doesn’t take into account pace, and pure point rating doesn’t tell us if a player is simply the beneficiary of a strong system. The statistical choices we make illustrate our own biases.

The last point has to be made...

Third and last, in the interest of full disclosure, I have always hated math. So it’s sort of weird that I’ve been drawn to statistics as a means to understand a league I’m unfamiliar with. In fact, I can say with confidence that I hated math in school…college…grad school…or looking at deductions from my paycheck (this could be used to build a much broader argument about the systemic flaws of school math, but that could fill a dissertation…and actually, I’m pretty sure it has filled more than a few).

My feeling has always been that in order to understand something well, we need to pay attention to the general patterns of the activity first and then work to figure out which patterns are significant. So the value of good observation is that it lays out the different dimensions of an activity so that we can better understand how all the pieces fit together. Good observations should not be dismissed as irrelevant or "biased" (what isn’t?) because they aren’t grounded in numbers; instead, it they are providing accounts of how a given activity operates so that we can better understand how people understand their own world or (in some cases) how to best intervene.

And that’s something I try to do here – identifying defensible patterns in how individual players and teams achieve the outcomes we seen in the box scores. To truly understand any game, even Risk, it helps to have some mental representation of how the action occurs. Within the best observations, we are able to identify the author’s dependent/independent variables, trends, and perhaps a low level association between variables. In other words, making good observations brings us closer to objectivity, not further from it, by contextualizing how things actually occur.

What it boils down to is that observations and statistics are just two halves of a whole truth. So what statistical half-truths can do is complement the half-truths that we observe. Rejecting one or the other – or favoring one over the other – is unproductive. To truly understand a game we need both. And I would hope that in the coming years, we’ll see more WNBA APBRmetricians trying to get WNBA statistics to reach the depth of NBA statistics.

In the meantime, I’ll just keep on doing my applied APBRmetrics work and waiting for others to do the tough stuff. The stats greatly enhance my experience watching the WNBA, and that’s why I continue to post them for others. But for those that doubt the value of statistics, I understand – I hated math too.