Over the last week or so, the NBA blogosphere has been filled with debate over the use of statistical analysis in basketball — specifically with regards to basketball writing.
Kevin Draper of The Diss is probably the one who started the conversation. Kevin — who has posted articles using statistical analysis on this very site — posited the following arguments in a recent post:
- Advanced stats haven’t “solved” basketball; stats aren’t “certain”.
- Most writers don’t understand advanced stats.
- It’s harder to write well when using stats.
In Draper’s opinion, these three factors have led to a decline in the quality of basketball writing. Or as he puts it:
…the movement towards empiricism has resulted in a reduction in creative forms of writing. There is more writing that serves to win an argument or be “right” about a point, and less writing designed to explore the unknown.
Over at Raptors HQ, Braedon Clark — inspired by Kevin’s article – wrote this:
Lost in all of the (justifiable) hoopla surrounding the analytical revolution is one dirty little fact, one unpleasant little secret: analytics makes watching the NBA less fun.
This idea — which, it should be made clear, is not one shared by Kevin Draper – is one that I vigorously oppose. Statistical analysis enhances my enjoyment and appreciation of sport. Knowing exactly how improbable it was for the Heat to come back and win Game 6 of the Finals made it all the more exciting. Watching Tony Parker make this shot is even more exciting when you understand how unlikely it was to have gone in. Moving outside of basketball, even a cursory investigation using basic physics makes this goal by Roberto Carlos all the more spectacular.
Statistics are simply another form of empirical analysis, and empirical analysis is the essential feature of the scientific method. Science shouldn’t impede your enjoyment of anything. If it does, you’re doing it wrong.
Back to basketball writing….
I certainly agree with Kevin’s last two points. Most writers are drawn to writing, practice writing, and study…writing. The majority of writers have not only not studied math or statistics, but in fact dislike math, and have probably avoided numbers because of this. In a society that increasingly relies on math, for some reason it’s still okay to be mathematically illiterate (unfortunately, this is particularly true for females).
I can also attest to the fact that it’s difficult to use stats to write a compelling article. It takes a lot more time to complete a stats-based article than it does to complete an article that is largely devoid of stats. When I’m writing a stats-based piece, I start out with a question, or perhaps an observation. From there, I actually have to do some research and uncover an answer. Sometimes I find out that my question is unanswerable, or — more accurately — that it would take too much time for me to come up with an answer to the question. Oftentimes finding one answer begs another question, which begs another question, and on and on until I’ve spent hours doing research and have no article to show for it. Or sometimes I find an answer that isn’t compelling. When I do find a topic worth writing about, there’s the matter of working the numbers into the text, fiddling with graphics, and resorting to using tables (which, let’s face it, aren’t the best tools for displaying data). And remember, all this work is on top of the work one would expect from a regular article.
However, I think it should be said that I do not completely agree with Kevin’s first point. While it’s true that no one has “solved” basketball, and that even the most objective stats do not give us 100% of the picture, I do think that we are much closer to “our simulations are 100% accurate” than “it’s impossible to tell if LeBron James is better than Adam Morrison”. There should be no doubt in anyone’s mind that LeBron James is a better player than Adam Morrison. Even a basic understanding of stats and/or basketball should make that clear. Unlike stats in football, hockey, or soccer, basketball stats are relatively consistent from year to year and give us a pretty good understanding of the contributions individual players have on team outcomes.
That being said, I don’t disagree with Kevin’s ultimate conclusion. With the increasing importance of statistical analysis in basketball, more writers with little knowledge of stats are trying to use stats to back up their arguments. And writers who are focused on stats-based arguments focus on stats to the detriment of their writing. Neither is ideal — although you can guess which of the two I prefer. I’m more interested in reading basketball analysis than I am in reading well-composed text about basketball; if I wanted to read “good writing”, I’d pick up a book by one of my favourite authors.
In an ideal world, writers would learn more about stats and analysts would learn more about writing. Eventually, I expect everything to sort itself out, as the best basketball writers — the ones who can both write well and understand the stats — will become more renowned. But that may take a while, especially given journalism’s current struggles.
In the meantime, the solution to this kind of problem is easily solved by reading basketball writing that you find enjoyable. Started an article and it isn’t grabbing you? Forget about your sunk costs and stop reading. Quitting has its benefits. Really intrigued by an article’s first paragraph? Keep reading. Hopefully, after enough of these events, you will start to see patterns and find writers who consistently meet your expectations.