Part II of Ant’s look back at the NBA decade that was.

In Wednesday’s post, I wrote about two mid-decade events that changes the stylistic landscape of the NBA. But there was another big factor that helped improve the quality of play league-wide – namely, teams got smarter. Or rather, the better teams got smarter. At the turn of the decade, these better teams were exploiting their superior overseas scouting to find all-star talent at scrub prices. The Spurs picked up two of their three core stars for a bag of a potato chips in 1999 and 2001. The Jazz found the incredibly talented if ultimately frustrating Kirilenko at the tail of the first round, and current centre Mehmet Okur was drafted (by Detroit) in the second round. It seemed like teams that were willing to put in the effort and were prepared to be patient in waiting for players to come over from Europe were reaping enormous rewards.

As that advantage was shut off by more teams investing more dollars into international scouting, these better teams changed tact a little bit and began playing with numbers. Pick out any one of the consistently good teams of the last decade, and chances are they have their own statistical research department. It’s not all that easy to pinpoint when exactly these were introduced since they tend to be fairly hush-hush, but when ESPN decided, in 2005, to hire a prominent egghead in John Hollinger to their writing staff, they opened the floodgates for statistical discussion to leak into the public domain.

Initially, damn near everyone was sceptical about what numbers could tell us. After all, where baseball (which had undergone its own statistical revolution much earlier) was a structured sport with well-defined parameters, basketball was far more freeform. A batter, for example, always hits from the same spot, whereas shots in basketball are spread all over the floor, with varying degrees of defense on each one. Whoever could possibly account for all that?

Yet some people quickly started to come around as the results of these numerical investigations lined up very well with what our own eyes told us. More often than not, good players were rated highly, and duds were scored as duds. The system worked. Terms like PER and pace and more recently plus/minus began to enter the basketball vocabulary – so much so that the NBA has started including the latter metric on every box score.

As a numbers head myself, I was positively giddy that some statistics other than the usual Pts-Rebs-Asts were getting some play. No longer would I have to hide out in the caves with my calculator! And for a while there it was really nice to be able to discuss some numbers in the context of NBA basketball. But unfortunately, the honeymoon didn’t last long. Instead of being widely ignored, NBA statistics quickly became one of the big divisive issues in NBA discussion. In this very American sort of way, you had to decide whether you were with us or with them; whether you were a numbers guy or a gut/instinct guy.

That sort of divisionism is usually the death of any good discussion. Even worse, folk on both sides begin to deal in half-truths and propaganda. New proponents of basketball stats began making some wild assertions about the power of their numbers – assertions that the people who come up with those numbers would never support. It’s not uncommon to see someone unequivocally state that Player A > Player B because their PER is half a point higher, or something equally silly. Meanwhile, those on the other side will dismiss PER wholesale based on one nitpicky flaw, or worse accuse Hollinger (or whoever else) of intentionally building in biases into their systems. The entire debate feels like a glorified shit fight at times, with those actually prepared to discuss the topic at hand rationally being hit by turds from both sides.

Here’s the thing – basketball statistics were never designed to supersede actual vision of games, they were simply designed to aid analysis or give us a rough outline of certain players. Basketball, by its very nature, will never be quantified the way baseball has. There are simply far too many variables. But it isn’t football either (whichever version you prefer), where so much of the game depends on physicality or one or two big plays – things that are beyond the scope of numbers.

That, in many respects, is what makes basketball such a beautiful game. It sits squarely in the grey between the white of quantifiable sports like baseball (or cricket) and the black of football. We can use numbers to guide us – and in a league where 30 teams play 82 games each that’s damn useful – but never to give us concrete answers. I can, for example, go to basketball-reference.com and get an idea of who’s playing well or not so well for the Pacers. I simply don’t get to see them very often, so I have to rely on the numbers (or various blogs, of course) to give me a rough outline. But that’s all it is – a rough outline.

That’s one of the real tragedies of this whole shit fight; here we have a sport where numbers and visual evaluation can co-exist happily side by side, yet for the most part we’re unwilling to do it. My hope for the next decade is that somehow it all gets worked out (right after the Middle East thing, I guess). It would require two simple things; integrity from the pro-stats guys, and open-mindedness from the anti-stats guys. That’s it.

Integrity simply means not making grand statements for cheap publicity. It means acknowledging that all statistics are inherently probabilistic beasts, and the evidence has to be overwhelming before you draw any strong conclusions. It means not saying “I would not sign [Kevin Durant]” for free (last paragraph of link) just because his adjusted plus/minus (a statistic subject to wild fluctuation) was bad for his first two years.

It’s no different to any scientific pursuit really – if the guys at the top fudge or misrepresent the numbers to push their own agendas, everybody’s poorer for it. It breeds distrust amongst the community at large and makes any discussion very difficult. And although the majority of “experts” in the field are first-rate, the few bad apples simply give ammunition to those that look to shoot down the numbers game at every turn.

I sincerely hope that those that do wish to play with numbers have the integrity (and understanding) to correctly represent their numbers as the probabilistic indicators that they are – rather than objective truths – and in the process push the discussion in the right direction in the coming decade. I hope that those still unconvinced respond in good faith. The result could be something truly wonderful, with our capacity for understanding and discussing the game of basketball greatly increased. And that’s what it’s all about, right?

But if this divisionist bullshit persists, we’ll inevitably see the two sides drift further and further apart in the coming years. And if my only choices are to rely strictly on numbers or alternatively to ignore them altogether, I’d rather just abort the discussion and stay silent. Both perspectives are as ignorant and misinformed as the other.


Tags: , , , ,

« « Previous Post: Sunday Six Pack: Gallo, Reke and the “T” word
» » Next Post: Tracking the Bogometer