Page 1 of 1

The only constant...

Posted: Wed Feb 27, 2013 6:38 pm
by schtevie
In the context of my intermittent mini-crusade to have the empirical history of the NBA be taken seriously, I feel compelled to offer comment on a rather recent TrueHoop post: http://espn.go.com/blog/truehoop/post/_ ... -is-change. I really like me my Henry Abbott, but the factual presentation and related argument about historical matters in this post is really, really very not good.

Front and center is a graph that motivates the post (and conversation with Stu Jackson). We are shown a familiar rise in 3PA (though the data for early years are both incorrect and incorrectly labeled on the axis) and a perhaps less familiar decrease in FTA. Story: the only constant is change.

First red flag. There is, in fact, at least one eternal constant in the NBA, and that is: it is a very bad idea to foul three point shots. Accordingly, this is very uncommon. And this fact fully accounts for the apparent secular decline in FTAs over time. Were the plot instead to show, say, FTA/2PA over time, one would see something very interesting, namely a wave, with two peaks and two (and on the way to three) troughs. This too is change, of course, but of a completely different (and far more interesting) nature, and one that requires a different narrative (primarily responses to rules change regimes, both in player actions and enforcement). Perhaps a blog post for another day.

From here we turn to the interview with Stu Jackson, where very little is heard that is both nutritious and true. But more to the point, the degree to which falsehoods are related is perhaps a bit shocking.

Stu Jackson could be forgiven for not giving an honest answer as to why the three point shot has been so slow to be adopted. What is he going to say, that hewing to norms of respectability was, historically, more important than winning for NBA job security? So, we first get the nonsensical "there was going to be a bit of a lag time" argument. Note: it hasn't been "just" 20 years; it's been 34 years, not including the 9 years of ABA experience. And there was that funny bump up in 3PAs between 1995 and 1997. Explaining that would have been a nice follow up question.

But then we get the second reason, 3PAs were also supply constrained (again, never mind the amazing effect on supply of the experiment with the shorter line): "it took some time for teams to start making the personnel decisions that allow them to incorporate three point shooting". And then we get an example: stretch 4s are required to stretch the floor, get adequate spacing, blah, blah, blah.

Huh.

OK. Plausible. (Henry seems to nod in agreement.) But is it true? Well, basketball-reference says it isn't, if that means one should expect an increasing number of F-Cs over time shooting threes proficiently enough to be a threat (I picked 3fg% of 35). If anything, the correlation is negative. Is there another interpretation that I am missing that redeems his point?

Then the interview continues about the effect of rule changes on the supply of players able to "work their magic around the rim". This is then followed by another factual curiosity from the VP of Basketball Operations. Unprompted, the claim is made that "the small guard has come back into the game...certainly guys under 6 foot, playing the point guard position...".

Seems reasonable, but, again, is this true? Well, basketball-reference says no. Both for guards under 6' and 6'2" (I stopped checking there) their absolute numbers and numbers playing starter minutes hasn't increased since hand-checking rules went in effect. If anything, their participation has gone done over the longer term.

Well. So what? What's the point of fisking the VP? Perhaps none. Everyone, pretty much, prefers the rule changes allowing zone and eliminating hand-checking, etc. Still, it is a bit disconcerting when the person nominally in charge of instituting rule changes so as to modify the nature of the game gives the impression that he both "knows" the specific consequences to be different than they actually are and misunderstands the casual relationships driving fundamental changes in the way the game is played.

Maybe I'm just being a bit too cranky. Maybe the whole point of the post was simply to plug HoopIdea, using up some All Star Weekend footage in the process. But then again, maybe there's some value in having a clear understanding of the evolution of the game.

Carry on.

Re: The only constant...

Posted: Wed Feb 27, 2013 7:33 pm
by Crow
On trends (but not the specific issues you raised above) I see that the league average offensive efficiency is up 0.9 per 100 possessions from last season but is still 1.7 pts below 2010-11. http://www.basketball-reference.com/lea ... stats.html So this season has only recovered about 1/3rd of the loss that occurred last season so far. Other than last season, it has not been this low since 2003-4. Still 2.8 pts below the recent high in 2008-9 (tied for all-time high).

Re: The only constant...

Posted: Wed Mar 13, 2013 4:44 pm
by schtevie
Ack! TrueHoop. Is. Killing. Me.

http://espn.go.com/blog/truehoop/post/_ ... -to-be-fun

OK. So TrueHoop likes to plump for "analytics". That's great. But for the love of John Naismith, please say things that are True and please tell coherent stories.

From paragraph 4: "...coaches have sought maximum efficiency since the day the league began." Screeeeeeeeeech! No, no, no! This can only be construed to be a true statement if one has the weakest possible interpretation of the word "sought".

It is not just the slooooow adoption of the 3 point shot (most recently referred to). It is also the multiple decades it took for coaches to gain the ginormous benefits of simply culling quick (i.e. bad) shots. And I would hazard the guess that any and all innovations not conspicuously introduced by championship teams (whatever these are or have been) have also seen decades of "seeking" before implementation.

What coaches have sought since the league began is job security. If this has happened to coincide with efficiency and wins, so much the better. Maybe things are different now in the brave, new world of analytics. I'm not holding my breath.

And then there is the weird George Karl narrative in service of the point of "analytics" also being fun. Setting aside the irony of a strategy of scoring dunks and lay-ups somehow being represented as an innovation owing to analytics (not dissimilar to the analytical innovation of recognizing that getting three points for a shot is more than getting two) there is the historical point that George Karl's adoption of this strategy is completely unrelated to the "analytics" revolution or Denver's investments in this area. (Indeed, he basically makes this point directly in the first quoted paragraph.)

Were 82games links of previous seasons not dead, we could go as far back as George Karl's last season with Milwaukee to check for continuity. But what is clear (using the links provided in the blog post to hoopdata.com and teamrankings.com) is that dating back to the coach's first year at Denver (i.e. even before he could possibly have come under the svengali-like influence and control of DeanO) Karl's style has always been to fastbreak and presumably also attack the basket (the hoopdata data on the latter point end at 2006-07).

So, when the written summary is "Translation: The analytics tell us the best way to play is in transition, and with maximum ball movement. That is, to give the fans what they want." never minding what the fans want, this basically is either besides the point or bass ackwards. It wasn't the "analytics" telling George Karl to play in transition at all; it was George Karl telling himself to do that, and that's because that's the way he had always done it. Perhaps that is because he has always "sought maximum efficiency" but in which case how do we construe his remarks early on in the cited interview in praise of Ty Lawson's mid-range game (this season's success rate, from 16-23 feet, courtesy of hoopdata: 37%, NBA average: 38.3%)?

Better stories, please.