Avoiding Indeterminacy (Predicting the Past)

February 15, 2012
Posted by Jay Livingston

One of the most difficult ideas to grasp or accept is indeterminacy, randomness.  We devote a considerable effort to making up stories to show that nothing is random, that everything is, or was, predictable. 

“We should have seen Lin coming” was the headline in Carl Bialik’s Wall Street Journal blog post (here) on Monday. 

Jeremy Lin – is there anybody who has not heard about “Linsanity”? – a guard for the Knicks, is the NBA’s new great white hope, though he’s Asian (his parents are from Taiwan).  Single-handedly he is bringing the NBA a new demo.  Even when the Knicks are on the road, the fans – in Washington, in Toronto, wherever – cheer for Lin.

Like one of those huge best-sellers by an unknown author that 39 publishers turned down before one took a chance, Lin’s talents might never have seen the light of NBA.  He played well at Harvard but in the NBA draft received not a nibble.   Eventually, he was picked up by the Knicks, but they too had no idea that Lin was star material.  Coach D’Antoni didn’t give him much court time until a game earlier two weeks ago when Lin went in to give Bibby a breather.  He scored 25 points, and since then, he has been headline news.

How did everyone miss him?  Carl Bialik is the WSJl’s “Numbers Guy,” and he puts together some numbers to show that Lin’s abilities were clear from the start.  Numbers like this:

Per 40 minutes this season, he’s taken 7.8 shots at the rim and made five of them. That’s the second-most made field goals from the rim for guards who’ve played at least 10 games and at least 10 minutes per game, and a percentage in line with the impressive Nos. 1 and 3 on the list . . .
If you thought a rim shot was something that followed a lame joke in a burlesque house, you might not find this convincing.  But Bialik has more such numbers, and he makes the case. 

Still, it reminded me of days at the horse track.  The Racing Form provides a wealth of information, mostly quantitative, on each horse in the race – the horse’s past performances.*   Horseplayers process all this data and make a bet.  Then, after the race, as they tear up their losing tickets, they go back to the past performances, and no matter which horse won, they can always find the bits of data that made it clear why that horse was bound to win. 

Prediction is very hard, especially about the future, as Yogi Berra or someone said.  Prediction about the past and present is much easier, as Bialik’s blogpost illustrates.  Or as Duncan Watts puts it in the title of his excellent book, Everything is Obvious . . . Once You Know the Answer.** 

There’s another reason the Knicks didn’t know how good Lin was.  Here’s Knick announcer Clemson Smith-Muñiz (la voz en Español de los Knicks):

I’ve asked the coaching staff the question this way: didn't you see this in practice? And the answer has been, invariably, “What practice?” Due to this condensed season, which included barely 4 weeks of pre-season, all teams are limiting their practices, especially the full-court scrimmages, on off days.
Their point, again, is that the Lin phenomenon was not indeterminate.  Given a chance to see Lin in practice, any good coach would have seen his abilities.

Of course, Lin might turn out to be a flash in the pan.  Maybe by his second time around the league, the other teams will have learned how to play him.  It hasn’t happened yet (last night, his three-pointer with less than a second on the clock won the game), but if it does, sports writers, maybe even Carl Bialik, will write columns and blog posts saying that his short-lived success was utterly predictable.

Linsanity is fine, Lindeterminacy is intolerable.


* When non-horseplayers appropriated the term “track record,” they distorted its meaning, just as Bialik does in his blogpost (I did not bother to quote it).  What they are referring to is what horseplayers know as “past performances.” 

Not only does the popular meaning of “track record” have nothing to do with its meaning at the race track, but in most cases, the speaker or writer could drop the “track” without changing the meaning, except perhaps to make it clearer. 

** Another blogpost on Watts ideas and horse races is here.

No comments: