Just Ask Nate

• Comment (2)
• Recommend (5)
• Tweet
• 

Nate Silver, the nerdy economist and New York Times blogger, apparently used big data algorithms to correctly project the presidential vote in all 50 states before a single ballot was cast. According to one report: Rather than reporting on nebulous aspects of a political race such as "feel" and "momentum," Silver's analysis relies solely on statistics.

Having seen Mr. Silver in action on “The Colbert Report,” I am not certain it would be any less painful to feature his predictions in place of the absurd pre-election analysis of say, a Fox or MSNBC News -- both of which traffic in unnecessary partisan hyperbole. But give math its due credit.

And now let's get the Nate Silvers of this world redeployed to more important tasks. Without question, the top priority is a better system to pick the nation's NCAA football national champion each year. Surely by now Nate has enough game stats from each BSC school to pick the top two prospects and save us from the other nearly 40 bowl games that are as pointless as they are often tedious to watch. Let's get it down to 10 teams playing just five bowl games, which spares players from having their seasons extended five or six weeks just to play in a bowl sponsored by a car part or a fast food item. As an aside, they should draw and quarter whoever decided six wins was enough to make a team "bowl-eligible." Move it up to eight or 10 and everyone saves a little face.

I don't personally know anybody who gets to cast a vote for the Academy Awards, but if there was ever a flawed system for deciding anything, this is it (nearly as stupid as the electoral college).  While nobody in their right mind would say to base winners on box-office success alone (since Hollywood panders to the lowest common audience IQ denominator), there should be a way for Nate to tap into audience reaction engines such as Flixster and decide winners based not on special-effect quantity, but script and acting quality.

I don't care who wins anything else having to do with entertainment, including Emmys, Country Music Awards, MTV Videos and Grammys, so as far as I am concerned Nate can take all that extra time and improve weather forecasts, or predict whether Brody will or won't be a terrorist.

Now, some in this industry spend a fair amount of time and effort developing "proprietary algorithms" to predict which ads audiences will best respond to. I suspect Nate could look at some basics -- like, is this a product or service anybody in his or her right mind would buy (something that should raise the barrier of entry considerably for most online ads), and if they would, is the creative approach somewhere beyond crotch jokes, animals and old ladies? Suddenly the pool shrinks like an Atlanta reservoir. Finally he can calculate who gets the attribution when the product sells. That ought to tie up his Excel for a few months alone.

One of the great new buzzwords in adland these days is "signals," which I suppose is more acceptable than data points, crumbs or user droppings. Surely Nate can amass enough "signals" from all of the time we spend online these days to predict pretty much anything. Why not tell a kid that he will die in the third level of “Meal of Honor” and save him a couple hours? Or that no matter how many times she goes back online or calls to confirm, there is nothing more the Family Travel Planner can do to assure adjoining seats on the flight, or earlier notice that they will cancel it altogether, anyway? Or, in spite of differently named sales every three-day weekend, that the best time to buy that car is, say, Aug. 14 or Jan. 8?

Or maybe Nate should put his money where his spreadsheet is, and take up picking stocks. Certainly more potentially lucrative and fun than picking Presidents.

• Comment (2)
• Recommend (5)
• Tweet
• 

2 comments on "Just Ask Nate".

1. John Grono from GAP Research
   

   commented on: November 9, 2012 at 7:28 p.m.
   You had me George until you suggested he tried picking stocks. There has to be some underlying patterns and 'sense' to make a prediction. It seems every night a company announces a new record multi-billion dollar profit and yet their stock price goes down - good luck predicting that!
2. Paula Lynn from Who Else Unlimited
   

   commented on: December 2, 2012 at 4:58 p.m.
   How about using his math and statistical skills to predict how budgets fall out in the immediate future and a few years down the pike ?

Leave a Comment