Clarion Content

The Clarion Content is a dedicated and loyal follower of Nate Silver's statistical analysis of elections in his blog the 538.¹ Long time readers find great irony in this, because as a baseball analyst the Clarion Content's editor is likely to be the first one jumping up and down screaming if Mike Trout somehow steals Miguel Cabrerra's Most Valuable Player Award. Silver, the creator of PECOTA, is renown in baseball circles as a statistical guru only slightly below Bill James on pantheon.

Statistics aren't everything. Math isn't religion. It is a language, descriptive, but limited by its own conceits and structures.²

The problem is that the average moron in the media doesn't speak math. In 2008, Silver correctly predicted the Obama vs. McCain results in 49 of 50 states. He predicted every Senate race correctly. In 2010, he did pretty well too, nailing 36 out of 37 gubernatorial races, although missing on some individual House and Senate races.

This year, using his rigorous statistical analysis and as many polls as he can lay hands on, he has been predicting an Obama victory for some time. His model did show Romney making gains after the first debate, but it was from getting beat handily in the Electoral College to getting beat barely. In the weeks since, the model has drifted back up to an 86% chance of Obama winning, with the two most likely outcomes being the President gets 300 electoral votes, rated a 14% chance, and the President gets 330 electoral votes, rated a 17% chance.

Rocket scientists, like MSNBC host Joe Scarborough, laughably conflate polling and statistics to understand the President and Romney each have a 50% chance of winning. Ah, the dim.³

In fact, Silver predicts that Obama will win only 50.6% of the popular vote. But as those of you who are old enough to remember the 2000 Presidential Election know, the popular vote is not the deciding factor for presidential elections in the good ol' U.S. of A. We have this rigamorole known as the Electoral College.⁴

Silver has done the legwork and the math. He is convinced Obama is going to win. One reason he cited as an example on the 538, in twenty-two polls of swing states Friday, not national polls, individual state polls, Obama led in nineteen, two showed a tie and Romney led in one. This is not what one would call a dead heat, regardless of the national aggregate numbers.

Silver is not alone in his quantitative analysis, Deadspin notes, "the criminally underrated Princeton Election Consortium... [which] nailed the electoral vote count in 2004, and missed it in 2008 by just one... [forecasts] Obama's re-election chances at 98%."

Vote! The Clarion Content's article is not a reason, not to vote. It is about math and bias. Please, please, please go vote! North Carolina is going to be razor close for the distribution of its electoral college votes. There are also a bevy of important state races which will effect everything from schools and state taxes, to energy policy and the environment.⁵


Notes
¹The 538 sadly now belongs to the New York Times, meaning we can read it no more than ten times per month.
²Show us past the Godel theorem, "For every consistent formalization of arithmetic, there exist arithmetic truths that are not provable within that formal system." And furthermore, almost all statisticians concede, small sample sizes can yield lots of surprising results, often called "noise." See the Central Limit Theorem.
³Deadspin notes in its article about Silver, just like math class in high school, some of the bullies use Silver's ostensible nerdiness and high-pitched voice to accuse him of one of their greatest sins, homosexuality. Of course, if he is smart and effeminate, he must be queer. Ahhhh, the sophistication of the narrative that comes from certain parts of the Right Wing.
⁴Interestingly, one might wonder if this will bring the Republicans behind a proportional distribution of states' electoral college votes, rather the winner take all model used everywhere but Maine and Nebraska.
⁴The era of fracking in North Carolina is about to begin.

Labels: 2008 presidential election, 2012 presidential election, Politics, Predictions