Silver says: Biden's Band-Aid; Obama's Opportunity

In my last post, "Pivoting from debacle to debating The Real Romney" **, I wrote that Joe Biden's piece of work on October 11th did succeed, as I'd hoped, in stopping the bleeding. 

To wit: 

"However, the one result I do feel confident about sustaining is that the prevailing winds of Romney momentum from the first debate has officially become yesterday's news."  [Ital. added]

Leaving aside the egregious subject-verb conflict, I felt that Biden had, by dint of personality and a cascade of facts upon facts, imposed a new set of images on the media campaign coverage.  The liberal-leaning commentators focused on Biden's hammering home the facts that Obama had disdained to mention, and the right-siders focused on Joe's historically, monumentally white teeth.  

In fact, Fox News and the other Repubican spokesplatforms were forced to sidestep recall of the Obama-Romney debate in order to press the indictment of Biden for behaving the same aggressive way as Obama supporters perceived Romney to have acted -- sans the grin, yes, but Romney was no less "rude".  

So, did independent polling feedback reflect that Biden's performance had stanched the loss of blood?  Nate Silver wrote this in his Monday column:

"National polls showed a modestly favorable trend for President Obama, allowing him to gain slightly in our forecast. (Mr. Obama’s chances of winning the Electoral College are now 66.0 percent, according to the FiveThirtyEight model, up from 63.4 percent on Sunday.)"

This is potentially a critical reversal of public opinion.  In his blog on Friday, Silver had written that Romney's broaching the 40% probability level would mean that the race was a "toss-up".  Romney had gotten to 38.9% on Friday -- 1.1% from toss-up city! -- which was well-nigh apocalyptic given that Silver's model had accorded Obama an 85% likelihood of winning the Electoral College eight days earlier.   

So what all this means is that the Silver model's predictive percentage for a Romney win has receded, increment by increment each day since Friday, from that 38.9% high water mark.  By Monday it was back down to 34.0%.

This turnaround trend is consistent with the Biden-to-the-rescue hypothesis. 

However, it's still a 20-point percentage drop for Obama, and the race still hangs in a delicate balance.  Indeed, that's precisely what Silver concludes.  After the quote I pasted above, he cautions the reader not to see the Obama recovery as a robust trend.  Biden's bump was really just a Band-Aid, not a cure.

Silver goes on from that caution to render a peek into the mind of an obsessive but very intellectually honest and ethical nerd. The rest of the column is not about the election but rather about the ethics of how one goes about doing poll meta-analysis and interpretation. Which is what Silver does, of course -- he is not a pollster.

In a post-truth, all-spin-all-the-time Orwellian political world where people make livings trying to tell others what is real (I try, but I no get paid), and in which too many of them are selling lies as reality, Silver is one of the last bastions of disinterested objectivity. He is a lighthouse in the dark of night.

Which is all well and good, but in the end we just want to see our stinkin' numbers going up and the other side's going down!

And that is where tonight comes in. 

Obama's chances of consolidating this fledgling recovery and getting a bit of daylight back between himself and the conviction-free, heartless, soulless cadaver -- i.e., the esteemed Governor Romney -- depend on Obama's showing up very PRESIDENTIAL-like at the debate on Long Island. 

Now, based on my unscientific chats with people as I canvassed for the Obama campaign in my neighborhood, I suspect that the majority of Obama's sharp sag in the polls was due to his own poor performance more than it was due to Mitt's good one.  The main person Obama has to "beat" tonight, then, is not Mitt as much as it's the first debate Obama.  The bar is low for him, but so are the expectations that can shape perceptions.

And finally, dipping into the FiveThirtyEight Blog well one last time, I looked to see today's updated percentages before posting.  It was not out yet, but in today's post he offers a classically Nate Silveresque labyrinthine historical review of presidential debate data to assess Obama's statistical chances of a second-debate bounce back.  After hiking us up a data trail with more back-and-forth switchbacks than Walter's Wiggles itself (on the way up to Angels Landing in Zion NP), Silver favors the bounce back by a hair's-width preponderance of the statistical trends. 

Hoping for the best.

** If you didn't read the L&LO October 12 post-Biden-Ryan post, "Pivoting from debacle to debating The Real Romney", please take a moment if you can to check it out.  It recounts an investigative report on Romney's imperious bullying of pregnant women as a Mormon church leader in the 80's. The report reveals the character of the Real Romney as candidly as the Boca Raton video revealed his social-economic world-view.