The New World of Election Coverage and Analysis: A Case Study

by Stuart Rothenberg May 15, 2014 · 10:35 AM EDT

My new statistical model of the open Wisconsin Senate seat suggests that Democrats now have only a 54.496 percent chance of holding the seat. That’s a dramatic change from just three weeks ago, when my model showed them with a 55.501 percent chance.

The change results from three main developments: (1) changes in the national generic ballot that are likely to filter down the ballot, (2) changes in my turnout model, specifically among voters with Scandinavian surnames, and (3) the unexpected development that Pisces has entered the House of Scorpio, indicating an increasing sexual energy that should benefit Republicans, who have had a long-term advantage with macho male voters.

In addition, multiple public polls conducted by high school students now suggest that the state, which went for Republicans Dwight Eisenhower (in 1952 and 1956) and Richard M. Nixon (in 1960), but flipped in 1964 to support Democrat Lyndon Johnson, could possibly be poised to either flip again or not flip again, depending on the meaning of the word flip.

One poll that was conducted by a couple of Republicans from Green Bay found that 102 percent of Republicans who attend church regularly plan to vote for the Republican nominee, assuming of course that Republicans have a nominee. At the moment they don’t.

But a survey conducted by a non-partisan group, Generic Women with an Opaque Agenda, found that unmarried women will constitute upwards of three-quarters of the midterm electorate, giving Democrats an “excellent chance” of winning the Wisconsin Senate race if the party nominates someone “to the left of Senator Tammy Baldwin.”

The GOP’s difficulty finding a candidate could pose a problem for the party, speculated a veteran political scientist who suddenly started an “institute” when his university president concluded that that could be the only way to reverse the decline in applications from high school seniors.

The Dane County Daily Bugle immediately hired the political scientist, Professor Wilmot Proviso, to write a blog, which will include ratings of all congressional, state and local races in the country, including city council races, state legislative contests, freeholder races and justice of the peace contests.

The newspaper’s editor, Hector Lopez, told Fishbowl DC that although he wasn’t sure whether any of the institute’s race ratings were accurate or even reasonable, that probably wasn’t all that important since the newspaper’s website would have “some really neat maps, with lots of colors, and plenty of statistics to convince people that we are really smart. Plus we are going to market the crap out of it — just wait ’til you see our headlines — so it really doesn’t matter whether our ratings are thoughtful anyway.”

So far, the 38 most likely Republican candidates in the Wisconsin Senate race have all declined to run for the open seat, with each offering a different explanation why he or she will pass.

County Commissioner Shmini Atseres said that his children are too young, while state Rep. Sue Ann Nivens said that she is spending all of her time trying to apply for health care on the government website and simply doesn’t have the time to run.

National Republican strategists blamed the party’s recruitment problems on “the cold weather,” while tea party groups blamed RNC Chairman Reince Priebus, and American Crossroads blamed Sarah Palin.

While Democratic prospects in the Wisconsin Senate race are fading, there is at least a 44 percent chance that Democratic prospects will improve over the next few months, because of the expected increase in sunspot activity over the summer.

Extensive research by people with doctorates has noted a strong statistical relationship between solar flares and higher turnout among Democratic voter groups — see the very excellent Washington Post piece, in the Style section, by Dr. Nathaniel Gold, “Sunspots, Voter Turnout and Midterm Elections in Odd-Numbered Years.”

Gold’s model uses regression analysis, Yule’s Q, analysis of covariance and Pearson’s chi-squared test, along with a dummy variable and three or four hard boiled eggs to predict how Wisconsin voters will vote. He then coats everything with a light batter before cooking it for 35 minutes at 350 degrees. Only then does Dr. Gold apply his secret formula “special sauce” and make his prediction public.

The Rothenberg Political Report/Roll Call rates the Wisconsin Senate race “We Haven’t a Clue.”

Oh, and by the way, Twitter can predict elections. I read it in the Washington Post and saw it on MSNBC last August, so it must be true, right?