Saturday, October 15, 2016

The Senate Races
15 October 2016

The Senate races are polled much less often than the presidential election and there are fewer states I would consider true battlegrounds. The news was generally good for the Democrats' chances to re-take the Senate, returning the race to about the same shape it was two weeks ago.

Let's begin with the chart showing the odds for the number of GOP seats that will be held.
Instead of a two week look as usual, here's the last three week's numbers by category.

Probability of Republicans holding the Senate outright:
15 October: 41.2%
8 October: 53.8%
1 October: 43.8%

Probability of a 50-50 split: 
15 October: 31.0%
8 October: 29.2%
1 October: 30.7%

Probability of the Democrats taking the Senate back outright:
15 October: 28.8%
8 October:18.0%
1 October: 25.5%

Technically, a 50-50 Senate belongs to the party who has the vice president, and right now the presidential race is not much of a contest.

Here is the movement in the closest current races.

Nevada: Joe Heck still leads Catherine Cortez Masto in the race to replace Harry Reid, but this week one poll showed Cortez Masto ahead. This coincided with Heck withdrawing his support for Trump. Trump is likely to want revenge, but his first priority right now is proving all the women who have claimed he groped them are delusional or too ugly to be molested or both. We'll see if Trump can multi-task enough to hurt the GOP brand as much as he has hurt his own.

Current odds: 75%-25% GOP gain

Wisconsin: On the flip side, Wisconsin looked like a cakewalk for former Senator Russ Feingold to unseat Republican Ron Johnson, the race has tightened here and the median poll has a three point lead for the Democratic challenger.

Current odds: 85%-15% Democratic gain

Indiana: Challenger Evan Bayh has held the lead consistently, but the median poll says he only leads by three.

Current odds: 83%-16% Democratic gain

Missouri: My system uses the median poll and freshness dating. Any time a new poll is entered, any poll more than three weeks old is no longer considered. Early on, it looked like Roy Blount was a prohibitive favorite, but a mid month September poll showed challenger Jason Kander ahead. Now a mid-month October poll has Blount back in the lead in a close contest, all the older polls now being ignored. Even though I'm a libtard Commie, I assume the free market of the polling industry will be intrigued by something that looks like a close race and I will get more data to play with. So far in Missouri, the market has been disappointing.

Current odds: 66%-34% Republicans keep the seat

North Carolina: In these last three contests, the polling industry definitely treats these races as interesting. In North Carolina, ten polls have been taken in the last three weeks. Incumbent Republican Richard Burr has the lead in five polls, challenger Deborah Ross leads in three and two have an even split. While it has been going back and forth, Burr looks better in recent polls and my freshness dating factor could help in the weeks ahead as older polls fall out of the data set.

Current odds: 60%-40% Republicans keep the seat

New Hampshire: The race between GOP incumbent Kelly Ayotte and Democratic challenger Maggie Hassan has five polls in the past three weeks. Each candidate has two polls showing her as the leader and the median has a flat-footed tie. Freshness dating doesn't favor either candidate.  This really looks like a nail-biter.

Current odds: 50%-50% Republicans on defense

Pennsylvania:In the past ten polls, Democratic challenger Katie McGinty leads in five, two show a tie and GOP incumbent Pat Toomey leads in the other three. Freshness dating moving forward gives Toomey a chance to climb up.

On a personal note, I love this all-Irish contest. It reminds me of old school East Coast politics. 
Current odds: 61%-39% Republicans lose the seat


  1. I find it tough to believe that Marquette poll about Feingold. The same poll has HRC over Trump by 19 points, but we're supposed to believe there is16 point enthusiasm gap? no way that 16% of Wisconsin is voting third party or staying home. something hinky there....

    1. Unless a poll is the median of a set of polls, I don't take it very seriously.