The House Next Week? 219/218 Republican
During basic combat training as an enlisted soldier, you crawl under barbed wire while a machine gun is firing rounds over your head. During ROTC Advanced Camp you…….hmm….drink a lot of beer….drink a lot of cheap vodka…..meet co-eds from other schools……OK, lets skip that one. During the Infantry Officer Basic Course you walk off a diving board, blindfolded, with weapon and backpack and need to get to the edge of the pool without panic. In life, I have called my “friends” in the Field Artillery on two occasions for “help.” Both times, I closed with “Danger Close.” The first time, I wondered if engaging the Taliban might have been a better idea. The second time, I knew better and trusted my friends in the Field Artillery, I should have known better the first time. You run the numbers, check them twice, then fire……
In my civilian profession as an actuary, I make financial “predictions” based on the probability of a whole series of events happening. They are not limited to a single event but rather a multiple of events. This has sparked an intellectual curiosity as to what may happen with 435 House races next week on the whole. This differs from pollsters/pundits who make a “bet” on each single race. I am only concerned with Party control in the aggregate.
As a result, I have burnt up too many hours looking at prior pollster predictions and how often their predictions actually occur based on certain classifications. What I have learned is that when pollsters and/or pundits provide a list of “x” number of incumbent seats and classifies them as “likely takeover” or “leans takeover” that not all of those seats will actually “lean” or be “taken over” on election day.
Probability tells me that when there are a list of “likely” or “leans” takeover seats that it is “possible” that all of the seats will be lost or all of them will be retained.
Using Novak, Election Projection and CQ Quarterly as a guide to eliminate “for sure retentions” I was able to apply “past history” success rates of pollsters and pundits.
Though I can not predict individual races, I am able to determine most likely scenarios. Using “past history” as a guide to “future performance” the next House will most likely consist of 219/218 Republican members.
Are there other scenarios? Of course!! The Republicans could retain every “at risk” seat and they could lose every at risk seat!! However, both of those scenarios have a very low probability. The most likely scenario is 219/218 Republican until we figure out how to “dice up” a Republican House Member to get a fraction. 
P.S. I have a distinct added advantage in observing this election. I am “observing” it solely online and that helps remove the “emotion.”
Subscribe to blog feed.