Vosem's Senate Prediction

[Vosem]

This will be updated semi-regularly until the actual election (I hope). No details; just a map.

Republicans gain 5 seats, Democrats gain 1. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. Republicans take the Senate, 51-49.

Obviously this map will change dramatically by Election Day.

[Vosem]

This will be updated semi-regularly until the actual election (I hope). No details; just a map.

Republicans gain 5 seats, Democrats gain 1. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. Republicans take the Senate, 51-49.

Obviously this map will change dramatically by Election Day.

Dems flipping NV but not MA seems unlikely.  Everything else looks very reasonable.  It's fascinating that the Senate is polling at essentially an exact tie right now. 

The three most recent firms to poll Massachusetts are Western NE College, UMass, and PPP. Western NE College found Brown ahead by a 47-42 spread; UMass found Brown ahead by a 41-38 spread; and PPP found Warren ahead by a 46-44 spread. This averages out to Brown leading by a 44-42 spread.

Only two firms have polled Nevada so far - PPP and Magellan Strategies. PPP found an exact tie, 45-45, and Magellan Strategies found Berkley ahead by a 46-44 margin. This averages to a Berkley lead 45.5-44.5; after rounding, 46-45.

I'm using RealClearPolitics as the source of all my polling; if I leave anything out, it's their fault, not mine.

[Vosem]

Scott Brown (R-MA) average poll numbers against Elizabeth Warren (D-MA) is at 44% while Warren is at 42%. That leaves the 14% undecided voters. Brown gets 1/3 of the undecided voters. Warren gets 1/2 of the undecided voters. Warren-D wins by a 49-48,50-48,51-48 margin. The undecided voters are unlikely going to vote for Brown but they are waiting to see how Warren performs until they commit to her.

Just out of curiosity, how come the undecideds are so magically likely to vote for Elizabeth Warren?

This will be updated semi-regularly until the actual election (I hope). No details; just a map.

Republicans gain 5 seats, Democrats gain 1. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. Republicans take the Senate, 51-49.

Obviously this map will change dramatically by Election Day.

Dems flipping NV but not MA seems unlikely.  Everything else looks very reasonable.  It's fascinating that the Senate is polling at essentially an exact tie right now. 

The three most recent firms to poll Massachusetts are Western NE College, UMass, and PPP. Western NE College found Brown ahead by a 47-42 spread; UMass found Brown ahead by a 41-38 spread; and PPP found Warren ahead by a 46-44 spread. This averages out to Brown leading by a 44-42 spread.

Only two firms have polled Nevada so far - PPP and Magellan Strategies. PPP found an exact tie, 45-45, and Magellan Strategies found Berkley ahead by a 46-44 margin. This averages to a Berkley lead 45.5-44.5; after rounding, 46-45.

I'm using RealClearPolitics as the source of all my polling; if I leave anything out, it's their fault, not mine.

The latest UMass poll found Warren ahead of Brown in MA by 4, 43-39. Here's the link:

 http://www.umass.edu/newsoffice/images/upload/UMass%20Poll%20detailed%20results_0.pdf

Thanks, but I'm using RCP as my poll source, and I would rather keep all states to the same standard, so I won't incorporate that poll. For your curiosity, if that poll was included, Brown would actually win by the slightly larger margin 44.67-42.33, which rounds to a 45-42 Brown victory; slightly larger, and with slightly less undecideds.

[Vosem]

Forgot about this to some extent.

Ironically enough, the only race to have a different leader since my last update was the one we argued about most; Elizabeth Warren has taken the lead in Massachusetts, 46-44, which seems to lead credence to nkpatel's theory that the undecideds tilt Warren. Still, 2 percentage points is within the margin of error.

Here's your map, with one change.

Republicans gain 5 seats, Democrats gain 2. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. The Senate is tied, 50-50.

Obviously this map will change dramatically by Election Day.

[Vosem]

Bumping this. I thought the recent Suffolk poll might shift Massachusetts back to Brown, but three most recent polls from different companies averages out to a narrow 45 1/3 - 45 lead for Warren, so the map in the previous post is still my prediction.

[Vosem]

I failed to note the CNU/Times-Dispatch poll giving Allen a 2-point lead, but it switches Virginia from a Democratic hold to a Republican pickup, with Allen at 44% and Kaine 43.66...%.

Republicans gain 6 seats, Democrats gain 2. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. The Senate has a Republican majority, 51-49.

Obviously this map will change dramatically by Election Day.

[Vosem]

Since I'm using RCP rankings in the event of no polling existing, Senator Snowe's retirement shifts Maine from the Republican to the Democratic column, and also brings the 'total' from a 51-49 Republican majority to a 50-50 tie (most likely, with Vice President Joseph Biden breaking the tie in favor of the Democrats).

Republicans gain 6 seats, Democrats gain 3. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. The Senate is tied, 50-50.

Obviously this map will change dramatically by Election Day.

[Vosem]

The three most recent polls from different companies in Massachusetts are a Rasmussen Reports poll showing Scott Brown leading Elizabeth Warren 49-44; a Suffolk/7News poll showing Scott Brown leading Elizabeth Warren 49-40; and a WBUR/MassInc poll showing Elizabeth Warren leading Scott Brown 46-43. These polls average out to a 47-43 lead for Scott Brown, flipping Massachusetts from Democratic pickup to Republican hold.

Republicans gain 6 seats, Democrats gain 2. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. The Senate has a Republican majority, 51-49.

Obviously this map will change dramatically by Election Day.
[/quote]

[Vosem]

If the Senate is tied 50-50 or if the Republicans take control, maybe they'll  be able to convince Joe Manchin to switch parties.  You never know what could happen.

I rather doubt something like that could occur - West Virginia is an ancestrally Democratic state, and some West Virginia Democrats are easily to Senator Manchin's right, such as Mike Oliveiro. Even if it does, the point of this model is to forecast the election, not the election's repercussions.

I see your prediction is rather more Democratic than my model's prediction, and you also seem to be assuming several things that are likely, but not certain, such a strong independent candidate in Maine.

--

In any case, the new Virginia poll from Marist shows a 9-point lead for Tim Kaine. This means that the three most recent polls from different companies are the aforementioned 48-39 Kaine lead from Marist; a Rasmussen Reports poll showing both candidates tied at 46-46; and a Roanoke College poll showing Allen leading 45-37. This averages out to a 43-42 lead for Kaine, which flips Virginia back into the Democratic column.

Republicans gain 5 seats, Democrats gain 2. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

The two independents are counted as Democrats for the purpose of this exercise. The Senate is tied, 50-50.

Obviously this map will change dramatically by Election Day.

[Vosem]

First off, although there are still no polls included, RCP has shifted Maine from 'Leans Democratic' to 'Tossup', which makes it essentially a conscience call; I say Angus King is the likeliest winner, and since he has failed to reveal what party he will caucus with, Maine is moved to 'Independent Pickup'.

The other shift is in Nevada, where Rasmussen Reports has released a poll showing Senator Dean Heller ahead of Democrat Shelley Berkeley, 47-40. This means the three most recent polls from different groups are Rasmussen Reports, which show a 47-40 Heller lead; the Las Vegas Review-Journal, which show a 44-43 Berkeley lead; and PPP, which showed a 45-45 tie. This averages out to a Heller lead of 45-43, shifting Nevada from 'Democratic Pickup' to 'Republican Hold.'

Republicans gain 5 seats, independents gain 1 (from the Republicans). Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

Joe Lieberman and Bernie Sanders are counted as Democrats for the purpose of this exercise, but since he has not revealed who he will caucus with, Angus King is counted as an independent. The Senate has a Republican majority, 51-48-1.

Obviously this map will change dramatically by Election Day.

[Vosem]

I've somewhat neglected this thread over the past several weeks; during this time, Virginia and Massachusetts have flipped.

Republicans gain 6 seats, independents gain 1 (from the Republicans), and Democrats gain 1. Should be obvious. I'm going with a mathematical model -- averaging the last 3 polls from different companies. If PPP's conducted 4 polls, I just take the most recent one, no averaging. If there's a single different one from a few months ago, it gets included. I get my polling from RCP; if a poll isn't included, ask them.

Joe Lieberman and Bernie Sanders are counted as Democrats for the purpose of this exercise, but since he has not revealed who he will caucus with, Angus King is counted as an independent. The Senate has a Republican majority, 51-48-1.

Obviously this map will change dramatically by Election Day.