Today, we have five new polls covering four states since the last update. These new polls – even though they didn't change the lead of any state – have moved the probable outcome to a tie (or ever so slightly in the favor of Republicans).
The probability that the Democrats will control the Senate after the midterm elections has fallen to 49.9%. This percentage, however, is within the margin or error for the SanityIID model.
Alternately, there is a different way to look at the output from the model. Instead of calculating probabilities, the model also outputs the expected number of seats won by each party. It does this by assuming that the polls will be their eventual outcome instead of assigning them probabilities.
Using this method, we can see that the Republicans continue to gain ground:
Using this data, the Republicans are leading in the short, medium, and long term outlooks for the model. Even the states that are considered ties in the medium and long outlook are expected to go to the Republicans in the short term.
This is exactly what we would expect to see if the Republicans are indeed gaining ground in their effort to take control of the Senate.
As for the individual polls, a new NBC/Marist poll was released for Iowa. The poll put Ernst (R) in front of Braley (D) by 46 to 44. This, in addition to how well Ernst has been doing over the past several polls, has increased the Republican chance of taking Iowa from 60.6% to 71%
Kansas received two new polls since the last time I updated: NBC/Marist and CBS/NYT/YouGov. These polls were for Roberts (R) 38 and 40% respectively, and Orman (I) 48 and 40% respectively. Because of the one tie poll, the SanityIID model increased the chances of Republicans taking the state from 17% to 19%. Keep in mind, however, that the model is still warning that there isn't enough data to make an accurate prediction. In addition, Roberts is just starting to get his campaign machine geared toward Orman as a challenger.
The latest poll for Michigan was from Lake Research. It has Land (R) with 36% of the vote and Peters (D) with 45% of the vote. The SanityIID model essentially stayed the same at a 2% chance that the Republicans will win Michigan.
Finally NBC/Marist released a poll for North Carolina. This poll has Tillis (R) at 40% and Hagan at 44%. Essentially this changed the prediction from Republicans having a 2% chance to have no chance in North Carolina. Or course, since the model's margin of error is greater than this, the chance will be held at 5% chance for Republicans.