Senate

35 seats are up. 33 in class 2, and two class 3. Unlike the 2018 elections, just about every race here will have a connection to the national stage. Candidate still matters, but local branding will be slightly harder in deep red/blue states.

The huge swing in this class was in 2014, when Republicans picked up a net of 9.

This cycle has been wild so far. Senate races can swing at a moments notice. Sometimes a scandal wrecks a candidate's chances, and sometimes, it does almost nothing.

Final Ratings

The GOP is favored (67.1%) to win the US Senate

High Confidence

Complete Predictions:

Final Results (note: look to the GA runoff page for the GA runoff results, this shows round 1 results):

I held my own from 2018, only getting one race wrong (so far). I already mentioned the thing about GA, so let's discuss the round 1 results. The class 3 result is exactly where I expected it, but Perdue underperformed by about 2-3% of what I expected. The class 3 seat is more in line with the House results, while class 2 was more like the presidential race.

I knew Susan Collins was going to overperform the polls, as she is quite well known for doing that, and she outperformed my expectations by around 6%. No RCV necessary for any election in Maine this time round.

Michigan I knew was going to be a tight race. Now, on my runoff page I still haven't declared him the winner, largely because the suspect activity in the state has a small but not insignificant chance that James could still win (this does not apply to Trump). It is a small chance. Peters won by about double the margin I expected James to win by (I had James +1, and Peters won by 1.8)

Aside from that I pretty much called everything perfectly and nailed a majority of the margins.