For the past 18 years, LatestPollResults has offered an unbiased, statistically-based probability analysis of the Presidential, Senate, and House elections. First, we compute the probability for each candidate in each individual race. From these individual race odds, we calculate each party's overall odds of winning by evaluating every possible combination of outcomes. For 2018, we're offering three distinct analysis methods in developing the individual race odds:
Probability of Democratic Control
After the 2016 election, Republicans held a 52-48 edge over Democrats. Currently, Republicans hold a 51-49 edge. This year there are 35 Senate seats which are up for election, 26 will be defended by Democrats and 9 by Republicans; 42 Republicans and 23 Democrats are not up for re-election. With 35 seats, there are 2^35, over 34 billion, possible combinations of outcomes. As you can see from the above probabilities, at this time, the odds that the Democrats will win enough seats to take control of the Senate are low.
Since 2010, we've been analyzing the Senate by using data from the latest poll for each race to calculate each candidate's probability of winning the race.
With these individual race probabilities, we calculate the odds for every possible combination of outcomes and each party's overall probability of winning.
Our poll-based analysis has correctly pointed to the winning party in 2010, 2012, 2014 and 2016.
Here is the current tracking for our poll-based analysis:
In 2016, we began market-based analysis of the Electoral College and the Senate using data from individual contract pricing at the Predictit.org marketplace, to calculate each candidate's probability of winning.
From these individual race probabilities, we calculate the odds for every possible combination of outcomes and each party's overall probability of winning.
For 2016, the market-based odds on election day pointed to a 65.8% chance the Democrats would control the Senate, with the help of a 50-50 tie based on the 87.6% market probability they would win the Presidency.
The Democrats had a 41.2% chance of outright control (51+ seats.)
Obviously, the market-based method missed the mark for the Presidency and the Senate.
Here is the current tracking for our market-based analysis:
For 2018, we're beginning our analyst-based analysis using the analyst ratings from
The Cook Political Report,
Sabato's Crystal Ball, and
Real Clear Politics.
Each offers a qualitative rating for every seat in the Senate and House and in each state during a Presidential election year.
Seats are rated either a "Tossup" or "Safe," "Likely," "Lean," or "Tilt" (Inside Elections only) for the Democrat or for the Republican.
For each analyst, using their final ratings for 2010, 2012, 2014, and 2016,
we evaluated the overall percentage accuracy for each rating for both the Senate and House and use this percentage as the probability for each analyst's current ratings.
With these individual race probabilities, we calculate the odds for every possible combination of outcomes and each party's overall probability of winning control of the Senate as well as the House.
We've been tracking each analyst's probabilities on a daily basis since late March:
After the 2016 election, Republicans held a 241-194 edge over Democrats. Currently, Republicans hold a 237-193 edge over Democrats. There are 5 vacancies: 3 Republican and 2 Democratic. With all 435 seats up, there are 2^435 possible combinations of outcomes - over 88 duoquadragintillion, a number that's 131 digits long and dwarfs the number of atoms in the universe! Since not every competitive district is being polled, we're unable to use our poll-based method for analyzing the House. Similarly, not every competitive district has market pricing available, so we're unable to use our market-based method either. We're analyzing the House using our new analyst-based method which converts each analyst's qualitative rating to their historical probability. From the individual probabilities for each seat's rating, we calculate each party's implied probability of controlling the House based on the current ratings. We've been tracking each analyst's probability on a daily basis since late March:
Note: To be clear, the above shows our calculation of the Democratic probability for controlling the House, calculated for each analyst, using their historical percentage accuracy for each qualitative rating as the probability of that rating. These results do not necessarily reflect the analyst's current or future projections. Here are recent quotes from the analysts:
Inside Elections - 9/14: "Our current outlook is a Democratic gain of 22-32 seats."
Sabato - 9/20: "...our best guess right now is a Democratic House gain of somewhere in the low-to-mid 30s."