Our History

LatestPollResults has been projecting the presidential election since 2000. We present an unbiased statistically-based probability analysis of the Electoral College based on each candidate's probability of winning in each of the 50 states and D.C. In 2010, we adapted our probability model to analyze the Senate. We post daily updates and analysis during the months leading up to Election Day.

Poll-Based Probabilities

After each party has had their national convention and the candidates have been selected, we begin the process of monitoring and compiling the latest poll in each state. We prefer polls of "Likely Voters" where the pollster has made a determination that there is a reasonable likelihood that the respondent will vote in the upcoming election. We will include polls of a more general sample of "Registered Voters" in a given state, if we haven't had a "Likely Voter" poll within the last month. We use polls reported by Real Clear Politics or HuffPost Pollster. Note, we disregard polls which are strictly internet polls.

With each state, based on the reported sample size and polling percentages for the candidates, we determine the mathematical probability of winning the state for each candidate or party.

The Overall Probability of Winning

Once we've calculated the state-level odds, we determine the overall odds of winning by calculating the probability of each possible electoral vote combination. For example, in the presidential election, we calculate the probability for each of the 539 possible electoral vote combinations: 0D-538R, 1D-537R, ..., 269D-269R, ..., 537D-1R, 538D-0R. We then calculate the sum of the probabilities for the combinations where each candidate wins. In the presidential election, the Republican's overall probability of winning would equal the sum of the probabilities 0D-538R, 1D-537R, ..., 268D-270R. The Democrat's over probability of winning would equal the sum of the probabilities 270D-268R, 271D-267R, ..., 538D-0R. In a 269-269 tie, since the House selects the President, the odds of a 269-269 tie are prorated according to the odds for each party to control the House.

The popular vote winner in each state captures all of the state's electoral vote, except in Maine and Nebraska where the winner of the popular vote in the state wins 2 electoral votes. In addition, the winner of each congressional district is awarded an electoral vote. With 50 states plus D.C., and the competitive 2nd districts in Maine and Nebraska, there are over 9 quadrillion possible outcomes, 2^53.

If it takes one second to process 20 states with over 1 million (2^20) possible outcomes, it will take 2 seconds to process 21 states, since there are twice as many outcomes than with 20 states. For 22 states it would take 4 seconds and so on. For 53 states/districts, it would take over 272 years!

In previous years, if a candidate's state-level probability was considered statistically significant, 97.5% or higher, we considered the state to be "safe" for that candidate. We assigned the electoral votes for each candidate's "safe" states and analyzed the remaining "battleground" states. In this way, computational time was drastically reduced. In the 4 presidential elections and 3 Senate elections we've projected, we've determined that over 200 contests were "safe." Fortunately, each contest did turn out to be "safe."

Our Track Record

We correctly projected Bush to win in 2000 and 2004 and Obama to win in 2008 and 2012 . In 2010, we expanded our analysis to the Senate races and we projected that the Democrats would win control of the Senate in 2010 and 2012 and the Republicans in 2014.

Market-Based Probabilities

www.PredictIt.org has emerged as the leading prediction marketplace for real-money politically-based contracts. The U.S Commodity Future's Trading Commission (CFTC) has issued a no-action letter for PredictIt.org to operate its marketplace and U.S. traders can legally buy and sell politically-based contracts based on an event occurring ... or not occurring.

Currently, for the Presidential election, PredictIt.org offers market contracts for all 50 states, DC, and the competitive 2nd congressional districts of Maine and Nebraska. Additionally, they offer 16 contracts for the Senate, and 18 for the House. Traders buy or sell contracts based on which party they believe will win in a particular state or district.

Each contract trades in cent intervals between 1 cent and 99 cents and has a specific expiration date. Traders can buy or sell at the market price, any time until the contract expires. At expiration, the value of the contract will be settled at $1.00 if the event occurs and $0.00 if it doesn't. If the transaction is profitable, PredictIt deducts a 10% commission from your winnings. From the contract pricing, we derive the implied market-based probability for each candidate. Assume that a trader purchases a contract for the Democrats to win Florida and that (s)he pays 60 cents for the contract. If the Democrats lose Florida, the value of the contract will be $0.00 and the trader will have lost 60 cents. If the Democrats wins Florida, the contract will settle at $1.00 and the trader will win 40 cents per contract, less 4 cents for the 10% commission. The net winnings will be 36 cents. The implied market-based probability of risking 60 cents to win 36 cents is 62.5%.

To obtain the most accurate probability for an event, we evaluate the market pricing for: Buy Yes for the Democrat, Buy No for the Republican, Buy Yes for the Republican and Buy No for the Democrat. For each state or district, we determine each candidate's market-based odds of winning.

Real-Time Calculations

For 2016, we are introducing our new powerful probability engine. Even though there are over 9 quadrillion possible outcomes, using each candidate's probability of winning in each state and the state's electoral value, our proprietary probability engine instantly calculates the probability of every possible electoral vote combination and the overall probability of winning for each candidate.

What-If Analysis

The Wizard Of Odds ℠ is the Ultimate Interactive Election Analysis Tool!

• Create your own electoral map.
• Assign states to each candidate.
• Use our live market probabilities or assign your own probabilities for each state/district.

Operation

LatestPollResults is operated and co-authored by Stephen and Mark Delano.