Red, White & Predictions: Why We're Analyzing Them

The presidential election is six months away, which is well beyond the 100-day mark when predictive analytics start to have meaningful rates of accuracy. But that’s not going to deter MediaPost from publishing a regular series of prediction data each month leading up to the election to see how it nets out.

Most are familiar with polling methods but many believe prediction markets may be a better method for predicting actual outcomes.

In addition to prediction markets, we’ll be monitoring the "13 Keys."

“They are votes up or down on the strength of the performance of the White House party,” American University Professor Allan Lichtman explained recently during an interview on CNN.



Lichtman is the author of 13 criteria for predicting presidential elections and he has found that if six or more of the variables go against the incumbent White House party, the party’s candidate he or she will lose. If fewer than six are negative, the candidate will win.

Lichtman’s 13 criteria:

  • Party mandate

  • Contest

  • Incumbency

  • Third Party

  • Short-term economy

  • Long-term economy

  • Policy change

  • Social unrest

  • Scandal

  • Foreign/military failure

  • Foreign/military success

  • Incumbent charisma

  • Challenger charisma

Lichtman’s method is important to include, because he has accurately predicted nine of the last 10 presidential elections. He was wrong in 2000, but claims that he was right because Al Gore won the popular vote even though George W. Bush won the presidency based on the Electoral College outcome.

If you know anything about predictive analytics, any success rate above 80% is beyond spectacular. Lichtman’s track record is outstanding with or without getting it right in 2000.

Lichtman even got the 2016 election right, which is amazing considering I checked the polls and prediction markets before and after that election and they were pretty much wrong across the board. Even the most famous predictive analytics person of our time -- Nate Silver of FiveThirtyEight -- missed that one.

Right now, the polls are leaning towards a Trump win in 2024. The average of the polls as reported by FiveThirtyEight have Trump ahead by less than one percentage point (0.7). They’re reporting Trump at 41.6 and Biden at 40.9. Keep in mind that this difference falls well within the range of error as reported by FiveThirtyEight – 40.3 to 42.8 and 39.4 to 42.2, respectively.

It’s also important to note that FiveThirtyEight’s averages are based on publicly reported polls that meet their quality requirements. They’re not straight averages. They’re weighted averages based on a couple of factors including a pollster’s empirical record and methodological transparency.

Another thing worth mentioning is that these numbers are being reported as polling averages not to be confused with election forecasts, which FiveThirtyEight promises to make available in a couple of months. Its forecasts will include additional inputs.

While polls are favoring a Trump victory, prediction markets currently favor a Biden win.

The Iowa Electronic Market’s (IEM) winner-take-all market favors a Democratic win by a large amount. Share prices for a democratic win are 0.92 vs. 0.18 for a Republican one. It’s important to note that IEM has a great track record, better than the polls, particularly when predicting more than 100 days in advance of an election.

PredictIt, another prediction market we’re tracking, also favors the Democratic party. It has Biden trading at 50 cents vs. Trump at 47 cents.

Professor Lichtman, is predicting Biden to have an advantage. 

Based on his 13 factors, Lichtman predicts Biden currently has the advantage, because as the incumbent, six or more would have to go against him to predict his loss. Currently, Licthman sees only two criteria failing:

  • Party mandate, based on House elections in 2022.

  • Incumbent charisma (see mainstream media coverage).

Though there are additional ones that could be problematic:

  • Third-party, depending on whether RFK Jr. emerges as a viable candidate.
  • Social unrest (see recent protests on college campuses).
  • Biden could lose the Foreign/Military Failure and Foreign/Millitary success criteria, depending on what happens in Israel and Ukraine.

If you only focus your attention on polls, Trump currently has a marginal advantage. But if you analyze prediction markets and Lichtman's keys, which have been proven to be more accurate predictors, Biden is in better shape.

Stay tuned to see how the methods trend in the months ahead.

Next story loading loading..