The political prediction market is a burgeoning industry, with several sites offering crowd-sourced analyses of upcoming election results and other outcomes in the realm of democratic governance. Essentially, a political or election prediction market is used to give people – and political campaigns – tools for how to predict market share for politicians and their policies.
For prediction markets, 2020 is shaping up to be the busiest and most important year ever, as the activities and trends on such sites are being relied upon as much as – or even more than – typical polling data.
To learn more about prediction markets and how they are used by analysts, pundits, bettors, and more, our brief guide covers the basics and introduces you to the two best prediction markets on the Internet. While these won’t change how you vote, they’ll definitely have an impact on how you’ll bet on the 2020 Presidential election odds and other prominent races this November.
Prediction markets are crowd-sourced markets that allow people to “trade” on various political outcomes. Commercially, prediction markets are sold to campaigns and news agencies, as they offer predictive analytics marketing that both entities use to craft their messaging.
However, outside of that business model, election prediction markets – in particular, the 2020 election prediction market – allow curious onlookers to get a grasp on whom their neighbors and fellow voters support and where they sit on any issue of the day.
Of course, it’s an imperfect science. Market predictions this week may be quite different from a given market prediction today. Politics is extremely volatile, with big swings in a candidate’s chances often occurring overnight as breaking events dominate the mainstream news cycle.
In addition to these research sites, sportsbooks with election odds are also considered to be prediction markets unto themselves, and political betting was indeed the first kind of prediction market for politics in general. The practice goes back several centuries and has been prominent in the US since at least 1884.
Yes! Presidential prediction markets – and all prediction markets across most industries (i.e. wider elections, the stock market, the real estate market, etc.) are legal. Most of these don’t actually process real-money trades from their members, and those that do – namely PredictIt – have been excluded from the oversight of US gambling laws. Still, even these aren’t actually betting outfits.
That said, presidential election betting sites are also considered to be prediction markets, and they process real-money wagers around the clock for US bettors in most states.
These political sportsbooks operate outside of United States jurisdiction, making them legal and safe to use, and the political odds hosted by such operators are cited in the mainstream media and used by campaigns just like PredictIt, PredictWise, and sites like FiveThirtyEight are.
For prediction markets, 2020 election outcomes are obviously the biggest mover and shaker. The most popular categories are things like Presidential election prediction markets, but there are several others, all of which have a high volume of “trading” or participation. These include – but are certainly not limited to – the following:
In the betting market, election prediction is big business, with gamblers all over the world turning a handle of hundreds of millions (if not billions!) of dollars on the outcomes of various primary and general races. Some analysts contend that for every major US election, the betting handle is roughly 50% of the cost of the campaigns in question.
While every country has election prediction markets and sees bettors placing wagers on domestic votes, the US Presidential election is the most popular by far, and prediction markets are a key tool that bettors can use to gauge public sentiment.
Prediction markets are considered to be far more accurate than traditional polls, though political predictions for 2020 are all over the place, and for good reason: Never in American history has there been an election with quite this kind of backdrop.
From the coronavirus pandemic to the so-called 1619 Riots and the open calls for communism in a traditionally capitalistic society, it’s hard for anyone to wrap their head around how the public is going to vote in November. Thus, if you want to make an informed bet, research is your friend.
If you’re looking for the best political prediction market, outside of actual political betting sites, there are two major players in the game. These include the market leader, PredictIt, and a competitor called PredictWise.
While FiveThirtyEight is considered a prediction market, it does not allow for the crowd-sourced “line moving” that these other sites use to meter public support for politicians and the many issues of the day.
If you bet on politics, PredictIt should be one of your most frequented resources on the Internet. The site allows for real money “trading,” where members can sign up and buy “shares” of the outcomes they favor. They can also trade these shares, as if the picks were stock holdings.
This is a pool-based system that is not considered to be gambling (as it’s more akin to pari-mutuel wagering and stock speculation), nor does it payout like a proper sportsbook. Still, it’s a great resource, and we advise all political gamblers to use PredictIt in their research.
PredictIt is designed for educational purposes only, and the non-profit was founded in November 2014 as a project of the University of New Zealand (Victoria University of Wellington). Because the site makes no money off the “wagers” being placed, it is not a true gambling operation. PredictIt has received a gambling exemption from the US Commodity Futures Trading Commission and has over 160 data-sharing partners.
PredictIt’s members sign up and buy “shares” of a given political outcome at the current price (between one cent and 99 cents). Then, whenever the event in question is settled, winning shares get $1, while losing shares get nothing.
At PredictIt, Democratic nominee odds were among the most “traded” in 2020, though Joe Biden was not the presumed winner for the majority of the cycle, demonstrating that the site is not without its own predictable flaws.
Right now, the PredictIt President 2020 category is the heaviest hitter, with some 88 million shares traded. Presumptive outcomes are expressed as cents on the dollar. For example, the PredictIt Trump 2020 selection shows that Joe Biden is trending at around 63 cents, while The Donald is at about 37 cents. This means that you can buy a share of Biden at 63 cents and make $1 if he wins. Similarly, you can buy a share of Trump at 37 cents and make $1 if he wins.
Yes, but not much. Whereas legitimate online betting sites can accept wagers of several thousands of dollars and payout several thousand more, the PredictIt max bet is $850, and no “question” or category can pool together more than 5000 individuals “traders.” As a betting site, PredictIt is a poor option for real gamblers looking to cash in on the 2020 election and other events.
PredictWise is the biggest PredictIt competitor, though it is relatively small in comparison, as it does not use a financial motive model for participants. PredictWise aggregates data using targeted passive participation, but with no “skin in the game,” there is less to temper participants from making their selections based on emotion rather than actual expectation.
Otherwise, the site uses the same sort of approach, crowd-sourcing various election outcomes and other political fare to achieve a wider, more holistic, more granular, and hopefully more likely series of outcomes. As a result, political bettors should also use PredictWise when assessing the odds at any Presidential election betting site.
The PredictWise 2020 numbers are more or less in line with PredictIt, and for the 2016 primaries, PredictWise accuracy was considered the best of any market election prediction service.
Instead of showing results as cents on the dollar, the site uses typical percentages to express its data. While you can expect differences, most of the action at PredictWise jibes with the status quo, with the site favoring a Biden win in 2020.
Unfortunately, PredictWise is more focused on business-to-business aspects of data collection, meaning that there isn’t as much to be gleaned by the curious bettor searching the site’s limited publicly-available numbers. However, many political sites and news agencies use PredictWise’s data and publish various hot takes based on it, so it can still be a part of your overall election betting strategy.
In general, no matter the prediction market, 2020 election outcomes can be bet on with more confidence by using these services for research purposes. They are absolutely better than political polls, as they apply to many more respondents and have far less inherent bias. Here are some of the questions people ask that rely on prediction markets:
Election prediction is exactly what it sounds like: election prediction! Regardless of the type of prediction market or analytical approach, an election prediction seeks to use empirical datasets to correctly pick the winner of a given race. Betting lines, prediction market chances, polling data, and academic historical models are all products of election prediction.
That depends on which race you’re talking about. There are hundreds of federal-level political races to be decided in November, and prediction markets have data on all of these. Betting sites, too, cover a large number of the most contested races. Right now, Democrats are expected to do well in the general, with Biden beating Trump and the GOP losing the Senate. Only time will tell how accurate these predictions are, of course.
Currently, all the best online sportsbooks with election odds are predicting a Joe Biden win in 2020, as are all the top prediction market sites. However, various analysts and academics disagree, with several prominent modelers calling Trump reelection inevitable.
In close races, not very often, especially lately. In the most recent elections going back to the 1980s, polls incorrectly determined that Ronald Reagan, George Bush, George W. Bush, and Donald Trump would lose, though each candidate went on to win. It is precisely because of the unreliability of polls that prediction markets and election betting sites have more merit when it comes to correctly picking the next President.
When it comes to who has accurately predicted Presidential elections in the United States, there are many analysts and data scientists who lay at least some claim to fame in the field. However, most models used by these academics are flawed or make assumptions that may not be germane to the given current election. That said, Professor Allan Lichtman has correctly predicted the last nine Presidential elections, including Trump’s 2016 upset victory over Hillary Clinton. Even more famous is Professor Helmut Norpoth, who developed the Primary Model that has correctly predicted the President for most of the last 100 years (as retroactively applied using the model’s constants). Northop called the 2016 election for Trump, and he’s calling the 2020 election for Trump, as well (91% to 9%).
Historically, the so-called purple states – or swing states – are what determines the outcome of any given Presidential election. However, of these dozen or so states, the 2020 election will likely come down to just four: Florida, Michigan, Pennsylvania, and Wisconsin. Florida is, in particular, the biggest swing state and most coveted prize in any general election for President of the United States.