How Prediction Markets Work
Gambling — the wagering of something valuable on an event of an uncertain outcome with the primary intent of winning additional money or material goods — is something man has been doing across his thousands of years of human history. Greek mythology tells the story of the three brothers Poseidon, Zeus, and Hades drawing lots to divide the world between themselves after they defeated their father’s generation of gods. Poseidon won the sea, Zeus the heavens, and Hades got the underworld. Thankfully, the land was left to the rest of us and we’ve been wagering on it ever since.
The history of gambling shadows human history, going back centuries, even millennia. For example, the writer Raymond Sauer quotes a contemporary description of crowds en route to the Circus Maximus in Ancient Rome as “already in a fury of anxiety about their bets”; and historians such as Roger Munting document a close association between gambling and sports such as cricket in nineteenth-century Britain and baseball in 20th-century America.
Both sports also had their share of gambling scandals. The “Black Sox Scandal” rocked professional baseball in 1919, with key players of the Chicago White Sox conspiring to throw the series to the Cincinnati Red, while Pakistan’s cricket bowlers were bribed to deliver no-balls on purpose during specific innings in a match against England in 2010.
As Rebecca Edwards explains in her article Gambling built baseball — and then almost destroyed it, “In baseball’s infancy, the sport was thought of as a boy’s game. But over the course of the 19th century, gambling deepened adult interest and investment in the sport, attracting cohorts of older fans.”
“Gambling’s popularity was helped along by the spread of statistics,” Edwards adds. These were developed so that box scores could be printed in the newspaper. “Statistics also created a pool of data that gamblers could use to inform their bets — many of which were made from the stands, in the middle of games,” Edwards says.
Today, statistics that are far more sophisticated than what these earlier gamblers had are only a cheap subscription away, if not free from many data websites. This data is also being used by a lot more people. Coaches make managerial decisions based on player profile, skill, and in-game performance data. Betting companies set odds by them. Bettors use them to try to profit by spotting inefficiencies in the market. Sports leagues use them to try to track and stamp out match-fixing to keep the game fair.
What is a Prediction Market?
So what exactly is a prediction market? A prediction market creates an exchange for players, or gamblers, to take a position on a specific event that will occur in the future, like whether the price of Bitcoin will be above or below at set price at the end of the year, the result of a political event, or even the U.S. going over 400,000 COVID deaths by June 1st, 2021, as Fairlay current offers rather morosely.
On the exchange, a contract for the specific event, like Biden winning the U.S. presidency (see Figure below), will trade at the market price according to supply and demand at that moment. Prices will range between the 0 and 100 and fluctuate as time progresses. With the Biden bet, price would follow polls and both his and Donald Trump’s activities. When a winner of the election is called or when the market-maker is comfortable enough to close the book and pay off the bet, the winning side gets the full 100 dollar or cent value, while the loser’s bet expires worthless. Prediction markets bets are also known as binary options, one side expires at full price, the other at zero.
Why Do Prediction Markets Work?
In his book Vox Populi, Francis Galton explains how he came up with the idea of statistical correlation while attending a weight-judging competition in Plymouth, England. Eight hundred people bought tickets estimating the weight of an ox, with the person closest to the weight winning a prize. Galton didn’t believe the average competitor would have a clue as to the correct weight of an average oxen but, to his surprise, the vox populi, or voice of the people, was within 0.8% of the animal’s correct weight.
Galton’s story was immortalized in his book Vox Populi and it became an early example of what has been dubbed the “wisdom of the crowds” or the idea that the aggregation of the opinions of many can be a surprisingly accurate predictor of outcomes, even when the opinions are being made by laymen who have no specific knowledge of the subject they are making predictions on.
The “wisdom of the crowds” is the main concept behind prediction markets and the fact that one is pitting one’s opinion — and money — up against a crowd of other gamblers makes winning these kinds of bets far more rewarding than beating a monolithic, faceless casino. “Money won is twice as sweet as money earned,” says Paul Newman’s character Fast Eddie Felson in The Color of Money and it’s especially sweet when cashing a prediction market bet.
“The market mechanism naturally aggregates information of prices, as a transaction happens only when there is a willing buyer and seller at that price. There is a monetary incentive to reveal the truth because of potential gains. In addition, there are long-term incentives for specialization in discovering new information and to trade on it. The markets themselves can do a better job of predicting prices when the prediction errors of each individual are distributed symmetrically around the true value, and with finite variance, so law of large numbers applies. This is known in probability theory as an application of the central limit theorem.”
Not only do these markets work but they are often some of the most accurate predictors of events.
When Prediction Markets Go Wrong
One thing a prediction market should never do is drive events they are taking wagers on. In 2001, the Pentagon started developing what has to be the worst idea for a prediction market ever devised. As John Schoen of NBC explained, the Pentagon wanted to set up a ‘futures market’ to “use market forces to help predict political upheaval in the Middle East”. “Possible ‘contracts’ could include questions like the overthrow of the king of Jordan or the assassination of key figures like Yasir Arafat. These contracts would deal with events in each country related to five major categories: military preparedness, civil stability, economic health, U.S. military involvement and U.S. economic investment,” said Schoen.
As Schoen described it the idea had some merit: “By placing thousands of individual bets, the thinking went, market participants would shape forecasts about future events. The value of the contracts would fluctuate based on market participants’ beliefs about the likelihood of specific events taking place — moving higher if they appeared likely and lower if unlikely. Contracts would expire quarterly and extend a year and a half into the future. Each contract would be fully valued at $1 — the payoff for an accurate prediction.”
“The market was also designed to allow trading in ‘derivatives’ and complex hedging strategies — linking, say, renewed economic growth in Israel to the prospects of approval of Palestinian statehood,” said Schoen. The prospect of parlaying an exacta on the assasination of a political figurehead and with a rise in GDP of a sovereign country is pretty abhorrent.
Forgetting for a second the astounding lack of moral clarity in creating markets on assassinations, bombings, and government overthrows the biggest problem with the idea was nothing stopped the terrorist from placing bets on the site and then fulfilling them and profiting handsomely when the assassination, bombings, and government overthrows were carried out. It actually made the prospect of these horrendous things happening because there was now a new financial incentive for them.
DARPA tried to claim that if terrorists bet on their own attacks warnings about impending attacks would be triggered, just what the system was set up to do, but Senate Democrats weren’t having any of it. They blasted the plan as nothing more than state-sponsored “gambling on terrorism” that might actually succeed in doing what t was trying to avoid.
In all fairness to the Pentagon they weren’t the first to think up this type of bet. As reported by Adam Mann in his article The Power of Prediction Markets “in September 2002, six months before the US-led invasion of Iraq, the Dublin-based betting site TradeSports.com gained international notoriety when it ran a prediction market on when Iraqi dictator Saddam Hussein would be ousted.” How’d the bet turn out? Well, “by the time the war began in March 2003, betters were 90% certain Hussein would be out by April and 95% sure he’d be gone by May or June. He was deposed in April,” says Mann.
The Future of Prediction Markets
Gambling will be with us for as long as humans exist. It’s something innate not just in humans but also in pigeons and monkeys. In her article Study shows pigeons like to gamble, Lin Edward describe research by psychologists Thomas Zentall and Jessica Stagner of the University of Kentucky in Lexington “found pigeons given a choice of a light that would deliver three pellets of food every time or one that gave them a big “win” of 10 pellets 20 percent of the time consistently chose the latter. When averaged out, these results meant the pigeons were preferring a payout of two pellets for each peck rather than three.”
Of course prediction markets are beatable, just like any fair bet. Unlike in a casino, it is possible to tilt the odds in one’s favor if one can spot an inefficiency in the market or you receive some secret knowledge that will adversely affect the market before the majority of other gamblers do. Or, you can build powerful models that will spot anomalies in the betting odds or help predict , like some people have done with the races in Hong Kong.
One could even say a direct gambling line stretches from the gladiators of Rome, who fought and died in the Eternal City’s midday sun, to today’s gladiators who fight in the mud-filled trenches of modern day coliseums like, well, USC’s Los Angeles Coliseum. Or, to the politicians of today who are just the latest in a long line of prevaricating characters who profess a love of democracy and self-sacrifice that is rarely achieved. In that Coliseum, during every USC home game, a galea and manica-wearing gladiator corrals the crowd in chants that probably differ little in vitriol or volume from the cries of the barbaric hordes of Rome, who were betting their hard-earned aureuses — gold coins — on the life and death of the gladiators before them.
At least today, we can say betting on assassinations and the fall of governments is a moral step too far, but sports betting is flourishing now, especially in America, where state after state pushes for legalization to tap into the enormous tax revenue potential it has. The states are throwing in the towel on prohibition, which is a good thing as “The American Gaming Association estimates that Americans illegally wager about $150bn on sports each year,” says The Guardian.
“Simply put, more betting on sports means more eyeballs watching and paying attention to those sports, and that equals more money flowing into the leagues’ coffers,” says Mashayekhi in his article Inside the Battle for the Future of Sports Betting. “According to an October study commissioned by the American Gaming Association (AGA), legal sports betting could result in an additional $4.2 billion in annual revenues for the four major North American sports leagues (NFL, NBA, MLB and NHL) — with the majority of that influx resulting from increased fan engagement with the product, rather than revenue coming directly from the gaming industry,” adds Mashayekhi. Prediction markets and Lepricon will be there to pick up its slice of this healthy and growing market.