Victoria University of Wellington’s Massive Gamble, the A-Bomb Kid, and Donald Trump.
WILL IRVINE, DARCY LAWREY
This story discusses gambling addiction. If you or a loved one has problems with gambling, please call 0800 654 655 or free text 8006.
On October 27, 2024, soon-to-be President Donald Trump held a rally to end all rallies in New York City’s Madison Square Garden. Hailed by liberals as a successor to the 1939 American Nazi rally at the same location, the event drew lines stretching around the block. Anxious Trumpists in a city surrounded by boutique coffee bars and decolonised tattoo galleries waited to hear their beloved leader speak, while “comedians” like Tony Hinchcliffe warmed up the crowd with a smattering of off-colour Latino jokes. As the lines trickled into the glammed-up arena, massive billboards showed something previously unnoticed by the vast bulk of society: election betting odds. For the first time, millions of Americans rushed onto one of a few sanctioned betting sites to try their hand at predicting the new president.
Amongst the big three election gambling sites (officially “prediction markets”), the third-biggest competitor was PredictIt: a research project run by Victoria University of Wellington. In just one day (November 4, 2024), 518,000 shares in various candidates’ futures were traded through the University-run site. Yet despite the massive flow of cash and options through their servers, the financially strained University is oddly silent about their project. Indeed, students around campus - whether it be Kelburn or Pipitea - are largely unaware of the giant betting operation running beneath their feet. Even the few Vic Students who have heard of PredictIt are surprised by the scope of the project.
PredictIt was once the biggest prediction market in the world, but years of litigation in US courts have tightened its requirements. There is an $850 cap on each bet, and traders must be US citizens in order to purchase shares. Conversely, on competitor site Polymarket, traders are explicitly forbidden from being US citizens or residents, although this rule, like PredictIt’s, is regularly flouted.
In 2016, 14-year-old James (named changed for anonymity) saw something very few others did. With next to no knowledge, aside from his parents’ assurance that “Americans are stupid”, he predicted Donald Trump’s victory as US president. For the intrepid young gambler, this sparked a lifelong side hustle as a prediction market expert. Today, James is a graduate of Victoria University working overseas in AI research. Throughout his betting career, he has made more than $170,000 on PredictIt alone (although he left the site in 2021). When pressed on how much he’d made on other websites, James cited legal reasons and refused to tell us more. “Just know it’s a lot of money”.
James chalks his success up to a clear understanding of electoral patterns. On betting against Democrats in the 2024 elections, he told Salient that “almost every incumbent party in the developed world has done badly in elections since 2022”, pointing to Europe as an example. He also employs more traditional prediction factors, like unemployment and inflation. “If anything, I was impressed by how well Democrats did in this election… you need to take into account that their opponent is Donald Trump. If the Republicans had run a sensible candidate, like Nikki Haley, it would’ve been a lot worse for the Democrats”.
Yet James doesn’t see PredictIt as a particularly useful tool for election research. He argues that the $850 limit on bets causes significant market distortions that limit the program’s research use. “It’s a monetary limit, not a limit on shares… If you’re buying Biden to win at 85 cents and somebody else is buying Trump to win at 15 cents, then you might need 6x as many Biden voters for the market to be even”. However, he notes that this hasn’t always been the case. “There was a better case for it as a research tool in 2016… prediction markets were new, and researchers could gloss over the fact that there are market failures”.
Voters, especially from the right wing, have pointed to prediction markets as an alternative in the age of failing polls. James isn’t averse to this: “the average person will probably get a better idea of the reality from Polymarket than they would from CNN or Fox News”. He points to a particular example from this election, where right-wing voters who disliked Trump were worried that Harris taking office could lead her to enact sweeping socialist reforms. James says that in this situation, prediction markets could disprove this and help people feel more confident in their vote, given that markets overwhelmingly favoured Republicans to win control of Congress.
As for his lucky wager in 2016? “It was a guess. If I’d done it now I probably would’ve bet on Hillary Clinton”.
Salient reached out to Deputy Vice-Chancellor (Research) Margaret Hyland, in the notable absence of any University official who was specifically responsible for PredictIt. Hyland has made public statements on behalf of the University concerning PredictIt, including the response to their 2022 temporary closure by the Commodity Futures Trading Commission. Hyland told Salient that the University does not make any profit from PredictIt, although she acknowledged that they receive a USD$2,000 payment each month to help keep up with expenses. Per Hyland, the user charges incurred on the website are associated with the cost of executing prediction contracts, and do not generate profit for the University or PredictIt itself. However, the same can not be said for the overseas partners that make the operation possible.
PredictIt’s US partner is Aristotle Inc., a data mining company run by John Aristotle Phillips, the infamous “A-Bomb Kid” who designed a nuclear weapon whilst still in college. In the University’s response to Salient, Hyland acknowledged that Aristotle Inc is attempting to generate a profit from the operation of PredictIt, but has not yet. This is in line with the approach taken by PolyMarket, which employs the tech-industry business model of operating at a loss for a sustained period of time in order to develop a large user base and secure investment.
Critics in the US have pointed out that prediction markets might lead to compulsive gambling behaviour, and there has been significant worry that traders who are drawn in by betting on their preferred candidate would be unable to make rational market bets. This was partially the rationale behind the CFTC’s decision to step in after previously guaranteeing Victoria University a no-action letter.
When Salient spoke to James, he echoed concerns about the rise of gambling addiction. Prediction markets like Polymarket and PredictIt have very few of the legislative restrictions put in place on similar behaviours like sports betting and casino gambling. This isn’t the case across the board (James spoke very highly of BetFair, a British betting site that called him multiple times when he lost on a bet) but broadly speaking, most scholars agree that the market is open to abuse.
This means that seasoned traders like James, with all their knowledge of statistics and predictive factors, can generate massive amounts of money against the trades of lesser-informed regular voters, who are often betting based on the information they get from mainstream media outputs and are thus easily swayed by partisanship and lack of foresight. With very few protections in place, James urged Salient readers not to get involved in election gambling on the hopes that they might make money - “I feel pretty comfortable taking from right-wing rich Trump voters, but I would feel awful making money off broke Vic students”.
At the end of the day, it’s unclear what the objective research value of PredictIt is, or whether its benefits as an academic tool outweigh its place in the growing and dangerous arena of prediction markets. For certain traders, tools like PredictIt offer the opportunity of a lifetime to use one’s political knowledge to get rich. For the vast majority of people, though, they’re yet another addictive sinkhole to syphon cash out of ordinary people’s pockets.