I made $15 betting on Donald Trump’s tweets (and you can too!)

PredictIt, which calls itself “The Stock Market for Politics” combines political punditry with legal gambling.




I made $15 betting on Donald Trump’s tweets (and you can too!)

PredictIt, which calls itself “The Stock Market for Politics” combines political punditry with legal gambling.

I got into gambling three years ago, at the age of 27, when I had to fly to Las Vegas to interview the British DJ duo Disclosure. I was scheduled to talk to them at one in the morning backstage at a nightclub in some casino — they were DJing the club at two — and so, with some time to kill, I took $20 and started playing the slots. I ended up winning something like $80 through sheer dumb luck, and even though Disclosure blew me off (long story, but it turns out famous British DJ duos enjoy drinking champagne with their friends more than talking to reporters), I at least made enough money to pay for the Ubers I had to take to and from various airports.

Ever since then, I’ve found that the main thing that sucks about gambling is that it’s illegal to do it on my phone. Everything else — the social stigma surrounding it, the low odds of landing a windfall, and the very real possibility of losing way too much money on it — is what makes it great (for me, at least).

Because I don’t live near a casino and I don’t know any Tony Soprano-types to invite me to their all-night underground high-stakes poker games in a crummy motel suite, I often find myself poking around in the liminal space between gambling and legitimate finance. I’ve lost money buying and selling Bitcoin and other, sketchier cryptocurrencies; trading high-risk futures contracts on the stock app Robinhood; and recently, I spent 45 minutes playing scratch-off lottery tickets and netted a grand total of $10. There’s a bar near my house that recently started selling scratch-offs, too, and after I finish writing this article I plan to go over there and buy one.

This, as much as anything else, is how a few weeks ago I found myself signing up for PredictIt, an online portal that bills itself as “The Stock Market for Politics.” The site functions as what’s known as a “prediction market,” allowing users to trade “shares” of a given outcome with the prospect of a sizable payoff if it comes to fruition. It’s kind of a hard concept to wrap your head around without the aid of a practical example, so let’s look at this actual PredictIt market you could bet on if you so chose:

As you can see, this is a list of popular contenders for the 2020 Democratic presidential nomination, each with a value next to their name. If, say, I’m convinced that after a million debates, a billion primaries, and a zillion news cycles, Democratic voters are going to decide that they’ve had enough of politics as usual and instead go with Andrew Yang, the oversaturated political neophyte who’s really into the idea of a Universal Basic Income, I would pony up $90 buy a thousand shares of Yang at nine cents each. If Yang actually secured the nomination, my $90 would turn into $1,000; if he dropped off the face of the earth and Democrats opted for somebody boring like Joe Biden or his son Beto O’Rourke (don’t @ me, this is canon now), I’d get nothing.

What makes a prediction market like this interesting, however, is that you don’t have to hold onto your shares until the nomination is decided. If Yang were to experience an upswing in popularity — or if, alternatively, QAnon turns out to be real and Biden, Sens. Bernie Sanders, Kamala Harris, Elizabeth Warren, and Mayor Pete Buttigieg all get arrested for being secret generals of an army of Antifa supersoldiers or whatever — and the value of your Andrew Yang shares shoot up to 60 cents a pop, you can always sell your shares for a nice profit if you think that Yang will be outpaced by another candidate and/or arrested for various QAnon-related crimes. (On the flip side, you can also bet against a candidate winning the nomination, which in the case of Andrew Yang seems like a pretty safe move.)

Media-watchers often derisively term stories of candidates’ personalities and electoral prospects that ignores their actual policies “horse-race coverage,” which is ironic, because trading on PredictIt is pretty damn close to betting on actual horse races. As someone who follows national politics fairly closely and has a clear set of ideological preferences, when I write about politics I often advocate for what I hope to see happen, rather than what I think will actually come to pass. On PredictIt, meanwhile, I found myself indulging my political cynicism for fun and profit.

Of the nine bets I’ve placed on the platform, I’ve made a profit off two, lost money on one, and am waiting out the other six. My biggest moneymaker was on a market asking users to pick which of the Democratic candidates not named Bernie Sanders, Joe Biden, or Kamala Harris would be the most popular on May 31. It’s an awkward and arbitrary thing to measure, but I found the market at the height of Mayor Pete mania, and his shares were trading for something like 90 cents each.

I once met a wizard, and that wizard told me to never doubt the fickle tastes of mainstream Democrats who are really into MSNBC, so I bet against the prospects of Mayor Pete. A few days later, everyone had decided that Mayor Pete was kind of lame and had gone off looking for the next Democratic presidential hopeful to get really into. My “No” shares, which had once been worth pennies, were now worth something in the neighborhood of 60 cents each. I sold them because while I don’t doubt Democrats will all suddenly decide to be obsessed with Kamala Harris or Cory Booker for a while, there’s no telling how quickly they’ll get bored of them and decide to give Buttigieg another tryout as the leader of the second tier of all presidential hopefuls. (Just before this article went live, I checked the market again. Mayor Pete is still in the lead, but Elizabeth Warren is giving him a run for his money.)

In case you were wondering how all of this is legal — again, it is basically gambling — it’s because prediction markets such as the ones I discussed above have genuine value to researchers in both academia and statistics-minded members of the press. Nate Silver of 538 has cited them as a worthwhile data point to consider when picking winners of political races, and has written that their odds tend to mimic his website’s forecasts more than conventional polling data does. (Silver has also criticized PredictIt in particular for being inherently “biased toward candidates that young white dudes who spend a lot of time on the Internet like.”) The knowledge that we will be given a reward, even a small one, if we correctly predict an event’s outcome tends to change the way we think about the event itself, putting us in a mindset which famed economist John Maynard Keynes described as “[devoting] our intelligences to anticipating what average opinion expects the average opinion to be.” In a prediction market related to electoral politics, I’m essentially trying to look at a market such as the one involving Buttigieg, assess whether the conventional wisdom of those participating in the market is correct in its prediction of popular sentiment, and if it’s not, place a bet and hope I’m right.

An early, and completely terrifying application of the quirks of prediction markets occurred in 2003, when it was reported that DARPA had quietly launched a website with the innocuous URL of PolicyAnalysisMarket.org which would allow people to place bets on the likelihood of various events occurring in the Middle East, including terrorist attacks. The idea was that the markets on the site might help the Pentagon crowd-source an opinion on where it ought to be devoting its resources in the region, but it quickly drew criticism because of the possibility that paramilitary groups could bet a bunch of money on a terrorist attack occurring, then carry out said terrorist attack and get paid for doing so.

This potential for exploitation of prediction markets by those with the power to determine outcomes is also present in PredictIt, albeit in a much more benign form. One of the quirks of the site is that you can bet on the number of Tweets that Donald Trump, Mike Pence, the White House, and other government Twitter accounts will issue in a given timeframe. Given that Trump has mostly hired crooks to work for him, it’s completely feasible that someone with insider access to the president could make tens of thousands of dollars by, for example, betting that Donald Trump will tweet the phrase “Sleepy Joe” eight times from April 30 to May 7.

So far Trump has tweeted “Sleepy Joe” zero times and the value of an 8x-Sleepy Joe is worth eight cents. But if a Trump staffer who has the ear of Donald Trump were put down a buttload of money on eight Sleepy Joes, then convinced him to tweet the phrase “Sleepy Joe” eight times on the evening of May 6, they’d be able to walk away with a fat payday. (It’s worth noting that any heavy action on something so specific would likely draw the attention of others participating in the market and cause them to bet similarly. Also, if you are this hypothetical Trump staffer of which I speak, I’ve got a few bucks riding on Trump tweeting the phrase “Sleepy Joe” twice, so any help you could provide me would be great.)

My other profitable PredictIt trade involved betting that Donald Trump was going to send 105 or more Tweets the week that the Mueller Report came out; I doubled my initial $7 investment by selling my shares while we were sitting in the eye of one of Trump’s Sunday tweetstorms. I lost $5, meanwhile, by taking some long odds that Bernie Sanders would be leading the Democratic field at the end of April.

I like Bernie Sanders; he and Elizabeth Warren are both people I would love to have as a president. I’ve also placed a bet on PredictIt that Warren will be the 2020 Democratic nominee. The rational, analytic part of my brain believes that, given Warren’s serious, well-thought-out position on pretty much every issue, as time goes on and Democrats test out various candidates of the week only to discover there’s no there there, they’ll come around on Warren and her stock, both metaphorically and on PredictIt, will go up. If that happens, the rational, analytical part of my brain will tell me to sell my shares out of fear that she won’t be able to sustain her momentum.

But another part of my brain, the part that hopes to live in a country that’s a more just and fair place, believes that there are politicians out there with the vision and expertise to make that hope a reality, and will tell me to hang on. Prediction markets imply that you have very little power over what will happen in politics, so you might as well make money off the inevitable mediocrity that elections will yield. And while betting is definitely extremely fun, it’s nowhere near as fun as it would be to live in a country that doesn’t suck shit. Maybe I’ll stick to scratch-offs from here on out.

[Correction: A previous version of this article contained incorrect calculations regarding the potential profit of a $90 investment in nine-cent shares of Andrew Yang stock on PredictIt. We regret the error, and would like to thank Twitter user @stan_vanwyn for catching it.]