Prediction one: Donald Trump and Hillary Clinton will win their parties' nominations. Prediction two: Clinton will take the White House handily. At least, that's what political wagering websites say.
These operations, which distill the wisdom of crowds, turn out to be surprisingly solid when it comes to predicting presidential races. Their record for accuracy outdoes that of opinion polls.
Most Web-based political investment markets -- they don't like to call themselves betting pools, not wanting to run up against online gambling bans -- are structured to run much like commodity futures exchanges.
According to the oldest such election bourse, the Iowa Electronic Markets, which has been operating since 1988, Trump has 85 percent odds of capturing the Republican nomination this year; Clinton has a 90 percent chance of gaining the Democratic prize. And the Democrats' candidate in the fall is the favorite, by 67-33 percent. (The Iowa market, run by the University of Iowa, doesn't put a person's name at the head of the ticket before the party's nominating convention.)
That's a slightly better result for Clinton than the Bloomberg poll finds: 54 percent for her and 36 percent for Trump.
Another prominent exchange that trafficks in the presidential contest is Predictit, a two-year-old organization run by Victoria University in New Zealand, with a Washington, D.C., office. Both the Iowa exchange and Predictit received OKs from the Commodity Futures Trading Commission to operate. The two organizations view themselves as more of an academic experiment than a business.
Then there are the storied British betting houses, Ladbrokes and Betfair Group. The two U.K. firms accept actual bets, chiefly on sports and casino games, but also have lively gambling action on how elections will fare. By law, however, U.S. dwellers can't deal with these outfits, which have similar American election prognostication numbers as the Iowa exchange and Predictit.
How accurate are these predictions?For nationwide races, as opposed to state presidential primaries and caucuses, they have been quite reliable, and more so closer to an election.
Iowa professors did a study of their exchange's results through 2008, its 20-year anniversary. They found it was accurate 74 percent of the time 100 days from an election, defining "accuracy" as being more precise than the polls. Opinion polls tend to be more exact right before a balloting, as voters focus with greater intensity on the candidates. So, five days prior to an election, the exchange was better than the polls 68 percent of the time. Predictit says it has an 84 percent accuracy rate, as measured against actual voting outcomes.
The task is tougher for primaries and caucuses. Predictit nailed the most recent contests, in Arizona, Utah and Idaho. But for Super Tuesday on March 1, it mistakenly projected that Vermont Sen. Bernie Sanders would lose to Clinton in Minnesota and Oklahoma. Instead, he won those two states by convincing margins.
The reason likely is that when traders from around the nation predict a specific state's outcome, they're not always clued into its particular situation. Brandi Travis, a spokeswoman for Predictit, said its Iowa-resident traders were more accurate for that's state's Feb. 1 caucus results than were non-Iowa ones.
For the next big contest, April 5 in Wisconsin, Predictit says Texas Sen. Ted Cruz will sweep to victory with 60 percent of the vote, versus 40 percent for Trump. For the Democrats, the exchange's traders believe, Sanders will outpace Clinton, 54-48 percent.
Why are they often so prescient?Various studies conclude that it's the nature of the people who participate in the electoral exchanges: mainly those who are highly informed about politics. Plus, they have a monetary stake in their election picks, which makes them take the endeavor more seriously than someone selected at random by pollsters for a phone survey.
"Polls are a static, one-time prediction," said Joyce Berg, a University of Iowa accounting professor who's the exchange's director. But traders on the exchange, she added, are continuously drinking in new information about the political state of play, and thus are more in tune with changing dynamics.
How much money can you make with these investments? Not a lot. Most limit an investment to $500 per contract. You might double your money. But the winnings are a pittance compared to what Wall Street sharpies can reap with their multimillion-dollar antes for stock and commodities trades.
People who invest in the political exchanges "do this because it's fun and interesting," said Tom Snee, spokesman for the Iowa exchange. Along the way, they give us a pretty decent crystal ball, although with this year's topsy-turvy race, you can be less sure that things will follow a predictable path. Last September, the exchanges gave little chance that Trump would advance this far.