Critics say that Augur—the crown king of smart-contract-based prediction markets—has a major flaw: outcomes of bets must rely on human “reporters” for verification. And humans, as we know, are notoriously corrupt and can collude to rig results.
Hodl Hodl, a non-custodial peer-to-peer crypto exchange incorporated in the Marshall Islands, thinks Augur needs a rework. In a Medium post yesterday, Hodl Hodl unveiled its own prediction market, “Predictions,” which eliminates Augur’s gang-of-reporters function. Instead, bettors deposit their wagers into a smart contract and debate the outcomes with their counterparties. The escrow account can only be opened once they have agreed upon the results—and if they can’t agree, the dispute is arbitrated by Hodl Hodl itself, a “non-custodial” exchange that can neither access nor hold funds.
If a person disputes an obvious outcome, said CTO and co-founder Roman Snitko in an interview, "this money would remain locked in escrow forever"
“All the Bitcoins are sent into escrow,” he explained. Specifically, it uses a two-out-of-three multi-sig escrow, meaning it requires at least two of three keys to unlock. “We [Hodl Hodl] have one key, the seller has one key, and the buyer has one key,” Snitko said.
He said this ensures that bettors don’t co-opt the verification system for their own ends. If two opposing bettors have to determine a market’s outcome, their mutually exclusive self-interest—that they benefit from different outcomes—stops them from collaborating to steal the funds. Bettors don’t have the incentive to cling onto a false conclusion either. If they lose the bet, they’ll lose their funds regardless, either to the counterparty or in eternal escrow, said Snitko.
In the case of a bettor disputing an outcome, Hodl Hodl would step in and, hopefully, side with whichever party is telling the truth. Snitko said Hodl Hodl can be trusted to mediate fairly, since it would become quickly obvious, and bad for business, if the exchange were to collaborate with outcome-deniers.
Even so, Snitko plans on setting up a market for freelance arbitrators, paid by bettors, who would stake reputation on making fair, impartial judgements.
Hodl Hodl also vets all incoming markets, meaning it can disable any that are ambiguous or illegal. So no endless puzzling over whether the weather was “good” on a particular day.
The quest for perfect predictions
Building a perfect prediction market has been a kind of holy grail among business thinkers for the past two decades. The basic idea is to tap into the so-called “Wisdom of the Crowd” by aggregating bets on the outcomes of any event. The data that emerges teds to be as reliable as the predictions proffered by experts, if not more so.
Augur, for instance, successfully predicted the outcomes of the Brett Kavanaugh hearings, and the 2018 midterms. (Though Veil, an Augur interface, faltered on the Oscars.) Decentralized prediction markets, of which Augur is one, need “oracles”: sources of real world knowledge that can submit the true outcomes to the blockchain, releasing the winnings.
Yet the mechanisms incentivizing Augur reporters to be honest brokers are limited. Reporters forfeit money if they deviate in their conclusions from the majority, the assumption being that anomalous reports are likely dishonest. Case in point: in November, an aberrant Augur bettor placed a large wager on a Republican victory in the midterms—after the Democratic Party had secured its commanding majority. Needless to say, the belated wager sparked concerns it was an attempt to exploit the market’s ambiguous wording.
A closer look
To illustrate the possibilities of Hodl Hodl’s solution, imagine two bettors wagering on a soccer game. Adam has bet $300 that Manchester United will win, while Beatrice has bet $30 that it will lose. Both place these funds into a lockbox, which requires two keys to open. A trusted third-party, a la Hodl Hodl, holds a skeleton key. After the match concludes, things go awry. Though Manchester United clearly won, Beatrice is in denial. The team cheated, she complains. It won on a fraudulent technicality.
On Augur, Beatrice could, feasibly, assemble a large enough share of like-minded deniers to stage a coup and produce a false outcome, releasing the winnings to herself and her accomplices. But on Hodl Hodl, Beatrice, locked into a codependent relationship with Adam, has zero clout. Adam has no incentive to cave to Beatrice’s demands, because he stands to gain from the correct outcome. Nor does the third-party arbitrator, Hodl Hodl itself, who has nothing staked either way.
Of course, in Hodl Hodl’s particular model, the third party cannot access the funds. The worst that can happen? All three parties dither and disagree, the funds remain locked away for eternity, and everybody loses. Now there’s an incentive to cooperate.