The biggest prediction market companies have been eager to position their products as “peer-to-peer,” pitting average Joes against other opinionated casuals. DraftKings CEO Jason Robins, who raved Friday to analysts about the underlying economics of his own predictions product, is taking a different approach.

“The thing that worries me is that some of the predictions guys out there are positioning this as peer-to-peer when it’s really not,” Robins said in an interview. “It is technically, but it’s really more peer-to-Wall Street, peer-to-professional bettor or peer-to-market maker. … You’ve got to present the product as what it is.”

They’re among the most pointed public comments that Robins has made on the underlying realities of prediction markets, a new form of betting that has upended the industry since the start of Donald Trump’s second presidential term. Startups Kalshi and Polymarket, private companies that have raised money at numbers higher than DraftKings’ market cap, have both taken the opposite approach, in marketing materials and public comments trying to frame their products as “peer-to-peer” both in structure and in skill level.

This sharks vs. minnows dynamic was central in the development of online poker, and it was also a vibrant debate when daily fantasy sports surged about a decade ago. DraftKings, ironically, was initially resistant to the framing that its DFS contests were not primarily casual fan vs. casual fan, though it did eventually start gate-keeping contests to separate beginners from more experienced players.

When speaking on Friday, Robins said the comparison to DFS was apt.

“Any time you have a peer-to-peer marketplace, where you have people that are sophisticated and using models, they’re going to win over the retail customer,” he said. “As long as you know that and you’re playing with that knowledge, there’s nothing wrong with that. … I just think we have to make sure in the prediction space that we’re doing the same thing, that we’re explaining how the product works so customers don’t feel deceived.”

DraftKings, which operates a legal sportsbook in more than half of U.S. states, has a different business than Kalshi or Polymarket and therefore a different set of incentives. Robins appears to be trying to frame prediction markets, particularly for people in states with legal online sports betting, as a place for algorithm-backed pro and Wall Street quant funds, while positioning legal sportsbook betting as a more suitable place for the common bettor. DraftKings and its peers have been criticized for limiting many users in their sportsbook apps that show betting patterns reflective of professional operations or an edge against the house.

To underscore that point further, DraftKings said Thursday as part of its second quarter earnings report that it estimated 80-90% of the volume on DraftKings prediction platform in those states came from “professional betting syndicated and institutional traders.”

Robins told Sportico on Friday that he wouldn’t be surprised if the actual number was actually higher than 90%. In states without online sports betting, such as California or Texas, the numbers look completely different.

In discussing his perceived differences between the two products, Robins also took aim at prediction market competitors that he said give too many economic concessions to the market makers on their platform. He compared that to the promotions and free bets that the DraftKings sportsbook gives to its customers. 

“Our competitors are giving the rebates, the equivalent of promotions, they’re giving to the market makers to incentivize liquidity,” he said. “It’s kind of a backwards system. The market makers are making money; they don’t need additional incentives. The retail customers that are mostly losing to the market makers are the ones that should be getting the give-backs.”

A DraftKings spokesman later clarified that while DraftKings’ market-maker program does have some rebate structures, no one to date has reached the threshold to receive them.

DraftKings stock (Nasdaq: DKNG) rose about 5% in Friday morning trading after the earnings numbers were released. Though the company missed analyst estimates on both earnings and revenue, sports outcomes were unfavorable in the quarter, and the company maintained its full-year guidance for both revenue and adjusted EBITDA.