NBA bans under parlays on low-minute players after Rozier charges

Scott Kaufman-Ross sent the memo on a Friday. By Monday, DraftKings and FanDuel had instructions: stop offering parlays on multiple statistical unders for the same player, stop taking wagers on opening-tip winners, stop pricing unders on anyone projected to play fewer than twelve minutes.

The mechanism the NBA is closing is not subtle. A player who knows he is leaving a game early — injury staged, illness claimed, minutes managed — can share that information before tip-off. Someone outside the building places a statistical under. The player exits after nine minutes. The bet cashes. Terry Rozier, the memo's proximate cause, allegedly ran exactly this sequence in March 2023 when he was still with Charlotte: prosecutors say he disclosed his planned early departure, co-conspirators wagered more than two hundred thousand dollars on his statistical unders, and he played nine minutes before leaving with a foot injury. He has pleaded not guilty.

Jontay Porter did not plead not guilty. Porter admitted it. He left a game after three minutes, an eighty-thousand-dollar parlay on his unders sat waiting, and the NBA banned him for life in 2024. Kaufman-Ross cited both cases in the memo as the analytical basis for the new restrictions.

What the NBA has done is sensible as far as it goes. Statistical unders on low-minute players are the softest target in the market — the information asymmetry is total, the manipulation is physically simple, and the tell, when it appears in betting data, arrives too late. Closing those markets removes one clean exploit.

The part the restrictions do not address is the information itself. The NBA can stop sportsbooks from offering a product. It cannot stop a player from knowing his own body, his own coach's rotation plan, or his own exit time. The gap between what a player knows at 6 p.m. and what a bettor can act on before tip-off is not a product-design problem. It is a disclosure and surveillance problem, and no memo to DraftKings fixes it.

Kaufman-Ross's framing — "seeing where we may have had issues in the past is one of the inputs" — is accurate but retrospective. The FBI probe that produced thirty-plus arrests in fall 2025 was built on patterns that were visible in betting data long before charges came. The NBA's authorized partners account for, by the league's own estimate, more than ninety-nine percent of U.S. legal handle on its games. That concentration means the data exists. The question is whether the surveillance infrastructure processes it fast enough to matter, or whether rule changes announced after charges are filed will always be the primary tool.