Shai Gilgeous-Alexander sat out on a Friday night in March, and 167 other NBA players joined him. The discourse, predictably, went straight to load management: stars resting, fans getting ripped off, the regular season becoming meaningless. It's a clean narrative. It's also aimed at the wrong target.
The 2025-26 NBA season has produced an average margin of victory of 13.1 points, the largest in league history. That number measures something specific: how often games are competitive. A high margin of victory means blowouts, and blowouts mean one team stopped trying to win. Load management does not explain 89 games decided by 30 or more points. Tanking does.
The Arms Race Nobody Wants to Admit Is Happening
The Utah Jazz signed Andersson Garcia and Bez Mbeng from the G League in early March and played them heavy minutes. Garcia logged 169 minutes across 5 games while the Jazz were outscored by 69 points. Mbeng posted a minus-146 net rating for the month, which, for context, means the Jazz were getting demolished every time he was on the floor. These are not players resting. These are players who, by every measurable standard, should not be in an NBA game.
The Memphis Grizzlies started 25 different players this season, an NBA record, while losing 15 of their last 17 games. Eight teams are currently running some version of this strategy, all chasing lottery odds for a loaded draft class. One team executive told ESPN: "It's a copycat league. All the models and ideas, there are always further iterations. That's what happens when it works." He's right, which is the problem.
The NBA fined Utah $500,000 in February after the Jazz benched Lauri Markkanen and Jaren Jackson Jr. in the fourth quarters of close road games against Orlando and Miami. Commissioner Adam Silver called overt tanking behavior that "undermines the foundation of NBA competition." The fine is $500,000 against a franchise worth over $2 billion. That is not enforcement. That is a strongly worded suggestion.
Why the Load Management Argument Keeps Winning Anyway
I'll grant the other side this: a star sitting out a nationally televised game is more visible than a team quietly signing G League players to lose. The fan who drove two hours to watch Giannis and gets a DNP-rest notice has a legitimate grievance. That frustration is real and the league should address it with stricter rest protocols on marquee games.
But load management is a player health decision with a statistical rationale. Teams that manage minutes across an 82-game season show measurable playoff performance gains; the relationship between regular season workload and postseason injury rates is documented enough that ignoring it would be analytically irresponsible. Tanking is not a health decision. It is a deliberate choice to field non-competitive rosters for months, which is why Silver used the word "epidemic" and why the blowout numbers are historically bad.
The fix is structural, not punitive. A flat lottery with equal odds for all non-playoff teams removes the incentive to finish last rather than 12th. The NBA has resisted this because it feels unfair to bad teams, but the current system has produced an arms race where eight franchises are actively competing to lose. The $500,000 fine tells teams the league is watching. Equal lottery odds would tell them the math no longer works.
Resting Shai Gilgeous-Alexander is a process decision. Signing Bez Mbeng to accumulate a minus-146 rating is a business decision. The numbers know the difference, even if the discourse doesn't.
