The NBA just told a group of UCSF doctors their Achilles finding is noise. I’m not buying the brush-off.
A paper published last month in the Orthopaedic Journal of Sports Medicine tracked every basketball-related Achilles rupture in the league from 2014 to 2025: 19 tears across 11 seasons. The researchers found something specific and uncomfortable. Minutes spiked in the window right before the tendon gave out. In the 10 games preceding injury, players averaged 29.0 minutes against a season baseline of 27.4. Stretch it to 15 games and it was still 28.6 versus that same 27.4. Both differences cleared statistical significance. Season-long workload looked ordinary. The short-term jump did not.
“Significant workload spikes were found in the 10 to 15 games preceding Achilles ruptures for NBA players,” the paper concluded. That is the entire argument in one sentence. Not total minutes. Not back-to-backs. Not age or position or early specialization. The study checked those and found nothing. The signal lived in the acute ramp.
The Spike Is the Mechanism, Not the Season Average
This is the part front offices should actually care about. Load management debates usually collapse into season totals or “games played” counting. The UCSF group isolated a different variable: a concentrated bump relative to a player’s own baseline in the fortnight before the pop. Veteran players in the sample carried heavier cumulative loads across the year. Younger ones ruptured under lighter observed use. Years of experience correlated with total season minutes and with minutes in the five games right before rupture. The tendon does not care about your narrative. It cares about the sudden ask.
The 2024-25 season made the stakes impossible to ignore. Seven Achilles tears. Three of them in the playoffs: Jayson Tatum, Damian Lillard, Tyrese Haliburton. The league had never cracked four in a single year before that. Last season the count dropped to one, Donte DiVincenzo. Correlation is not causation, and n=19 is small enough that any honest analyst has to flag it. But seven in one season is not a rounding error. It is a red flag waving at every medical staff still treating minutes like a flat ledger instead of a rate-of-change problem.
The League’s Data Stays Locked
The NBA’s response arrived through Dr. John DiFiori, its Director of Sports Medicine. “While we welcome continued learnings, we believe the recent paper on this topic has important methodological limitations,” he said. “Our own extensive review of league data to date, conducted with independent epidemiologists and consultants, spanned the 2012-13 season to the present and found no relationship between Achilles tendon injuries and minutes played before injury.”
I believe they ran the review. I do not believe that settles it. The league has not released the methods, the covariates, or the null results in any form a team analytics department can pressure-test. Commissioner Adam Silver said last summer the majority of Achilles injuries over the prior decade hit before the All-Star break, which undercuts simple “too many games” explanations. He also floated AI ingesting every frame of game film to hunt for patterns humans miss. Fine. Publish something. A working group and a private finding of “no relationship” is not the same as transparent science when stars are still exploding tendons on national television.
Silver put it plainly: “We’re taking it very seriously.” Taking it seriously means letting independent researchers see the full workload and tracking data, not just issuing a statement that the UCSF sample is limited. Of course it is limited. Nineteen events in a league that plays 1,230 regular-season games a year will always be limited. That is why the spike finding is interesting rather than dispositive. It generates a testable hypothesis about acute load, not a season-long average that every trainer already tracks.
GMs Should Treat the Ramp, Not the Total
Here is the practical read for anyone running a roster. If the only dial you turn is total minutes or games played, you are fighting the wrong battle. The paper’s signal lived in the 10-to-15-game window. That is a coaching and medical-staff problem more than a CBA rest-policy problem. Monitor the slope. When a guy who has been sitting 26 suddenly logs 32-plus for two weeks because the rotation tightens or the playoffs arrive, that is the moment to intervene, not the March back-to-back in Charlotte.
Study co-author Dr. Nirav Pandya put the stakes in plain language when the paper dropped: the minutes spike showed up even when season-long workload looked normal, and seven tears in one season made it a real, growing problem. The NBA can keep insisting its internal review saw nothing. Until that review is public, the UCSF numbers are the best independent look anyone outside the league office has. Nineteen cases is not a law of physics. It is a flashing light on the dashboard. Ignore the short-term ramp at your own peril. The next star who goes down will not care that the league’s epidemiologists found no relationship.
