Did the NFL Make Players Safer in Practice—and More Vulnerable on Sundays?

Two policy changes need separating. The 2011 collective bargaining agreement cut five weeks from the offseason program, reduced organized team activities from 14 to 10, eliminated traditional two-a-day padded practices, and capped regular-season padded sessions at 14. This was the first major turning point in modern restrictions on team practice contact.

Then came the 2020 agreement. With the transition to the 17-game regular season beginning in 2021, training-camp padded sessions were limited to 16, down from the previous limit of 28, with a five-day acclimation period. The regular season expanded from 16 to 17 games; the preseason moved from four games to three. Those simultaneous changes make attribution difficult: fewer exhibition games, another competitive game, changing coaching practices and a new ramp-up regime all arrived together.

THE THREE-DECADE FRAME / POLICY TIMELINE

1996–2010 Historical comparison era, before the 2011 CBA restrictions. Comprehensive, consistently coded soft-tissue rates have not been established here.

2011–2019 First reduced-contact CBA era. Useful pre/post analyses exist for subsets of these years.

2020 Exceptional pandemic season: no preseason games, substantially altered team activities and an unusually disrupted preparation calendar.

2021–2025 Five completed 17-game seasons under modern training-camp constraints.

This is why a comparison of the five most recent completed seasons against the preceding 25 seasons is a useful editorial frame, but not automatically a sound statistical experiment. The rules changed more than once, and the methods for recording injuries did not necessarily stay constant.

COVID-19 canceled the NFL's exhibition schedule in 2020. The league also altered camp and offseason access. Players still had to make the transition to full-speed regular-season football, but without the usual sequence of preseason games. This made 2020 an unusually informative—but deeply confounded—test of the preparation hypothesis.

In a 2022 retrospective study, Sclafani and colleagues compared publicly reported soft-tissue injuries in the 2018 and 2020 regular seasons. Their reported hamstring injury rate rose from 5.31 to 9.98 per 1,000 athletic exposures, an approximately 88% relative increase. Groin injuries rose from 2.46 to 5.56; calf injuries from 1.61 to 4.08. The study also identified statistically significant increases in quadriceps and thigh injuries. These were reported injury-rate estimates, not evidence that any specific CBA provision caused the differences.

Another 2022 study, examining 2019 against 2020, reported 632 versus 765 musculoskeletal injuries in its defined dataset. Its counts for calf/Achilles strains increased from 17 to 39, and for groin injuries from 25 to 46; both comparisons were statistically significant. Hamstring counts increased from 87 to 114, but that comparison did not reach the study's statistical significance threshold. The two papers used different ascertainment methods and reference seasons, so their injury counts should not be merged.

THE EVIDENCE / 2018 VS 2020

There are at least three separate exposures that the football argument often compresses into one: general physical conditioning, football-specific running and movement, and full-contact tackling or blocking. A player might get strong and fast without absorbing repeated collisions. A player might also perform repeated contact drills without getting the high-speed running volume needed to prepare for competitive sprinting. The CBA limits certain types of team activity and padded practice; it does not prohibit players from maintaining physical fitness.

That difference is crucial when the outcome is a hamstring strain rather than a concussion, shoulder injury or fracture. Hamstring injuries often occur under high-speed running loads. A credible mechanism links insufficiently progressive exposure to those speeds with inadequate readiness, but evidence of a plausible mechanism is not an estimate of how many NFL injuries a policy produced.

It is possible that a program with less collision exposure but well-designed sprint progression makes players safer. It is possible that a rushed transition from controlled drills to maximal game speed raises muscle-injury risk. Both statements can be true, because preventing collision injuries and preparing muscle for competition are different problems.

There is a second distinction: the protective benefits of fewer practice collisions may be real even if a particular training program increases muscle strain risk. A serious safety evaluation would measure both sides, including head injuries and contact trauma, rather than judging an agreement only by its effect on hamstrings.

The requested comparison is 1996–2025: the most recent five completed NFL seasons against the preceding 25. The time frame is sound as a research ambition. The full comparative result has not been established in this investigation, however, because a uniformly coded, exposure-adjusted, publicly reproducible series of hamstring, groin, calf, quadriceps, Achilles and other soft-tissue injuries across all 30 seasons has not been assembled.

Raw injury counts cannot simply be lined up. League injury-reporting practices have changed, injured-reserve eligibility and roster management have evolved, more players are evaluated and documented, and 2021 introduced a 17th regular-season game. Comparing totals without accounting for games, plays, player exposures and changing reporting incentives can create an apparent rise that is not a rise in underlying individual risk.

The cleanest future analysis would track injuries per 1,000 player-games, player-hours or meaningful football exposures; compare the same diagnosed injury categories through time; separate practice events from games; and measure injuries in the first quarter of each season, when preparation is most directly testable. It would identify each injury as a new event rather than repeatedly counting weekly injury-report entries. The ideal analysis would also adjust for position, age, roster changes, weather, field surface, cumulative workload and changes in rules.

An especially valuable natural comparison would separate teams or position groups with different objectively measured preseason sprint loads and contact exposures, while checking whether the benefits of preparation outweighed injuries suffered during preparation itself. Without those inputs, attributing a 30-year injury trend to the CBA is beyond what the evidence allows.