What research exists on number-of-rounds fired in police shootings and how training influences that outcome?
Executive summary
Research on how many rounds police fire in shootings and whether training changes that pattern is substantial but fragmented: large-sample analyses and narrative reviews show low hit rates and that multiple-officer incidents correlate with more rounds fired [1] [2], while experimental and training-focused studies suggest realistic, decision-focused training can reduce inappropriate firing and improve shoot/don’t-shoot decision-making [3] [4]. The evidence points to training design—scenario realism, decision-focused drills, and human-performance principles—shaping both accuracy and the propensity to continue firing, but methodological limits and mixed findings leave firm causal claims tentative [5] [6].
1. The macro picture: hit rates, rounds-per-incident and the multiple-officer effect
Longitudinal and department-specific analyses repeatedly report that officers often fire multiple rounds and that overall hit rates are lower than public expectation; studies of officer-involved shootings found hit rates commonly under 50 percent and, in some samples, about half of incidents produced at least one hit [1] [5] [7]. Research going back decades documented typical strings of fire rising with added shooters—single-officer incidents averaged roughly 3.6 rounds, two-officer incidents near 5.0 rounds, and three-or-more shooters averaged over six rounds—establishing a clear correlation between the number of shooters and total rounds fired [2]. These aggregated findings provide the empirical backdrop for concern about “wayward rounds” and bystander risk [1].
2. Contagious fire and human reaction limits: why volleys happen
Behavioral and forensic work highlights two mechanisms that raise rounds-per-incident: “contagious fire,” where officers feel compelled to shoot after colleagues do, and sensorimotor delays that make it physically impossible to stop firing instantly once a stop decision is signaled; one experimental simulation found participants still fired an average of ~2 extra rounds after a clear stop cue [8] [9]. Experts and reviewers cite these human-performance constraints to explain why volleys of shots can appear reflexive rather than deliberative, with important implications for training and policy [9] [5].
3. What training studies say: realism, decision-making, and transfer to the street
Intervention studies and reviews converge on the idea that training that mirrors operational complexity—moving targets, time pressure, stimulus discrimination, and force-on-force drills—improves decision-making and marksmanship under stress more than static range practice [3] [5] [4]. A recent video-based, 4C/ID-inspired intervention improved shoot/don’t-shoot decisions in a lab experiment by training inhibition and gaze control rather than purely mechanical shooting skills [3]. Narrative reviews find multiple trainable factors (grip, stress inoculation, realistic scenarios) that can raise accuracy and reduce inappropriate firing when properly implemented [4].
4. Simulation and transfer: promising but limited evidence
Simulation and simulator-based training often produce better short-term performance and higher hit rates in controlled tests, and some field exercises show increased shot discipline after simulation—but many evaluations rely on small samples, single departments, or simulator outcomes that may not generalize to chaotic, real-world encounters [6] [10]. Meta-level critiques warn that traditional static qualification courses may create paired responses that fail under stress, supporting moves toward human-performance-centered curricula—but causality from training change to reduced rounds fired in real incidents remains imperfectly documented [5] [11].
5. Policy implications, dissenting views and research gaps
Policy recommendations from reviews and policing institutes push for scenario-based, decision-focused, and bonded-team training to reduce unnecessary volleys and improve accountability, while some practitioners resist wholesale changes citing cost, logistics, and the difficulty of replicating lethal stress [12] [13]. Significant gaps persist: many datasets cover only fatal shootings or single departments, experimental studies use lab or simulator proxies, and longitudinal causal evidence tying specific trainings to measurable reductions in rounds-per-incident in field shootings is still sparse [1] [6]. Until larger, multi-site, longitudinal evaluations are performed, conclusions must acknowledge promising mechanisms without overstating definitive proof [5].