A flashpoint of controversy in some officer-involved shootings is when officers do not immediately cease fire the moment a deadly threat ends and they are no longer in mortal danger.
An officer’s ability to instantly stop pulling the trigger once a “stop shooting” signal becomes evident is not always considered. Instead, the officer behind the gun may face harsh media criticism and daunting legal action alleging deliberate excessive force for firing “unnecessary” extra rounds.
This is a conundrum that the Force Science Institute has explored in pioneering research, and a review of its findings is published in the current issue of Law Enforcement Executive Forum, a peer-reviewed journal.
The report, authored by FSI’s executive director Dr. Bill Lewinski, Dr. William Hudson, dean of the College of Engineering, Mathematics, and Science at the University of Wisconsin-Platteville, and Jennifer Dysterheft, a Force Science research associate and doctoral candidate at the University of Illinois, focuses primarily on four human-perception, decision-making, reaction-time experiments conducted with 102 experienced LEOs in Arizona.
“Our findings, obtained under stressful but nonthreatening laboratory conditions, comprise a starting point for understanding the human dynamics involved in promptly concluding a shooting episode,” Lewinski told Force Science News. “They very clearly illustrate the challenges of responding instantaneously to a rapidly changing situation.
“The infinitely more complex circumstances of a real-world, life-threatening gunfight are likely only to magnify what our officer volunteers experienced.”
In the experiments, the officers one at a time were equipped with nonfiring 9mm Glock training guns that were rigged so that trigger pulls could be precisely timed to thousands of a second. In a training room, they then faced a 3X3-ft. “stimulus board” studded with nine clusters of colored LED lights that could be remotely activated by computer in unpredictable patterns of increasing complexity. Each officer responded with five “trials” to each of a series of four tests as monitors measured their trigger-pull reaction times.
To establish a simple typical reaction time, officers were instructed to watch a specific cluster of lights on the board and when a green light came on, they were to pull the trigger once, “as quickly as possible.”
The fastest time between the light flashing on and an officer beginning to move the trigger was 0.17 second, with the slowest being half a second. The average time to perceive the change cue and initiate trigger pull was 0.25 second. This is starting with the officer’s gun already aimed at the threat, with the officer’s finger on the trigger and the officer primed to respond.
For these five trials, the officers were told to begin “shooting” as quickly as possible when the green light came on and to “continuously pull the trigger” as they might in an actual gunfight until the light blinked off, representing an end of threat. Then they “must stop instantly” or their “score” would be penalized. The duration of the shooting time was randomly varied among the trials.
Responding to this simple stimulus, some officers were able to stop immediately, but the slowest to stop completed six more trigger pulls after the light went off before releasing the trigger for good. On average, officers shot one more round and started a second trigger pull that would likely be completed in a real-world situation after the “threat” stopped.
Officers were to watch a full row of light clusters, which consisted of three bulbs each. If only one or two lights in a cluster came on, the officers were not to shoot. Only when a complete cluster was simultaneously illuminated were they to fire.
This relatively simple increase in the complexity of decision-making roughly doubled reaction times. Now, on average, 0.56 second passed between the time a full cluster lit up and the officers initiated a trigger pull.
In the final and most complex trials, officers were to focus on the entire stimulus board. They were to pull the trigger “as quickly as possible” once all the green lights in any row were lit. As distractions, yellow and red lights in the clusters might turn on or the green lights in a row might not all be lit.
The average reaction time to start shooting–0.46 second–actually improved slightly for this experiment. The officers learned to “anticipate a pattern evolving and simply had to recognize that pattern,” the researchers explain.
The decision-making in the experiments was the “simplest possible” compared to the challenges facing officers in real-world deadly force encounters, the researchers point out. In street confrontations, LEOs must deal with “a multitude of stimuli; ambiguous circumstances; poor ambient light; and a complex, dynamic, and often evolving threat situation”–all of which will tend almost inevitably to impact on an officer’s ability to rapidly evaluate options and react to contextual changes.
“It is always expected that officers perform at expert levels of shooting,” Lewinski says. “If they fire excess rounds or make any mistakes, they are highly criticized and held accountable. Yet this study suggests, among other things, that many officers may be unable to cease firing instantaneously when the suspect is no longer a threat.
“Everything an officer does takes time. It takes time to perceive that a threat level has changed and it takes time to decide to stop shooting and to mechanically activate that decision. When officers are engaged in continuous rapid fire, as their training requires for defending their lives, the stopping process is more complex and generally takes longer.”
In their paper, the researchers note that “if an officer were to take [merely] 0.56 seconds to react to a stop-shooting signal, three to four [extra] rounds could be fired by the officer as an automatic sequence after the signal to stop had already occurred.” The slower an officer’s reaction time, “the greater number of shots [can] be fired before a conscious stopping can occur.”
The researchers also comment on the number of mistakes officers made during Tests #3 and 4. In Test #3, 3% of rounds fired were “false positives”; that is, officers misread the stimulus and fired when they shouldn’t have. “That number more than doubled [to 8%] with the addition of pattern recognition” in Test #4.
“This directly translates into officer-involved shootings, suggesting that with complex decision-making components, in addition to movement patterns, there is nearly a 10% risk of officers making false positive errors or shooting when the pattern appears to represent an evolving threat but in reality it never reaches that point.”
What suggests even more physical danger for officers is the number of false negatives that occurred in the tests, “when the officers did not shoot when they should have.” This represented only 1% of the trials in Tests #3 and 4, but on the street the consequences could have been grave, giving “deadly suspects the advantage” and putting officers’ “lives at risk.”
Among other troublesome aspects of “extra shots” incidents, the researchers also address controversial complications that often arise from rounds fired at moving vehicles and at suspects who are falling to the ground after already being hit.
“Before analyzing real-life shootings, it is necessary to understand the basic reaction times and other data recorded in this study,” Lewinski says. “The principle purpose of the study was to create a foundation of such knowledge. We anticipate conducting more sophisticated testing of pattern recognition and decision-making in the future.”
The full study, titled “Police Officer Reaction Time to Start and Stop Shooting: The Influence of Decision-making and Pattern Recognition,” will soon be available on the journal’s website as well as on the Force Science Institute site. We will make an announcement once the study is posted.
The study findings and their implications for investigators and use-of-force reviewers are discussed in detail during the five-day certification course in Force Science Analysis.