It’s crazy to think that it’s 2017, and even with all the protests physically and digitally, people being killed by cops is still happening consistently. And guess what else? Apparently, the numbers of those killings aren’t even being recorded accurately.

U.K.-based newspaper and media company, Guardian, created a project called “The Counted” to “count the number of people killed by police and other law enforcement agencies in the United States throughout 2015 and 2016.” The federal Centers for Disease Control and Prevention (CDC) also does the same thing in America, but not very well though. “The Guardian counted 93 percent of the U.S. police-related deaths, while the federal Centers for Disease Control and Prevention (CDC) statistics counted only 45 percent,” according to the Huff Post.

Interestingly, the Guardian relied on news stories and crowd-source information for their count, while the CDC’s source was official records, which seem to often omit police killings. Before you get alarmed, the killings appear to be omitted mostly by mistake. A medical examiner or coroner often failed to mention law-enforcement involvement on the death certificate, and police killings by weapons other than guns (for example, Taser shocks) were also missed.

In 2015, there were an estimated 1,166 deaths. The America’s CDC missed 643, while the United Kingdom’s publication, Guardian, only missed 80.

Accuracy of police-related killings are, obviously, important, something “The Counted” senior author Nancy Krieger, also a professor at Harvard T.H. Chan School of Public Health, made very clear. “As with any public health outcome or exposure, the only way to understand the magnitude of the problem, and whether it is getting better or worse, requires that data be uniformly, validly and reliably obtained throughout the U.S.,” the professor stated.

Hopefully this will motivate U.S. data organizations to pay more attention and carefully gather data in the near future. How will the immensity of killings by the hands of police ever stop if the precise number of deaths continue to be inaccurate?