John McAfee: Dark web data shows 10% of fatal US police shooting victims in 2015 were unarmed

On 5 November 2015 in Marksville, Louisiana, officers Norris Greenhouse Jr. and Derrick Stafford shot and killed a six-year-old autistic boy while he was strapped into his seat in a car. The officers fired six bullets into the boy's head and upper chest. The boy's father, unarmed and with his hands held over his head, was also shot and critically injured.

The officers said that they were trying to serve a warrant, that the father had tried to run them down, and that the father had fired a gun at them. Video from a third officer at the scene proved that every aspect of that testimony was false. State records further showed that there were no warrants whatsoever that had been issued for the father.

What is the dark web?

The dark web is a section of the internet that requires specialist software tools to access, such as the Tor browser. Originally designed to protect privacy, it is also associated with illicit activities.

It is often confused with the deep web, which is a vast section of the surface internet not indexed by search engines such as Google. The deep web comprises around 95% of the internet.

I wish that I could say that this is an isolated and extreme circumstance, hopefully never to be repeated, but unfortunately such is not the case.

There does not exist, in America, a National Police Shootings Database, not in the FBI, nor the NSA (which, probably due to the overwhelming demands of spying on American citizens, does not have the time nor manpower to address such issues), nor in any other agency within our government. The scale of such incidents remains obscure at the best.

But the Internet underground – the deep web and the dark web – contains, believe it or not, detailed records and sometimes analyses, of almost every event that has ever happened on the face of this planet since approximately 2003. If you know how to find it. So I went to the Internet Underground.

The beauty of us hackers, both black hat and white hat, is that sometimes we do hard work simply in the anticipation that someone, somewhere, at some time, will find a use for it. I don't know why this is. Maybe our social lives lack a certain zest, or our collective conscience is different than the average citizen. Maybe it's the challenge. I myself don't know. I only know that it's true. Operating under this assumption, I quickly located a number of databases, created by hackers, relevant to police shootings.

I wanted more than just the number of shootings. I needed names, ages, locations, race, whether the suspect was unarmed or not, whether the suspect was in their right mind or not, what the suspect was armed with, if armed, etc. There was so much data in the dark web on this subject, that it was unmanageable. Some files listed all shootings and abuses, whether by handgun, Taser, bean bag, projectile tear gas, paintball guns, rubber bullets, fists, batons, etc.

Others listed only certain regions or only certain races, or certain age groups. The task was too large for my tiny resources, so I settled on shootings that resulted in death, for which, according to Internet Underground sources, there were just shy of 1,000 so far this current year. A list that size I can deal with. My subsequent analysis of that list is revealing.

As an interjection at this point, the oddest story appeared to be a man, armed only with a paper stapler, who was shot and killed by police in Polk County, Florida, in September. I have accidentally stabbed myself with a stapler a few times, and while painful, it never occurred to me that I might be able to intimidate an armed police officer with it. But, apparently, I was remiss.

Death by numbers

There were 879 fatal shooting by US police from 1 January 2015 to 25 November 2015. Of these 879, 82 victims were completely unarmed. But this does not tell the full story. Of the 797 armed suspects, 375 were armed with weapons that were no match for a gun. These weapons ranged from box cutters and screw drivers to beer bottles and cordless drills.

Of those who were armed, 13 were over the age of 70, and 2 were over the age of 80 - one of whom could barely walk and the other could not see. There were 19 minors who were killed, and 62 of those killed were under the age of 21. Four of those killed were under the age of 15.

There were 224 Blacks that were killed, 418 Whites, 143 Hispanics, and 11 Asians. The police did not identify the race of 83 of the killed. 846 males were killed compared to 33 females.

The most telling statistics came when we analysed killings by state. There were a number of glaring anomalies. Oklahoma, which ranks 5th in number of police killings, ranks number 28 in state population size. Arizona ranked number 4 in police killings, yet number 15 in state population. Colorado ranked number 8 in fatal police shootings, yet number 23 in state population. New York, oddly, ranks number 2 in state population but ranked way down at number 17 in the number of fatal police shootings.

These numbers are huge statistical anomalies. Most of the rest of the states fell within one or two rankings of their position in both number of shootings and population size. I cannot explain the reason for these anomalies, but it is clear that New York is to be applauded, while Oklahoma, Arizona and Colorado should be avoided if at all possible, if you look or act in a suspicious manner.

I would like to thank Jaque Donahue, my lead data analyst, for help in acquiring and analysing this data.