Although the most dangerous cities are mostly well-known: the perennial top ten of Detroit, St. Louis, Baltimore, Memphis etc. when you get more granular the results change pretty dramatically.
From Forbes, Jan 28:
While data shows that crime is actually down in the U.S. in recent years, many Americans believe that the country is becoming more violent than ever—and high-profile incidents like mobs rioting at the U.S. Capitol and the Department of Homeland Security issuing a threat bulletin to warn of ongoing potential for domestic violence only underscore those fears. So the timing of a new study from the risk assessment app Augurisk Now, revealing the country’s 10 most dangerous neighborhoods, couldn’t be better.
Augurisk Now—which launched last month—helps people track the likelihood of crime and national disasters, making it a useful tool for everyone from travelers to concerned citizens to potential homebuyers. The way it works: Whenever you enter a dangerous block in the U.S., the free app uses proprietary risk-scoring algorithms to alert of potential dangers, warning users that they have entered a predicted or observed high crime area. Similarly, the app can predict risks when it comes to floods, earthquakes, wildfires and storms.
“Many years ago, while I was searching for a new property that would be safe from floods or earthquakes, I was surprised no service provided such essential information in an intuitive form,” says founder Mohamed Mezian. “That’s where the idea for Augurisk emerged—our main mission is to help people and businesses better prepare for the future.”
In order to zero in on the 10 most dangerous neighborhoods in the United States, Augurisk analyzed its data and issued a report on the riskiest places by looking at Census block groups, a statistical division of tracts that usually contains 600 to 3,000 residents. Augurisk then measured each Census block group by predicting violent crime occurrences and violent crime rate (per 100,000 residents). “The data underlying this crime analysis are taken from the predictions of our proprietary machine learning crime prediction algorithm,” says Augurisk’s lead scientist Simon de Bonviller. “Other contextual factors presented in this article were adapted from the American Community Survey 2014-2018 5-Year estimates.”
Surprisingly, some of the cities in the U.S. that have recently been named most-dangerous places—like Detroit and Memphis—didn't make this Augurisk list. According to de Bonviller, one explanation for this relates to the resolution considered. “We focus on block groups— i.e. neighborhoods—while such rankings consider crime statistics in the entire city (or law enforcement agency),” he says. “However, one risky neighborhood doesn’t necessarily make a whole city dangerous, and vice versa.”
Another explanation: Augurisk trained a machine learning algorithm to predict crime occurrences based on 188 predictors (including socioeconomic, demographic, climatic and law enforcement variables, among others). “This algorithm was trained using the exact location and type of crimes committed in 11 U.S. cities between 2014 and 2018, including Detroit,” says de Bonviller. “The result was used to generate this ranking. Thus, we use predictions of the crime expected by our algorithm given 188 predictors related to a place, not past crime—as in the mentioned rankings—which can explain some differences.
Nevertheless, de Bonviller says that the he and team at Augurisk were surprised by the results—especially by the high number of dangerous neighborhoods in California. “Initially, when we were exclusively focusing on crime rate, the most dangerous places in the U.S. were found to be parks, airports and other busy places with low to no population. If the population is low, then crime risk can be overestimated when using such an indicator. Subsequently, we therefore used the number of crimes committed as an additional criterion. Moreover, we were surprised to notice the high occurrence of block groups located on the West Coast in the ranking, including in Los Angeles and San Francisco.”
Read on for more details about the most dangerous neighborhoods in America. For people who are moving from one city to another, Augurisk’s public county risk assessment can provide free insights into their new county, while people acquiring a new property can use the Augurisk website to generate a free natural disaster and societal risk report, to make sure the property is not at risk.
1. Los Angeles: Block group delimited by E. 5th St., S. Los Angeles St., Boyd St., and S. San Pedro St.
Predicted violent crimes per year: 239
Predicted violent crime rate: 347 crimes per 1,000 residents
2. Los Angeles: Block group delimited by Santa Monica Fwy., Maple Ave., E. 9th / S. Los Angeles St., E. 7th St., Gladys Ave., E. Olympic Blvd., and S. Alameda St.
Predicted violent crimes per year: 407
Predicted violent crime rate: 293 crimes per 1,000 residents
3. Los Angeles: Block group delimited by E. 7th St., S. San Pedro St., E. 5th St. / S. Central Ave. / E. 4th, and S. Alameda St.
Predicted violent crimes per year: 456
Predicted violent crime rate: 291 crimes per 1000 residents
4. San Francisco: Block group delimited by Division St. / 9th / Brennan / 10th / Bryant / 11th St., Harrison St., 6th St., and Townsend St.
Predicted violent crimes per year: 296
Predicted violent crime rate: 298 crimes per 1,000 residents
5. Kansas City, Missouri: Block group delimited by E. 8th St., Prospect Ave., E. 12th St., and Cleveland Ave.
Predicted violent crimes per year: 182
Predicted violent crime rate: 304 crimes per 1,000 residents
.....MUCH MORE
All that being said, we should remember:
