Killings in New York City have jumped 350 percent compared to this time last year, the New York Post claimed yesterday. It’s a classic Post argument: bloody, provocative, and utterly false when divorced from its own narrow and purposefully misleading premise.

Under the headline “Killings in NYC spike,” a team of three Post reporters conveyed the impression that the city is in the throes of a historically murderous period—surely thanks to mayor Bill de Blasio, the publication’s enemy number one. Their argument appears legit at first glance: During the week ending October 31, 2015, there were seven killings in the city, compared to just two slayings over the same seven-day period in 2014—less than a third as many.

But comparing murders week-to-week puts an arbitrary window on the data and tells us hardly anything at all. For whatever reason, last week was a particularly bloody one, and the Post cherry-picked it in the service of making New York City look violent and in need of saving—the editorial project it has labored over since the 1970s, which has reached surreal new heights under de Blasio’s mayoralty. It’s true that murders are up from last year, but the “spike” is close to six percent, not 350.

To illustrate the feebleness of their argument, let’s look at the NYPD’s most recent weekly crime report, which documents the seven-day period ending October 25, 2015—shifting the article’s window backward by six days. This seemingly minor tweak produces startlingly different results: during that week, there were five murders, not seven, and over the same period in 2014, there were 10. A rag with a different agenda could use these numbers and produce a totally different headline: “Murders in NYC down by half.”

NYPD records for years past similarly deflate the implication that killings are skyrocketing. In 2013, there were eight murders during the week ending November 3, and in 2012, there were two, a police spokesman who was miffed about the Post article told me. (The department’s weekly reports run from Monday to Sunday, making this the closest available seven-day period to the Post’s.)

Judging by the last week of October alone, you’d think the murder rate in New York has peaked and valleyed wildly in recent years: it leaped by 400 percent from 2012 to 2013, floored again from 2013 to 2014, and went back up by 350 percent this year. These percentages are misleading in two ways: they highlight unrepresentative data and visually inflate what are actually pretty small numbers. In this case, a 350 percent increase means just five people. When you zoom out and look at an entire year’s data, the number of homicides in New York City has declined steadily every year since 2010. New York may end 2015 with more homicides than it had in 2014, but it will take a major crime wave to beat the number from a mere three years ago.

That the city saw seven murders last week is worth noting, but statistically, it means almost nothing when considered in context. Fortunately for the Post’s reporters, meaning and context are not high on their employer’s list of priorities.

Photo via AP. Chart by Jim Cooke.