The triumph of Black Lives Matter got countless black lives murdered and splattered on the asphalt by lying about the death toll. The NYT continues the lying in this graph.
Another problem is that this is not a graph showing the absolute number. Even though it's low key designed to give that impression. What it is is the percent difference using 2015 as the zero point baseline. In which case, they would have done better to use 2013, the final full year pre-Ferguson, for the zero point.
Tacky to self-respond, but I think I can ring this out.
No matter where you look within a relevant range of recent history, there are way many more homicides than police killings.
Let's say that 100 people died of a lightning strike last year and 200 this year. And 10 million people died of heart disease last year and only 9 million this year. This means that in a NYT style graph of this, lightning strikes are up 100%, heart disease down 10%. A lot of people would interpret that there are way many more people all of a sudden dying from lightning strikes compared to heart disease.
So the WaPoo collected the data and didn't use it when it was relevant. Guess they weren't willing to misplace St. Floyd's Day like the NYT did or tell the truth as Steve does.
True. Also, shootings, rather than homicides, offer an even starker picture. As do the rise in non-gun assaults by random individuals or mobs. I miss the prescience of Colin Flaherty and wish every city had websites like Hey J**k**s, which offers accurate Chicago shooting statistics; Second City Cop, and CWB Chicago, citizen journalism that follows cases through the courts.
2020 was such a crazy time. When it was happening, the George Floyd hysteria just kind of felt like it was part of the same continuation of craziness, so I personally didn't really think of it rationally, it was all part of, "All of this is nuts!"
But looking back, the George Floyd hysteria was a special kind of crazy all on its own. A convicted criminal died during an arrest for yet another crime, it was all very clearly an accident due to his own conduct and his own poor health, yet the whole world lost its frickin' mind because... why? And all four cops went to jail, because... why? And cities were burned down and destroyed and thugs took over sections of some of them for weeks at a time because.... why?
And it was all - covid, BLM, the Great Cancelation, Me Too - because they hated Donald Trump so much and couldn't stand to lose their power and their privilege to the candidate of the unwashed masses.
The election in 2024 was partly a reaction to all of that, but there are still plenty of people out there who think some kind of outrage occurred to this criminal, and somehow the entire country was justified at blowing up and tearing itself apart over it? I just don't get it. A saner society would say to itself, "Well, too bad he died, but at least that POS is no longer around to cause anyone else trouble".
I see the WNBA celebrated George Floyd here within the last couple of days. The same George Floyd that (may have) held a pregnant woman at gunpoint. You'd think that women would take such transgressions personally. Then again, WNBA = LNBA, so I don't think they're that personally invested in women being in the family way.
I think black women in general have demonstrated that they have an incredibly high tolerance for dysfunctional behavior from black men. Toss in the symbolism of Floyd and how it’s viewed as a mechanism for racial leverage and they’d overlook anything.
The graph represents the same sort of argument that progressives made on inflation and violent crime generally. It goes like this: for the past several years I have punched you in the face, kicked you in the nuts, and pissed on you when you were down. Last month, I stopped pissing on you. See how much I have improved your life.
The label on the graph is difficult to understand unless you know the numbers. I think it means killings relative to 2015. Taken more literally it would mean that killings were higher than the year before, every single year since 2015.
It looks to me as if the NYT got the line denoting 5 May 2020 right but deliberately shifted the graph of deaths to the left by over a year. In any event it's just another NYT lie.
"Notice that they put the vertical line denoting the effect of the beginning of the George Floyd racial reckoning on May 25, 2020 on homicides in the wrong year: between 2020 and 2021 when it should, of course, be between 2019 and 2020?"
That is so underhanded and morally inexcusable. Sometimes I feel like I'm the only one who sees it when they do this stuff. This is why I've been a faithful reader for almost 30 years -- it's comforting to know I'm not alone.
What level of hell would Dante put these people in?
I really hope some psychologist or historian studies these last ten years. The lying has been so blatant and consistent in an era where the access to information has been unprecedented in human history. Maybe that’s the paradox the more information we have and the more data we have and the more experts we have the easier it is to pull off the big lie.
Even reading your explanation here I had to look back at the graph several times to understand your point. NYTimes readers will have no reason to look twice and think about the line placement.
In happier times, we could have assumed someone at the NYT just looked at the year and flubbed the timeline. In the immortal words of Inspector Clouseau, "Nut any more."
>Wow, it’s almost as if the triumph of Black Lives Matter on May 25, 2020 subsequently got a whole lot of Black Lives Murdered and Splattered on the Asphalt.
But the New York Times is not going to tell you that, even after 5 years of BLM induced mayhem.<
Correction:
"But the New York Times is not going to tell you that, even after 5 years of BLM hysteria and mayhem that they helped to whip up with intentional lying."
I realize journalists don't tend to be the most math savvy people on the planet. But c'mon, the people the NYT hires presumably went to decent schools and had to stumble through the SAT with at least a 600+ math score. These aren't the kids I tutored at my kids' HS who'd failed the WA state math exam and needed help to get their diploma.
The sheer brazen lying here is through the roof. But hey, I guess that's nothing new. They still right articles that saying things like "since the murder of George Floyd ..."
I read the NYT story, and Sailer's commentary on the graph could be generalized to the entire piece. No individual statements are provably untruthful, but the tapestry is woven so as to mislead the reader.
Steven Rich, Tim Arango, and Nicholas Bogel-Burroughs are the reporters credited with the story. They will doubtlessly do (relatively) well in their dumpster-fire of a profession.
.
Me: Describe Joseph Goebbel's position on truth and falsehood in effective propaganda.
.
ChatGPT: Goebbels' position can be summarized as follows:
* Truth is useful only if it serves the cause: Goebbels stated that propaganda does not need to be truthful—it needs to be effective.
* Lies are acceptable if they advance the goal: He endorsed manipulation of facts, distortion, and outright falsehood if they promoted the regime’s objectives.
* Partial truths are most powerful: He often used selective truth—cherry-picking facts and stripping them of context to create a misleading but emotionally compelling narrative.
* Big Lies are more believable: Goebbels held that a colossal falsehood, repeated often enough, would be more easily accepted by the masses than a small one.
In essence, for Goebbels, truth was subordinate to utility—what mattered was the persuasive impact, not factual accuracy.
The percent change in police killings seems relatively insensitive to the percent change in homicides, which is interesting. Have you (Mr. Sailer) ever examined this relationship?
There’s a noticeable uptick at 2022-23. The percentage increase in illegal border crossings in those years was huge relative to 2021. A connection there seems plausible, although NPR has assured me that they were all peaceful migrants and blameless refugees.
> May 25, 2020 … in the wrong year: between 2020 and 2021
> when it should, of course, be between 2019 and 2020
I don’t understand. Is this some kind of humorous point?
Or I miss some point about how the graph represents change, so… The change caused by 2020 should be visible in 2021. But what is this mention of 2019? I’m confused.
#1: The data reported as "2020" refers to homicides that occurred from Jan 1, 2020 through Dec 31, 2020. So the 2020 label on the graph could read "12/31/2020" to be more clear. Floyd died in May of that year, so the line for his death should be shown between the data points labeled 2019 (12/31/2019) and 2020 (12/31/2020), if we wish to draw conclusions as to whether his death had anything to do with deaths reported in 2020 and subsequent years.
#2: Yes, Covid also green. In the past I've provided Steve with Excel charts and the colors were garbled when he posted them. It seems to be a some sort of software incompatibility, perhaps between Excel and Substack. So you just have to roll with it when it occurs.
> The change caused by 2020 should be visible in 2021. But what is this mention of 2019? I’m confused.
The x-axis of the NYT's graph (2015, 2016, ... , 2023, 2024) seems to be about one thing (years, duh!). But it's actually about two.
For "Death of George Floyd, May 25, 2020," the point on the axis for "2020" represents Jan. 1, 2020, and the one for "2021" represents Jan. 1, 2021. So the dotted vertical line for George Floyd is placed to be at the 146th (of 366) day of 2020.
For "Homicides" and "Police killings," the point on the axis for "2020" represents "percentage change relative to Jan. 1 - Dec. 31, 2015, of deaths that took place between Jan. 1 and Dec. 31, 2020."
If the NYT had (more properly) depicted deaths as a bar graph, then the "2020" dot would be the upper left corner of the "2020" bar, and the "2021" dot would be "2020's" upper right corner.
Using six-month bars would have revealed the temporal correlation (Floyd's death preceded 2020's murder spike) that the Times' formatting conceals.
If only reporters Rich, Arango, and Bogel-Burroughs were aware of Sailer's work! Then they could have graphed "CDC figures for black homicide victimizations and black traffic fatalities by week from 2018 to 2023."
Their readers would have learned something true, and would have the tools to start thinking about *why* Floyd's death caused so many excess black homicides and traffic fatalities.
As a journalist of better character once quipped, "It isn't what we don't know that gives us trouble, it's what we know that ain't so."
FWIW, Steven Rich, clearly the data analytics member of the reporting team, and probably the senior member, offers the following bio, below. Turns out he is intimately familiar with the Washington Post data. Reporter #2, Tim Arango, is also an experienced hand, having reported international stories from the Near East for over a decade. So "rookie mistake" doesn't look like a good candidate for the misleading graphic.
Ken
==================
Steven Rich
I am a data reporter at The New York Times, using data analysis to investigate major issues and contextualize current events.
What I Cover
As a member of The Times’s data journalism team, I dig deeply into a wide range of subjects. I analyze data to identify trends, add quantitative context to stories, and help readers understand current events. This work facilitates both longer-term investigations and fast-paced stories on deadline.
My Background
Before joining The Times, I spent a dozen years as an investigative data reporter at The Washington Post, where I worked on stories across many topics. I developed particular expertise in data related to policing and opioids.
For 10 years, I helped to build and maintain a database of every fatal police shooting across the country in near-real time, an effort that was recognized with both a Pulitzer Prize and a Peabody Award. I mapped unsolved homicides in major cities to find neighborhoods where police departments had failed to solve the vast majority of murders, leaving victims and their families without full justice. My colleagues and I sued to get 14 years of data on shipments of prescription opioids across the United States. That data, which we made public, was used by hundreds of news outlets to examine opioid shipments to their communities.
I spent six years serving on the board of directors for Investigative Reporters and Editors, a nonprofit that helps journalists develop their skills. I am a graduate of Virginia Tech and the University of Missouri.
Journalistic Ethics
Like all Times journalists, I am committed to upholding the standards of integrity outlined in our Ethical Journalism Handbook.
I strive to use data to illuminate our world in a way that is accurate, transparent, and properly contextualized. I try to help readers understand the work behind my reporting, as well as its limitations. I maintain a healthy skepticism of everything and everyone, and I subscribe to the old journalism adage that says, “If your mother says she loves you, check it out.”
Thanks. Steven Rich certainly writes a catchy biosketch. In the context of this article, I am particularly taken with his three concluding sentences, which sum up his professional strengths and his aspirations:
"I strive to use data to illuminate our world in a way that is accurate, transparent, and properly contextualized. I try to help readers understand the work behind my reporting, as well as its limitations. I maintain a healthy skepticism of everything and everyone, and I subscribe to the old journalism adage that says, 'If your mother says she loves you, check it out.'”
Uh hunh.
If his article allowed reader comments (it doesn't), I would have quoted that paragraph.
Reading the companion "Five Years A Martyr" articles that do permit them, I note that a plurality of NYT readers do not (any longer?) appreciate being led around by the nose.
To answer your question on twitter, in 1968 Ray Oyler was such a bad-hitting shortstop that Mayo Smith moved Mickey Stanley there from CF for the Tigers in the World Series. While they won, it didn't work out for Stanley at SS so they traded for Tom Tresh from the Yankees in June
Another problem is that this is not a graph showing the absolute number. Even though it's low key designed to give that impression. What it is is the percent difference using 2015 as the zero point baseline. In which case, they would have done better to use 2013, the final full year pre-Ferguson, for the zero point.
Tacky to self-respond, but I think I can ring this out.
No matter where you look within a relevant range of recent history, there are way many more homicides than police killings.
Let's say that 100 people died of a lightning strike last year and 200 this year. And 10 million people died of heart disease last year and only 9 million this year. This means that in a NYT style graph of this, lightning strikes are up 100%, heart disease down 10%. A lot of people would interpret that there are way many more people all of a sudden dying from lightning strikes compared to heart disease.
The graph says it’s from WaPo. But the WaPo story on the fifth anniversary has no graphs.
So the WaPoo collected the data and didn't use it when it was relevant. Guess they weren't willing to misplace St. Floyd's Day like the NYT did or tell the truth as Steve does.
True. Also, shootings, rather than homicides, offer an even starker picture. As do the rise in non-gun assaults by random individuals or mobs. I miss the prescience of Colin Flaherty and wish every city had websites like Hey J**k**s, which offers accurate Chicago shooting statistics; Second City Cop, and CWB Chicago, citizen journalism that follows cases through the courts.
The graph is designed to be as sophistic as possible.
2020 was such a crazy time. When it was happening, the George Floyd hysteria just kind of felt like it was part of the same continuation of craziness, so I personally didn't really think of it rationally, it was all part of, "All of this is nuts!"
But looking back, the George Floyd hysteria was a special kind of crazy all on its own. A convicted criminal died during an arrest for yet another crime, it was all very clearly an accident due to his own conduct and his own poor health, yet the whole world lost its frickin' mind because... why? And all four cops went to jail, because... why? And cities were burned down and destroyed and thugs took over sections of some of them for weeks at a time because.... why?
And it was all - covid, BLM, the Great Cancelation, Me Too - because they hated Donald Trump so much and couldn't stand to lose their power and their privilege to the candidate of the unwashed masses.
The election in 2024 was partly a reaction to all of that, but there are still plenty of people out there who think some kind of outrage occurred to this criminal, and somehow the entire country was justified at blowing up and tearing itself apart over it? I just don't get it. A saner society would say to itself, "Well, too bad he died, but at least that POS is no longer around to cause anyone else trouble".
The left in this country is dangerously crazy.
I see the WNBA celebrated George Floyd here within the last couple of days. The same George Floyd that (may have) held a pregnant woman at gunpoint. You'd think that women would take such transgressions personally. Then again, WNBA = LNBA, so I don't think they're that personally invested in women being in the family way.
I think black women in general have demonstrated that they have an incredibly high tolerance for dysfunctional behavior from black men. Toss in the symbolism of Floyd and how it’s viewed as a mechanism for racial leverage and they’d overlook anything.
Well said. It was insane. And if you argued against any of it with logic or statistics, they’d just scream at you.
A saner society would laugh its ass off at trannies, women’s rights grifters, race hustlers, climate hoaxers, teacher unions….
Did you think fumbling down America is not intentional? That the Dept of Ed existed to improve education? Seriously?
Per an X post from Andy Ngo I saw reposted on Instapundit, today's WaPo article on Floyd states he was "shot and killed in police custody."
Still crazy after five years.
The bigger problem is this kind of stuff only gets blown up during presidential election years. I wish the majority would wise up.
Left behind again.
The graph represents the same sort of argument that progressives made on inflation and violent crime generally. It goes like this: for the past several years I have punched you in the face, kicked you in the nuts, and pissed on you when you were down. Last month, I stopped pissing on you. See how much I have improved your life.
And why arbitrarily index to 2015?
The label on the graph is difficult to understand unless you know the numbers. I think it means killings relative to 2015. Taken more literally it would mean that killings were higher than the year before, every single year since 2015.
The low point for homicides was 2014: during the first 7 months, nobody had heard of "BLM."
It looks to me as if the NYT got the line denoting 5 May 2020 right but deliberately shifted the graph of deaths to the left by over a year. In any event it's just another NYT lie.
The graph says it’s from WaPo. But I can’t find it.
Graph was made by NYT:
https://www.nytimes.com/2025/05/24/us/police-killings-george-floyd.html
It uses, like I did in my recent Takimag.com column, data from the Washington Post's big Police Shootings Database.
OK. Thanks.
The summer of 2020 showed how thin the thin blue line can become when the left is running amok.
"Notice that they put the vertical line denoting the effect of the beginning of the George Floyd racial reckoning on May 25, 2020 on homicides in the wrong year: between 2020 and 2021 when it should, of course, be between 2019 and 2020?"
That is so underhanded and morally inexcusable. Sometimes I feel like I'm the only one who sees it when they do this stuff. This is why I've been a faithful reader for almost 30 years -- it's comforting to know I'm not alone.
What level of hell would Dante put these people in?
I really hope some psychologist or historian studies these last ten years. The lying has been so blatant and consistent in an era where the access to information has been unprecedented in human history. Maybe that’s the paradox the more information we have and the more data we have and the more experts we have the easier it is to pull off the big lie.
Even reading your explanation here I had to look back at the graph several times to understand your point. NYTimes readers will have no reason to look twice and think about the line placement.
In happier times, we could have assumed someone at the NYT just looked at the year and flubbed the timeline. In the immortal words of Inspector Clouseau, "Nut any more."
>Wow, it’s almost as if the triumph of Black Lives Matter on May 25, 2020 subsequently got a whole lot of Black Lives Murdered and Splattered on the Asphalt.
But the New York Times is not going to tell you that, even after 5 years of BLM induced mayhem.<
Correction:
"But the New York Times is not going to tell you that, even after 5 years of BLM hysteria and mayhem that they helped to whip up with intentional lying."
I realize journalists don't tend to be the most math savvy people on the planet. But c'mon, the people the NYT hires presumably went to decent schools and had to stumble through the SAT with at least a 600+ math score. These aren't the kids I tutored at my kids' HS who'd failed the WA state math exam and needed help to get their diploma.
The sheer brazen lying here is through the roof. But hey, I guess that's nothing new. They still right articles that saying things like "since the murder of George Floyd ..."
I read the NYT story, and Sailer's commentary on the graph could be generalized to the entire piece. No individual statements are provably untruthful, but the tapestry is woven so as to mislead the reader.
Steven Rich, Tim Arango, and Nicholas Bogel-Burroughs are the reporters credited with the story. They will doubtlessly do (relatively) well in their dumpster-fire of a profession.
.
Me: Describe Joseph Goebbel's position on truth and falsehood in effective propaganda.
.
ChatGPT: Goebbels' position can be summarized as follows:
* Truth is useful only if it serves the cause: Goebbels stated that propaganda does not need to be truthful—it needs to be effective.
* Lies are acceptable if they advance the goal: He endorsed manipulation of facts, distortion, and outright falsehood if they promoted the regime’s objectives.
* Partial truths are most powerful: He often used selective truth—cherry-picking facts and stripping them of context to create a misleading but emotionally compelling narrative.
* Big Lies are more believable: Goebbels held that a colossal falsehood, repeated often enough, would be more easily accepted by the masses than a small one.
In essence, for Goebbels, truth was subordinate to utility—what mattered was the persuasive impact, not factual accuracy.
That makes them may be a millimeter above the WaPo in the media sewer.
Per an X post by Andy Ngo, reposted on Instapundit, today's WaPo article on Floyd states he was shot and killed in police custody.
The percent change in police killings seems relatively insensitive to the percent change in homicides, which is interesting. Have you (Mr. Sailer) ever examined this relationship?
There’s a noticeable uptick at 2022-23. The percentage increase in illegal border crossings in those years was huge relative to 2021. A connection there seems plausible, although NPR has assured me that they were all peaceful migrants and blameless refugees.
He's just been in that area:
https://www.takimag.com/article/bullet-proof/
Thanks, I’d overlooked that.
> May 25, 2020 … in the wrong year: between 2020 and 2021
> when it should, of course, be between 2019 and 2020
I don’t understand. Is this some kind of humorous point?
Or I miss some point about how the graph represents change, so… The change caused by 2020 should be visible in 2021. But what is this mention of 2019? I’m confused.
Also, the second graph is confusing:
the legend is Covid: yellow, Floyd: pink,
but the graph has green and pink lines.
I guess Covid: also green?
> the legend is Covid: yellow, Floyd: pink, but the graph has green and pink lines.
It is a simple mistake: the legend should show Green (not Yellow) to match the Green line that marks the arrival of the Covid virus in the U.S.
Let me try to help with your two questions.
#1: The data reported as "2020" refers to homicides that occurred from Jan 1, 2020 through Dec 31, 2020. So the 2020 label on the graph could read "12/31/2020" to be more clear. Floyd died in May of that year, so the line for his death should be shown between the data points labeled 2019 (12/31/2019) and 2020 (12/31/2020), if we wish to draw conclusions as to whether his death had anything to do with deaths reported in 2020 and subsequent years.
#2: Yes, Covid also green. In the past I've provided Steve with Excel charts and the colors were garbled when he posted them. It seems to be a some sort of software incompatibility, perhaps between Excel and Substack. So you just have to roll with it when it occurs.
Ken
> The change caused by 2020 should be visible in 2021. But what is this mention of 2019? I’m confused.
The x-axis of the NYT's graph (2015, 2016, ... , 2023, 2024) seems to be about one thing (years, duh!). But it's actually about two.
For "Death of George Floyd, May 25, 2020," the point on the axis for "2020" represents Jan. 1, 2020, and the one for "2021" represents Jan. 1, 2021. So the dotted vertical line for George Floyd is placed to be at the 146th (of 366) day of 2020.
For "Homicides" and "Police killings," the point on the axis for "2020" represents "percentage change relative to Jan. 1 - Dec. 31, 2015, of deaths that took place between Jan. 1 and Dec. 31, 2020."
If the NYT had (more properly) depicted deaths as a bar graph, then the "2020" dot would be the upper left corner of the "2020" bar, and the "2021" dot would be "2020's" upper right corner.
Using six-month bars would have revealed the temporal correlation (Floyd's death preceded 2020's murder spike) that the Times' formatting conceals.
If only reporters Rich, Arango, and Bogel-Burroughs were aware of Sailer's work! Then they could have graphed "CDC figures for black homicide victimizations and black traffic fatalities by week from 2018 to 2023."
Their readers would have learned something true, and would have the tools to start thinking about *why* Floyd's death caused so many excess black homicides and traffic fatalities.
As a journalist of better character once quipped, "It isn't what we don't know that gives us trouble, it's what we know that ain't so."
FWIW, Steven Rich, clearly the data analytics member of the reporting team, and probably the senior member, offers the following bio, below. Turns out he is intimately familiar with the Washington Post data. Reporter #2, Tim Arango, is also an experienced hand, having reported international stories from the Near East for over a decade. So "rookie mistake" doesn't look like a good candidate for the misleading graphic.
Ken
==================
Steven Rich
I am a data reporter at The New York Times, using data analysis to investigate major issues and contextualize current events.
What I Cover
As a member of The Times’s data journalism team, I dig deeply into a wide range of subjects. I analyze data to identify trends, add quantitative context to stories, and help readers understand current events. This work facilitates both longer-term investigations and fast-paced stories on deadline.
My Background
Before joining The Times, I spent a dozen years as an investigative data reporter at The Washington Post, where I worked on stories across many topics. I developed particular expertise in data related to policing and opioids.
For 10 years, I helped to build and maintain a database of every fatal police shooting across the country in near-real time, an effort that was recognized with both a Pulitzer Prize and a Peabody Award. I mapped unsolved homicides in major cities to find neighborhoods where police departments had failed to solve the vast majority of murders, leaving victims and their families without full justice. My colleagues and I sued to get 14 years of data on shipments of prescription opioids across the United States. That data, which we made public, was used by hundreds of news outlets to examine opioid shipments to their communities.
I spent six years serving on the board of directors for Investigative Reporters and Editors, a nonprofit that helps journalists develop their skills. I am a graduate of Virginia Tech and the University of Missouri.
Journalistic Ethics
Like all Times journalists, I am committed to upholding the standards of integrity outlined in our Ethical Journalism Handbook.
I strive to use data to illuminate our world in a way that is accurate, transparent, and properly contextualized. I try to help readers understand the work behind my reporting, as well as its limitations. I maintain a healthy skepticism of everything and everyone, and I subscribe to the old journalism adage that says, “If your mother says she loves you, check it out.”
Thanks. Steven Rich certainly writes a catchy biosketch. In the context of this article, I am particularly taken with his three concluding sentences, which sum up his professional strengths and his aspirations:
"I strive to use data to illuminate our world in a way that is accurate, transparent, and properly contextualized. I try to help readers understand the work behind my reporting, as well as its limitations. I maintain a healthy skepticism of everything and everyone, and I subscribe to the old journalism adage that says, 'If your mother says she loves you, check it out.'”
Uh hunh.
If his article allowed reader comments (it doesn't), I would have quoted that paragraph.
Reading the companion "Five Years A Martyr" articles that do permit them, I note that a plurality of NYT readers do not (any longer?) appreciate being led around by the nose.
O/T
To answer your question on twitter, in 1968 Ray Oyler was such a bad-hitting shortstop that Mayo Smith moved Mickey Stanley there from CF for the Tigers in the World Series. While they won, it didn't work out for Stanley at SS so they traded for Tom Tresh from the Yankees in June