That doesn’t make them any less true? I think you’re confusing selection bias with the fact that something happened.
Say a news outlet says “X happened and it’s bad”. Another news outlet says “X happened and it’s good.” Both have their respective slant. Fine. But BOTH say X happened. Stuff that’s not newsworthy doesn’t go into news. You can’t just dismiss facts with prejudice as in “Don’t look up”.
The thing is that "warrant served without gunfire" doesn't get clicks, so it doesn't get reported. Looking at news sites tells you nothing at all about the relative prevalence of those two cases.
The relative prevalence is irrelevant. The claim was about shootings by felons on parole preferring to be killed by police rather than going back to jail. That claim was used to support the hypothesis that jails are horrible. The claim was not A is more common than B.
Reporting outlier anecdotes is more profitable than reporting systematic data analysis.
Outliers are true in the sense of factual, but false in the sense of accurate. The "Don't Look Up" metaphor applied here would be that the media is telling you to look, when reality is that you shouldn't.
Say a news outlet says “X happened and it’s bad”. Another news outlet says “X happened and it’s good.” Both have their respective slant. Fine. But BOTH say X happened. Stuff that’s not newsworthy doesn’t go into news. You can’t just dismiss facts with prejudice as in “Don’t look up”.