Re: Page 3 of this thread
http://www.theanswerbank.co.uk/News/Question1502054-3.html
and the Talbot/jno exchange, I'm not sure jno's point about
proportionally was acknowledged fully.
Even the Grauniad's statistics page didn't address this
http://4.bp.blogspot.com/-No2377MRFYU/VW9pszU2b0I/AAAAAAAAg5Q/hhWKOKg8DCw/s1600/guardian-the_counted-1018.jpg
choosing absolute numbers (which serves Talbot's stance) in preference to proprtional ones, which supports the #BLM side of things.
The Facebook post linked to was very well written and bang on the money: if it's Baltimore, with a BME majority population, it will "ALWAYS be more" whether it is black on black killings or killings by cops. Clustering of populations of minority groups is a fact which should be acknowledged. Nobody likes to move that far from their close family and/or social network. "The fruit doesn't fall far from the tree".
But if they're only 13% (say) of the population then, proportionately, they shouldn't be suffering nearly half as many deaths as whites. But, as per the Facebook post, tracking like-for-like arrest situations, suspect behaviours and outcomes is a complex piece of analysis, beyond a single individual's capacity to work out in their head. Computers might manage the task but who do you trust to do the programming?