>>3662892>>3662913Anon you're not very good with statistics.
Take your your 25x figure for instance.
Blacks are roughly 10% of the population. If their targets were following a random, uniform distribution on the whole population, out of 10 black attacks, you'd expect 9 to be on whites.
If the whites did the same, out of 10 white attacks, you'd expect 1 to be on blacks.
I.e. with no increased criminality and exactly same targeting and violent behaviour, you'd expect 9x more attacks to be by blacks on whites.
So your 25x figure is cute by a factor of 9, it now becomes 2.7x.
Next you should look into conditional expectations.
Find the (statistical) correlation between poverty and crime. Find the distribution of blacks with regards to poverty, then the same for whites.
And then calculate the conditional expectation of E[black attack on white| black is poor] and E[black attack on white | black is not poor]. Then do the same for whites.
So in the end even *if* your data are right, and you have
E[black attack on white]= 2.7x E[white attack on black], when you break down to conditional expectations and use Bayes' formula, this 2.7 factor can be explained by the fact that the expectation of a black person being poor is 2.7 (if not more) times the expectation of a white person being poor: E[black is poor]=2.7x E[white is poor].
And if you compare the *same* thing, i.e. E[black attack on white | black is not poor] vs E[white attack on black | white is not poor], as well as E[black attack on white | black is poor] vs E[white attack on black | white is poor], you'll see they follow population averages very closely.
What you're doing with the ""infographic"" is hiding the fact that the distribution with regards to poverty is very different for blacks and whites, and also the fact that there's strong statistical correlation between poverty and crime.
>TL;DR: your /pol/ infographic a shit and you can't into statistics 
