I can tell you're an academic. That's actually a rather sensible comment.
The conflict in Syria has many shades of grey; but neither Assad nor ISIS are profoundly misunderstood 'good people' that are somehow justified by their moral ambiguity and rhetoric to kill Syrians because they feel their power and status are slipping away. A particular social hierarchy or dynamic does not justify itself by mere existence.
The effects of homophobic and racist actions cannot be lessened or ignored because their perpetrator is wavering in his conviction, and perhaps ten years from now will denounce and reject his position. So although it's quite possible to show empathy to some lost soul with open prejudices due to a number of compounding conditions, from upbringing to environment to lack of means, it does not absolve them in and of itself. Indeed, say I had battered you for ten years straight every time I met you, and then turned up on your doorstep with an apology, "Sorry, mate, I didn't really mean any harm. Just really been stressed these past few years. Hey, you're still alive, and I hope we can put it behind us." I somehow doubt you'd smile and offer me a warm embrace.
So when black people statistically kill more people, it's racist to say they're more violent?
Yes. It's quite a jump to make, and statistically illiterate to boot.
You're asserting that the statistics given measure something that they were not designed to do; namely, measure 'intrinsic violence capacity' of a population. In itself a rather quack effort, since you'd have to somehow link skin pigmentation to aggressiveness (the predominant differentiation that crops up in discussions of race), and define a biological concept of race suitable for analysis, sharing presumably some unique aggressive characteristics genetically which no other group has.
Using America as an example, I could argue by the same token that us white folk are intrinsically more prone to mass murder, but is this -- our whiteness -- really the cause of those crimes? Access to weapons, history, politics, compounding crimes, socio-economic factors, culture, geographic concentration and distribution of each population, amongst other things play a part. If you're not leaping to conclusions, you'd have to rigorously dismiss all other variables that offer context to crimes committed in the population to isolate your definition of race as the single most important factor or the sole factor. Such analysis is not offered in the images you've quoted.
Lastly, generalising from one sample country about 'race' and its effects in general is profoundly flawed, since it is quite possible to find less violent and more violent countries for any type of majority required, in this instance, and so on for other nonsense justified by the same means. Unless you're asserting that your 'races' aren't general, and that African-Americans are somehow different from other people of Afro-Caribbean descent; likewise for White Europeans. Which still does not rescue you from the statistical analysis required.
"There are more murders/higher murder rate in X community/country" as a statement of statistical fact is not in itself racist; the generalisation of "because Y people live in a more violent community/country they are intrinsically more violent than X in another community/country with fewer violent crimes" is. It's the racialist equivalent of saying that if one is born poor, one deserves it, and equally as laughable.
I could put a Gaussian line on top of another Gaussian line, doesn't mean one caused the other.
And your factor analysis is... where?