A few things:
First, I just want to say how very sad this chart is. Regardless of whether it represents what it seems to on the surface, or something deeper (as I get to later), it should not be this way.
Second, I don't think you did this, but others have: It's dangerous, IMO, to say that because a trend exists, it applies to individual cases. When you say, "I think it was the determining factor in this case," that's fine, as you've qualified it as a personal belief, rather than some undisputable truth. When someone (not you) claims the chart indisputably shows that it applied in this case, that's not correct. It may have been the determining factor, but it may not have been. After all, there do exist cases where whites are not acquitted, and others where blacks are. So it's important to remember to distinguish between trends and their potential application to individual cases (which I think you've correctly done).
Finally, (and I have to load this full of huge disclaimers considering the emotions running through this thread: I am NOT suggesting that discrimination against blacks does not exist; I believe it does, to a significant extent) a potential factor in the chart that could account for some of the discrepancy is gang feuds or other crimes that are heavily over represented statistically by black people, and would be far less protected by self-defense laws than isolated incidents. Of course this doesn't mean that white people are better than black people, obviously. It's important to acknowledge that violent crime happens far more often among black folks due to the economic disadvantage they've been put at. So it's more likely (statistically speaking!) for a black killer to be involved in a scenario that makes self-defense difficult (like a gang fight, or a drug feud) than for a white killer.
So my point is, the chart could be skewed by crime statistics generally. Again, I am not suggesting discrimination does not happen; it does. But the statistics of crime should at least be a consideration when we view charts like this. Again, I'm talking pure correlated statistics that have nothing to do with the merits of each race, and are only correlations based on the economic disadvantages at which blacks have been placed. I hope I put in enough disclaimers that my point is clear.