Why do we care about "other factors"?
California's large cities are not any more safe than Louisiana's large cities and if there is a difference, it's because of demographics, not gun control.
Homicides are certainly going to be higher in cities with disproportionately high numbers of blacks, Hispanics and poor people...
Seems like you know exactly why. If there are factors other than gun control that affect homicide rates, then they have to be accounted for before you can use homicide rates to prove or disprove anything about gun control.
In fact, when I started using EXACTLY the same tactic you did to compare cities, YOU immediately brought up "other factors". Now, if you want to go back and edit your posts to eliminate those objections, I think maybe you could legitimately go back to arguing that "other factors" don't matter. But as it stands, it's abjectly ridiculous for you to continue to argue against considering other factors when you, yourself, just pointed out that other factors are important to consider.
Furthermore, I'm not interested in proving anything. I'm interested in disproving something; specifically, the hypothesis that strict gun control leads to decreased homicides.
Because you acknowledge that there are other things affecting homicide rates and because it's easy to see from the data that if there is an effect from gun control, it is much smaller than those other effects, then until you can compensate for the effects from the other factors and isolate the effect (if any) due to gun control, you don't know if gun control is having an effect or not.
All you know is that the effect (if it exists) is smaller than the other factors by a significant amount.
By the way, if you would just slow down and read that last sentence--the one that is highlighted--and think about it for a minute or two instead of just immediately starting to compose a response, you might just have an epiphany.
It either does or it doesn't and the data can't really be weighted differently because of these "other factors".
Massive oversimplification. It may have a positive effect in some areas, none in others, a negative in others. For one thing, not all the states that are primarily anti-gun have the same mix of gun control. Second, as you acknowledge there are other factors affecting homicide rates which change from one to another.
If you want to prove that there's NO effect, (as opposed to just proving that the effect, if it exists, is very small-hint, hint) then you have to go through all the possible effects and compensate for them so you can see what (if any) effect is actually due to differences in gun control laws.
Go back to the water tower example. Without turning the pump off, you'll never see the effect due to the hole. The hole IS there, it IS having an effect, but you won't ever be able to see it in the water level data unless you can compensate that water level data for the effect of the pump to isolate the effect due to the hole. I'm not claiming that gun control is having an effect on homicide stats in the U.S., this is just an example of how one effect can be present but masked by other more significant effects.
...any attempt to throw out the statewide per 100,000 rate is nothing but a statistical trick to lend support to a disproven hypothesis.
At times, I think you are very close to finally "getting it" and then you will say something like this...
If you could just read through some of what's been posted here with an open mind, you would see that you can already effectively rebut the argument that gun control can be shown to have an effect based on a simple assessment of raw homicide statistics. Unfortunately, because you're stuck in "Reject! Reject! Reject! Mode", all you can do is repeat the same logically bankrupt arguments over and over, and continue to fight against statistics which if you were able to stop arguing just long enough to think about them, actually prove almost exactly what you want to prove.