You guys are just afraid you might lose, I bet.
Lose what? To lose, you'd have to argue a perspective not based in fact. As I said, the research is conflicted, so a factually grounded position in favour of
specific firearms-controls can be an elusive beast at best.
It's the internet. Feel free to reveal your half-formed opinions on gun control, criminology and racial science.
Now you're just baiting

Besides, I don't have half-formed opinions on any of the above, one of my degrees is in sociology/psychology, with a heavy emphasis on criminology coursework

I'm interested to hear some examples of "rates of success varying by country", MP-Ryan. The impression I got from skimming a few Reddit articles on this matter was that if there is an effect of gun control on crime either way it is almost negligible, and any evaluation of "effectiveness" is going to have to deal with far too many confounds for it's conclusions to be meaningful. You can't just yell "Success!" when a graph bounces one way or another after policy changes.
Well, such things are always bound up in the correlation != causation mess, but it really depends on the way you measure success. Violent crime is declining in Canada - has been since the '70s. Firearms-related crimes are also way down, especially since the 90s - when the Firearms Act was introduced. Firearms-related homicide, suicide, injury, and accidental deaths are also all down, also since the introduction of the Act. All of those are pretty fair measures of success, provided changes in firearms-related laws actually precipitated them. There's also the number of possible suicides and accidents prevented (by training and storage requirements), which isn't actually measurable (but given the decline in rates in general, can be analyzed as a trend). Success of firearms-control measures isn't just a function of crime rate. In fact, I tend to argue that the most compelling reason for some controls (including licensing and safety training) has more to do with the safety of the owner and their family (and the public at large) than anything to do with crime rate. By that measure, the controls introduced in Canada have been an overwhelming success.
Of course, by contrast you need only look at the handgun ban in Washington, D.C. to see abject failure of a control measure.
So it's really not that clear-cut. Evidence is mixed, success depends on the how, where, and whats of implementation.
And then there's the lesson from the drug war, that efforts to outlaw something is probably going to create some crime in itself.
Disingenuous argument. Prohibiton (outright bans) always fails due to supply/demand considerations. Regulation of a substance has more nuance. Banning alcohol created a new category of crime; regulation of alcohol alleviated some social problems [and crime] while permitting a small criminal sector to exist around it (as it did when alcohol was unregulated, albeit in a different form). Regulation of drugs may in fact be considerably more effective than either full prohibition or full legalization.
Firearms are best viewed from the analogy of vehicles (which was mentioned earlier). Unregulated operation of motor vehicles is probably not desirable from a public safety standpoint, but that doesn't mean they should have been banned. Instead, and large and flexible network of safety regulations exist: both on vehicles themselves, and the people who operate them. There is no compelling reason for firearms to be treated any differently, 2nd Amendment considerations momentarily put aside (since we're talking about principle, rather than specifics).
Perhaps the lesson here is to be careful declaring a win when the facts don't support it (and much as research is conflicted, the facts do not support completely unregulated ownership and use of firearms in developed nations).
EDIT: Also, I find myself oddly positioned as, in Canada, I'm actually firmly in the camp advocating for fewer restrictions on firearms than the general populace. Heh.