Why would suicides using guns be considered "puffing" the numbers in favor of the anti stance?
They used a gun to kill themselves, right?
Do you want statistics, or just statistics that support your theory.
There are "statistics"...and then there are "statistics".
Statistics:
the practice or science of collecting and analyzing numerical data in large quantities, especially for the purpose of inferring proportions in a whole from those in a representative sample.
A raw number with no context and no informative data on the components that make up the statistics is meaningless. In fact, the context and informative data on the components of the statistical sample is crucial to accurate/meaningful interpretation of the data.
A "homicide" is, legally, simply the killing of a human being by another human being. That's what the components of the word means..."homo" referring to "man" (from "homunculus") and "cide" meaning "the act of killing" (from "-cidium"). It's from old French, derived from Latin.
Since we're talking about the legalities of homicide, it's the LEGAL definition which is important. Not a layman's interpretation of the word.
So the start of our problems with statistical analysis begins with the collection of the data itself and incorrectly/inaccurately lumping killings which are NOT "homicide" into the sample data.
Moving from there, we progress into the intent behind any given homicide. Not all homocides are, in fact, "illegal".
"Justifiable homicide" is a killing of another human being which is legal under the jurisdictional precepts of the law.
"That which is committed with the intention to kill, or to do a grievous bodily injury, under circumstances which the law holds sufficient to exculpate the person who commits it." If you kill someone in the act of self-defense, and the law finds that you are justified in doing so, then you have not committed murder, or even manslaughter. How, then, does blindly lumping every justifiable homicide into the general statistics skew the interpretation of the overall data?
"Involuntary manslaughter" is
the unlawful killing of another human being without intent. Thought there are shades of involuntary manslaughter which progress from criminally negligent to unlawful-act, the key here lies with intent. Lack of intent means involuntary manslaughter is not murder. Again, if you blindly lump every instance of involuntary manslaughter into "homicide" without an interpolative understanding of each circumstance, how does that skew the interpretation of the overall data?
"Accidental death" is a death strictly due to accident, typically excluding such things as acts of war, death caused by illegal activities. Again...blindly lumping accidental deaths into the general category of "homicide" without knowing/understanding the circumstances skews the overall conclusions.
We could go on, and certainly we could make a case where each side will interpret the data to their own ends.
HOWEVER, we all have to be honest with how the "facts" are presented and deal with the underlying reasons behind all those facts. How truly relevant is the data for the conclusions being drawn? In addition to an indepth analysis of the actual data (meaning and context), a common test of relevancy is whether or not changing anything based on the conclusions drawn actually has a significant effect on the statistical sample being analyzed. Another would be parallel studies, "double-blind" studies, etc.
"Homocide" by itself, as a singular measuring stick, is too general and too wide-ranging to be an effective measure by which to judge the overall impact of privately owned firearms.