I have always found that 1 million self defense uses per year claim to be fascinating. There were 5.2 million reported crimes last year in which the victim saw the perpetrator. Of those 3.5 million were attempted or threatened violence.
http://www.ojp.usdoj.gov/bjs/pub/pdf/cvus05.pdf
As near as I can find the largest % of an adult population with conceal carry permits in any states is about 7.65%. Most states have less than 5% of their adults with a permit. If we assume those 1 million self defense uses were part of that 3.5 million it would mean that your odds of being threatened with violence are way more if you have a conceal carry permit than if you don't.
Then of course maybe all these crimes that were stopped happened in a person's home. There again, we can look at the crime statistics.
about 25% of the violent crimes occurred at or near someone's home. But wait. The completion rate of the violent crimes was actually HIGHER when it occurred in the persons home. 28% of the completed violent crimes occurred at or near a person's home. So in other words, you were more likely to stop a crime away from home than you are in your home.
So, 730,000 crimes were stopped in people's homes. 2.449 million were stopped outside person's homes.
That leads us to some really odd stats. Either people that carry guns are a lot more likely to be attacked or people that don't have guns never prevent an attack in their home. Lets for a moment consider that all attempted crimes are committed and the perpetrator never knows who has a gun. 50% of homes have guns and 7% of people carry guns. That means 370,000 homeowners stopped crimes and 171,000 people carrying guns did. That would leave us with 540,000 crimes stopped with a gun. But that again is a hard one to buy since it would require that EVERY homeowner with a gun actually used that gun. Half of those 750,000 attacks prevented weren't at the home but near it. Many of those would have occurred as the homeowner was on his way home and he wouldn't yet have his gun.
Now the argument will be that many gun owners don't report their use to prevent a crime. That may be so but why would a gun owner's rate of not reporting be different from a non gun owner? Even if it was different then it would require that gun owners are more likely to be attacked.
No matter how you parse the stats it seems that if you don't have a gun you are less likely to be victim of an attempted robbery, rape or assault if the 1 million figure is true.