Limnophile...I posted a 5 year data set taken of individuals using their weapons in self defense situations. You felt this data inaccurate and typed in some statements about LA police shootings in response to crime scenes and felt this was a more accurate representation?
Andy,
That is correct. The data in your link was gathered haphazardly at best; thus, no valid statistics can be drawn from them. That's the problem with "street data."
I'm unaware of any valid database of civilian defensive gun uses. I believe LE DGU data offer a fairly reasonable model for the armed civilian, and LE departments typically have policies requiring all gun discharges to be reported (non-NDs at the range excluded). The LAPD even requires the drawing of a weapon not fired to be reported.
Thus, the numbers in the LAPD's 2010 Use of Force Report do not constitute a sample (valid or otherwise), but rather a complete census of DGUs for that year.
The 2010 LAPD data are in general agreement with the Rule of 3s with regard to average number of shots fired in that the average per officer per incident was 2.5 as I recall. The average per incident, because in most incidents more that one LEO fired, was 4.5 shots as I recall. But, at least one incident involved 23 shots being fired.
By the way, the average distance involved in the 2010 LAPD shootings does not conform to the Rule of 3s, as it is in excess of 30 ft. While most of the shootings by far were with sidearms, some were with shotguns and rifles, which likely helped drive the distance up. The way the data are tabulated all I know is the longest shot of the year was greater than 60 ft.
The important point I wish to make is that planning for the average DGU of about 3 shots is risky. If we plan for the average DGU we wouldn't bother carrying a sidearm, because on the average day, year, or lifetime one has no need to use a gun for self defense. Of course, if the need ever does arise, no gun means trouble.
If one does opt to carry a handgun, the typical DGU involves drawing one's weapon but not firing. Thus, one might argue an empty gun is good enough on average. Of course, if the need to fire the gun arises, no ammo means trouble.
If one opts to carry one's gun loaded, the Rule of 3s says that about three rounds of capacity should suffice on average. Knowing this one might feel content carrying a compact 5-shot revolver with no reloads, but if you find yourself having to defend yourself against multiple armed assailants you may find yourself wishing you had carried a few reloads.
Capacity was a major criterion for me in choosing my carry pistols, but I've come to understand there is a role for the pocket pistol when deep concealment or convenience are more important. A pocket pistol will have a limited capacity, so it would be prudent to always carry at least one reload and preferably two.
And sometimes I've got a spare mag for the pistol on me 
My theory is go the biggest that is comfortably concealable. We're all still sucking oxygen so we can't be doing that bad whatever the method