Spree killings are incredibly rare. As such, policies designed to cut down the death toll of a spree killing that does occur can easily have their positive side effects outweighed by negative ones.
For example, if you had tens of thousands of armed people in our movie theaters, day after day, year after year, to guard against the once-per-lifetime event of a movie theater spree shooting, what would the result be? Maybe, that one time, you'd have five or ten fewer people killed during the event. But maybe, over those many years of having many guns in many other theaters, you'd have dozens of additional deaths (accidental shootings, minor events that became deadly because someone had a gun on hand, etc.) So, you could have a huge net increase in killings.
That's why it's seldom a good idea to let your policies be driven by exceedingly rare events. By far, the vast majority of people who are killed by guns are killed one at a time -- accidental shootings, suicides, and single homicides. Those are the scenarios that should guide our gun policies.
