Humans are not good at detecting patterns…nor are we good at recognizing randomness when it presents itself as a pattern (or part of one).
The tragedy at Newtown brings up the question of whether mass shootings (counted as 10 fatalities or more) are increasing. If you do you use 10 fatalities as a baseline, then it certainly seems that we’ve seen an increase in these incidents over the last two decades. The increase in frequency as a percentage would appear to go down if you defined mass shootings as 2 fatalities or more (or even 3, 4, 5, and 6). So, it all depends on what constitutes a “mass shooting.” Defining these incidents is not merely an academic exercise. Indeed, the definition of “mass shooting” determines the abilities of policy makers to make informed decisions. And the definition of these events impacts the arguments of gun control advocates and champions of the 2nd Amendment. (I have not done formal research to verify these numbers, but based on a preliminary search it does seem there is an increase of incidents with 10 fatalities or more.)
Super-storm “Sandy” brings up another problem with pattern detection. Some commentators attribute the cause of the storm to climate change. It’s a convenient argument. There is a belief among climate experts that super storms will increase in frequency as climate change progresses. Nevertheless, there is zero empirical evidence linking this storm with climate change. We simply don’t know if “Sandy” is part of a large emerging pattern of super storms or if the storm was simply an occurrence unassociated with the variables of climate change.