Don't forget that statistics do not deal with a particular event, but all events. You could toss "heads" seventeen times in a row, but in the overall scheme of things it's still a 50-50 chance you'll toss tails.

The same thing is true of "grading on the curve." The "bell-shaped curve" of probabilies is based upon a huge population, and to apply it to a population of the 20 people in a classroom is wrong. The classs could all be "A" quality just as likely as they could all be "F" quality.

Statistics deals with populations, not discrete events. Thus this year could be the coldest in the last 450,000 and still not be any indication one way or the other about "global warming."

More pragmatically, go ask the Arctic peoples what they think.