When some sort of "bad" phenomenon is experiencing a random peak, stories about this peak flood the media. When the same "bad" phenomenon has an extraordinarily quiet year, there are no stories in the media. This (mostly) innocuous media habit (based on their incentives) creates the impression among average folks that the "bad" phenomenon is on the rise, even when there is no such trend.
Case in point: tornadoes. How many stories have you seen this year about what may well be a record low year for US tornadoes?
Extreme lack of extreme with tornadoes. Will need "second season" to stop it from being quietest year on record! pic.twitter.com/PuBvgjJBvn
— Joe Bastardi (@BigJoeBastardi) November 8, 2016
Postscript: By the way, some may see the "inflation-adjusted" term in the heading of the chart and think that is a joke, but there is a real adjustment required. Today we have doppler radar and storm chasers and all sorts of other tornado detection tools that did not exist in, say, 1950. So tornado counts in 1950 are known to understate actual counts we would get today and thus can't be compared directly. Since we did not miss many of the larger tornadoes in 1950, we can adjust the smaller numbers based on the larger numbers. This is a well-known effect and an absolutely necessary adjustment, though Al Gore managed to completely fail to do so when he discussed tornadoes in An Inconvenient Truth. Which is why the movie got the Peace prize, not a science prize, from the crazy folks in Oslo.