Do Tornado Centroids Show a Tornado Alley Shift?

NOAA's Storm Prediction Center came out with a very interesting statistical tornado map recently:

GET MORE OF MY UPDATES! FOLLOW ME ON...

For the purpose of this blog, I have made one small change: I overlaid the "Tornado Alley" boundaries (in red), from the map found on The Online Tornado FAQ (also run by SPC). What this map now shows is that every year examined has (on average) fallen outside of what is classically defined as Tornado Alley (the closest years would have been 2007 and 2010, both slow years). Of course, there are other ways to define Tornado Alley ...but this dovetails with hearsay from storm chasers on social media that Tornado Alley isn't what it used to be.

In the article linked above, SPC's Harold Brooks refuses to define any tornado alleys, while AccuWeather's Mike Smith says there are two: "the classic stretch from Dallas to Des Moines, Iowa and Dixie Alley -- northeastern Arkansas, western Tennessee, northern Mississippi and Alabama."

It's an interesting question. I'll leave you with this video showing tornadoes by month from 1950-2011:

P.S.: The explanation of the SPC map, when shared on Facebook by the SPC, was: "Preliminary tornado reports centroid YTD, 2005-2014. The year is the third slowest start over past 10y with the center of activity about average." The caption on the map reads: "Marker size based on total preliminary reports. Centroid based on average of preliminary lat/lon in Local Storm Reports."