Do college students accurately predict #FeelTheBern? A #skeptic view

This story has been making the rounds of the Internet, about how students at Western Illinois University predict that not only will Bernie Sanders win the Iowa caucuses, he will then go on to beat Hillary Clinton for the Democratic nomination. And then, he'll win the general election, with detailed Electoral College predictions per the map.

(And then he'll become Tsar of All the Russias, or Pope of All the Catholics, or the Lubavitcher Rebbe, or something.)

How realistic is this?

Well, here's the nutgraf:

Dr. Rick Hardy and Dr. John Hemingway have been leading Mock Presidential Elections since 1975. During that time, students who have participated in these mock elections have chosen the winning party with 100% accuracy and have an astonishing record in selecting presidential winners. 

Interesting. 

First, there's a time for every streak to be broken. Ask the 18-0 New England Patriots of a few years back, for example.

Second, even if Clinton's support continues to weaken, and looks thin, even if wide, I simply cannot believe he'll win by a 3-1 margin.

Period and end of story.

Third, it predicts Jeb! will win the GOP nod. Erm, uh sure!

Fourth, a Sanders-lovers website was reporting incomplete primary returns from the mock elections. Here's the actual, for analysis.

Some of these are just laughable.

ZERO Clinton delegates in Colorado? ZERO for Sanders in Kansas or Kentucky, conservative as they may be within their Democratic as well as GOP parties? But a 31-8 split for Sanders in even more conservative Oklahoma? ZERO for Clinton in inside-the-Beltway, Democratic side, Maryland? As well as ZERO for O'Malley? Yes, Martin O'Malley may be out of the race by late April. Or, he may not. ZERO Sanders delegates out of the People's Republic of Madison, and the rest of Wisconsin?

Oh, and Republicans winning Hawaii, per a Facebook friend? Or, Democrats South Carolina?

So, no, I'll pass on these predictions. More below the fold.

(Note: I have emailed one of the two professors involved, asking for further explanation. I'll post any I get.)

And, here we go

Prof. Hardy has now weighed in, admitting the mock election isn’t scientifically controlled:

This was meant to be an educational experience for the students. This year we had a high percentage of students from Chicago area, many from depressed areas of the state. In the past, we had random selection of students but this year we did it by class. That had a significant effect. This was just a simulation and the student learned a lot and had fun. Best, Rick 

Sorry, professor, but, if that’s the case, then shouldn’t you not hype that you’ve been accurate every time since 1975?

I don't blame Berniebros for running with the story as much as I blame Hardy and Hemingway for not noting they'd changed student selection methodology, and any MSM who uncritically ran the piece without asking the questions I did above.

Maybe the two profs were under orders by administrators to "sell" their mock election, too?