The Bern Report reports
Dr. Rick Hardy and Dr. John Hemingway have been leading Mock Presidential Elections [at Western Illinois University] 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.
On the Democratic side for the Primaries Sanders won[1] by close to a 2 to 1 margin over challenger Hillary Clinton.
[...] After nominating Sanders, the student Democrats put him up against Jeb Bush, who received the Republican nomination, with Bernie winning the electoral college by nearly a 4 to 1 margin and a decisive win for the popular vote.
This simulation always takes place the year before the presidential election year, and three months before the actual Iowa caucuses. The genesis of this mock presidential election began at the University of Iowa in 1975 with two political science doctoral students, John Hemingway and Rick Hardy. In that year, students selected Jimmy Carter over Gerald Ford--long before anyone really knew of Jimmy Carter.
In the years that followed, Rick Hardy expanded the format and engaged thousands of students at the University of Missouri-Columbia where students registered a perfect record of selecting the subsequent winning presidential party. In 2007 and 2011, Hardy and Hemingway teamed up again to conduct a massive campus-wide simulation at Western Illinois University. In 2007, Western students selected Barack Obama as president at a time when no one thought he could win! And, in 2011, students narrowly re-elected President Obama.
[1] Content is behind scripts.
(Score: 3, Insightful) by wonkey_monkey on Saturday November 14 2015, @05:30PM
University With Record of 100% Accuracy for Picking the Winning Party Says Sanders Will Win
So they've been doing this since 1975 - that's, what, 9 elections? So that's a 1/512th chance of doing it even by luck alone - hardly astronomical odds.
How many universities are there in the US? One of them probably has 0% accuracy for picking the winning party. They should just pick the same as these guys so at least one streak is going to end.
systemd is Roko's Basilisk
(Score: 0) by Anonymous Coward on Saturday November 14 2015, @06:42PM
There are probably a couple hundred universities in the US with both the size and political motivation to do a mock election with valid numbers. Makes sense that one and only one would have a perfect score. Odds are they will get this one wrong.
(Score: 2) by wonkey_monkey on Saturday November 14 2015, @09:43PM
Odds are they will get this one wrong.
Nope.
https://en.wikipedia.org/wiki/Gambler's_fallacy [wikipedia.org]
In fact I'd say the odds are they'll get it right. An election is not quite a coin toss - there are myriad factors, but that doesn't mean it's impossible to get some sense of which way it might go, even now. While their record so far is not completely inexplicable by chance, it does at the very least suggest that they might have a good enough grasp on politics to actually make a meaningful prediction - even if it only gives them, say, a 1% edge over guessing (which would give them a 1/428 probability of reaching their current record, I think).
systemd is Roko's Basilisk
(Score: 0) by Anonymous Coward on Saturday November 14 2015, @10:56PM
ISTM that this college's student body has simply been a reliable model of the USAian electorate in the past.
Fristy pretty much said this.
Unless the admissions practices have changed significantly there, I see no reason why the accuracy of the results should change.
a good enough grasp on politics to actually make a meaningful prediction
That's not at all what's happening here.
It's much simpler.
The kids are casting votes for the individual that they want to be president.
-- gewg_
(Score: 2) by wonkey_monkey on Sunday November 15 2015, @10:49PM
I see no reason why the accuracy of the results should change.
Nor do I. But nor do I see completely compelling reason to make any valuation of their accuracy much above 50/50.
systemd is Roko's Basilisk
(Score: 0) by Anonymous Coward on Sunday November 15 2015, @06:17AM
Their odds of success is very small because the odds are always very small.
(Score: 2) by wonkey_monkey on Sunday November 15 2015, @10:57PM
50/50 - which is realistically their minimum probability of success (they might have such a bad model that they are actually more likely to pick the losing candidate, but this could be turned into a winning model quite simply) isn't "very small."
It's also not "very big." It's right slap-bang in the middle.
systemd is Roko's Basilisk