[marktd]

In 2000, both Gore and Bush were always the frontrunners, and then they both won the primaries in landslides. I believe they both won every state, except Bush lost New Hampshire to McCain. Also in 2008, you're exaggerating the extent to which the punditry favored Giuliani; he was a frontrunner, but not at all considered a shoo-in the way Hillary is.

The problem with Bernie is that he has very limited appeal outside of the areas where he's campaigning (overwhelmingly white, also a lot of college towns). He has a chance if he can improve his standing among minority voters somehow, but even then it's a long shot.

[eevilspock]

Cornel West, a highly respected person in the black community, just endorsed Sanders[1]. What's interesting, and not surprising if you know West, is that he's support of Sanders is much stronger than his support of Obama 8 years ago.

I think it will be far easier for black voters to get behind Sanders than behind Clinton. I'll suggest the same is true for Hispanic voters.

Since blacks aren't going to get behind any of the GOP candidates, the question is more about whether they will sit out this election.

[1] http://www.nytimes.com/politics/first-draft/2015/08/25/berni...

[jerf]

"the punditry favored Giuliani; he was a frontrunner, but not at all considered a shoo-in the way Hillary is."

That is all I was claiming. Hillary is an unusual case; I've not seen this level of "shoo-in" before.

Also, just to be clear, I'm not making grandiose claims that I know something about the race, nor am I claiming that this is somehow identical to previous situations [1]... it's more that I know that something isn't true: The punditariat is wildly overconfident, and their current confidence in their beliefs is wrong. Even if they are ultimately correct, they're still overconfident right now. The echo chamber does that.

[1]: It never really is, and there's a certain amount of anti-inductiveness to similar situations in politics: http://lesswrong.com/lw/yv/markets_are_antiinductive/ If you see a similarity that favors X, in general, so do X's opponents, and they're likely to do something that will, as a side effect, break the similarity somehow.

[eli_gottlieb]

>In 2000, both Gore and Bush were always the frontrunners, and then they both won the primaries in landslides. I believe they both won every state, except Bush lost New Hampshire to McCain.

I think the question is: does the ability of pundits to predict elections correlate or anticorrelate with voter participation rates?

Or in simpler terms: do pundits predict well when primaries are democratic and reflect the will of the people, or when they're isolated, low-information games that mostly reflect punditry itself?

[slg]

It is also worth noting that in 2008 the Clinton campaign was a lot weaker than it currently is and Obama's campaign was a lot stronger than Sanders' current campaign. [1]

[1] - http://fivethirtyeight.com/datalab/bernie-sanders-youre-no-b...