That said, this is what I meant by it always seems nailbiting at the time to predict that the frontrunners at this phase (remember, well over a year until the election), because surely this time we've got it all figured out? But... no, no we don't, and if anything I think it gets worse every cycle. The news cycle keeps speeding up, and other than Trump's apparently locating the teflon breastplate created by Reagan and then found again by Bill Clinton, every news cycle is an opportunity for a candidate to completely flame out.

Out of seven races under something approximating the modern primary system where an incumbent President or Vice President from the party isn't running, Democratic early (Summer-to-Fall of the year preceding the election) front-runners have apparently once actually won the nomination, have failed to win five times, and there's been one race with no clear early front runner (the one apparent win may not actually have been as early as Summer-Fall preceding, but I can't clearly rule it out, so we'll call it in.)

(Incidentally, I restrict this to Democrats because the way primaries and caucuses factor into nominations is different between the two major parties, such that one cannot validly assume that similar, from external qualities, positions in the two are similarly situated with regard to nomination.)

And the no-clear-frontrunner and frontrunner-wins elections were the first two opportunities -- everything since then (everything after 1984) has featured a clear early front-runner that lost.

2012 obviously wasn't in issue, because Obama was an incumbent President. 2008 Clinton was the early front-runner, Barack Obama the nominee 2004 Dean was the early front-runner (something lots of people making comparisons between Dean and Sanders forget), John Kerry the nominee 2000 wasn't an issue, because Gore was an incumbent VP. 1996 wasn't an issue, because Clinton was an incumbent President. 1992 IIRC, Tom Harkin was the early front-runner with Paul Tsongas second place in the early period, Bill Clinton ended up with the nomination. 1988 Gary Hart was the early frontrunner, Michael Dukakis the nominee 1984 I think Mondale was the early frontrunner (can't find any clear information easily that places that status back into Summer-Fall of 1983, though), and was the nominee. 1980 Carter was incumbent President 1976 -- the first nomination using something like the modern primary-dominated system rather than nominee selection dominated by party bosses -- no clear early front-runner, and Carter -- a relative unknown nationally before the primaries got started -- won the nomination.

> Focusing on the latter, fivethirtyeight did a comparison of Clinton's chances today compared to 8 years ago and found that she is in a much better position [1].

You omitted the detailed citation, but presumably you are referring to the piece "Hillary Clinton's Inevitable Problems" [0].

There's not really a coherent argument in that for why she is better positioned, just some scattered observations without any clear analytical framework (or even strong rationale for the other elections that are offered as comparables.)

> As much as I want Sanders to win, I trust fivethiryeight and I think they've got it right on this one.

The Fivethirtyeight brand was built on a richly deserved reputation of doing a good job of providing a useful framework for aggregating and interpreting general election polling, and I haven't seen many places do that better. Beyond that, I don't see much that Fivethirtyeight institutionally deserves "trust" on.

[0] http://fivethirtyeight.com/features/hillary-clinton-scandal-...