craziness of 2004 & 2008 primaries: should we all just give up on predictions?

[Mr. Morden]

OK, now that it looks like McCain is a virtual lock for the GOP nomination, let's revisit just how crazy the last two presidential nomination cycles have been.  Particularly the 2004 Democratic primaries and the 2008 GOP primaries.  In both cases, you had a candidate (Kerry in the first case and McCain in the second) who was given up for a dead months before the primaries started somehow stage an amazing comeback.  Yes, there were people who never completely gave up on them, but there were stretches of several months in which each candidate was considered an extreme longshot by virtually everyone, and no one would have rated them the favorite.

Which raises the question: At least in these races in which there's either no frontrunner (GOP in 2008) or a fairly weak frontrunner (Dean with the Dems in 2008), does it even make any sense to be making any predictions months before the primaries start?  Do we really know anything useful about what's going to happen until maybe a couple of weeks before Iowa, when we can look to see if anyone's surging in the polls in either Iowa or New Hampshire?

I mean, we spend all this time analyzing polls from months in advance and making predictions, but in each of these two cases, our best guesses were pretty much all wrong.  Any predictions in either October 2003 or October 2007 about what was going to happen in Iowa or New Hampshire or who was going to win the nomination were completely wrong.

Or is there some way to analyze this stuff that would allow one to assess the race more than just a few weeks pre-Iowa in a meaningful way?

[Meeker]

You're absolutely right. There's really hardly anyway to tell more than a couple weeks before the process starts.

That, of course, won't deter anyone from doing it all over again in four years

[Harry Hayfield]

Whatever for? So far this election season I am doing very well indeed thank you for asking!

Republican Tally: 22 state projections correct
Democratic Tally: 16 state projections correct

[Silent Hunter]

Am I right in thinking that it's impossible for Obama or Clinton to now get enough delegates to lock up the nomination pre-convention?

[Erc]

Am I right in thinking that it's impossible for Obama or Clinton to now get enough delegates to lock up the nomination pre-convention?

Barring one of the two of them dying or eating babies, it is impossible for either of them to win on pledged delegates alone.  (Obama would need to win 78% of the remaining pledged delegates and Clinton would need to win 84% of the remaining delegates).

That said, it doesn't mean this is going to be decided at the convention.  It's quite possible (and is rather likely, given the pressure not to have a convention fight) that one or the other will have a large enough lead when you count superdelegates who have endorsed a candidate that the outcome will be clear wll before the convention.

[Sam Spade]

Primaries are very unpredictable - that's not really news.

Hell, we have enough trouble predicting these races 3 days out, much less three months out.

[ZamboniDriver]

OK, now that it looks like McCain is a virtual lock for the GOP nomination, let's revisit just how crazy the last two presidential nomination cycles have been.  Particularly the 2004 Democratic primaries and the 2008 GOP primaries.  In both cases, you had a candidate (Kerry in the first case and McCain in the second) who was given up for a dead months before the primaries started somehow stage an amazing comeback.  Yes, there were people who never completely gave up on them, but there were stretches of several months in which each candidate was considered an extreme longshot by virtually everyone, and no one would have rated them the favorite.

Which raises the question: At least in these races in which there's either no frontrunner (GOP in 2008) or a fairly weak frontrunner (Dean with the Dems in 2008), does it even make any sense to be making any predictions months before the primaries start?  Do we really know anything useful about what's going to happen until maybe a couple of weeks before Iowa, when we can look to see if anyone's surging in the polls in either Iowa or New Hampshire?

Does it make sense?  No.  But, like a sporting event, people will make them anyway because it's fun to do.  A viewer must make an individual determination about how much confidence he/she wants to place in those predictions.