Ghosts Of Elections Past

One nice thing about blogging for so long is you've got archives, so you can look back and see what you thought, and how it holds up.  Since we're close to a presidential election, I figured I'd look back at how we covered the last few around this time of year.

I didn't want to get bogged down with lengthy arguments, so here are some posts that are relatively straightforward.

October 2012:

We'll have the final presidential debate tonight.  The Washington Post suggests it'll be pivotal, but I doubt it.  First, the candidates have created strong impressions from the first two debates that likely won't change short of some sort of meltdown.  Second, this debate is about foreign policy, which is simply not as pressing as domestic issues right now.

As the race comes down to these final days, I only hope the polls stay close.  For a while there it looked like it might be an easy victory for Obama, and it brought back memories of how boring it was four years ago when election night was over before it began.  I just want some excitement.

We'll likely have a pretty good idea who'll win when the eastern states come in--Florida, Virginia, North Carolina and Ohio (not to mention New Hampshire, Pennsylvania and Michigan). But at least before then both sides can hope. And who knows, it could even be a late-night race that comes down to Colorado and Nevada.

October 2008:

There are some who claim the polls showing Obama clearly ahead are wrong. It's true different polls use different methods, so some of them must be a little off, but they can't all be wrong. I'd say every poll being way off is about as likely as McCain winning right now.

PS One of the more positive polls for McCain is this one, but I wouldn't call it reliable. Listen to this: "[McCain is] also gaining momentum in the suburbs, where he's gone from dead even a week ago to a 20-point lead." Do you believe that? Would anyone?

October 2004:

The latest Harris Poll gives us two choices for potential likely voters, with greatly varying leads (2% and 8%) for Bush. Apparently, they're not sure if they should measure likely voters in a new or traditional mode.

This won't do. Any pollster (heck, any person) can give you a bunch of different results based on different methods saying one of them is probably correct. The Harris Poll is a venerable institution, but if they can't tell us what they actually think is the best data, then who needs 'em?