Monday, October 6, 2008

The Electoral College Map (10/6/08)

Another slow weekend poll release, but at least Sunday's new polls were in states that are on or have recently been on our Watch List (states closest to switching categories or sides according to our average). Put very simply, Minnesota is all over the place. Within the last week, CNN has it at +12 Obama, Survey USA at +1 McCain and now the Minneapolis Star Tribune has it at Obama +18. So which is it and can we believe any of those polls anyway?

Throughout the last month or two Minnesota has been the topic of conversation in the comments sections of these electoral college posts here at FHQ. That is largely attributable to the erratic nature of the polling in the North Star state. The Survey USA poll is the first to show a McCain lead since March, but polling had shown a range from around 0 to the mid- to upper teens across the entire year's polling in the state before that. Now, why is it so hard to poll Minnesota? Well, part of it has to do with election day registration. If there isn't a filter question (or series of questions) in the survey that asks the likelihood of someone both registering and voting on election day, then those folks don't count as registered and they obviously don't count as likely voters. Some of the electorate is potentially being missed then. But which poll is closer to right? The truth, as our custom around here may suggest, is somewhere in the middle. Minnesota looks to have tightened some, but we are beginning to get some information that indicates the North Star state is following the national polls in that it is trending toward Obama. But by 12 to 18 points? Probably not, but it does indicate that the state is fairly strong for Obama at the moment.

New Polls (Oct. 5)
Star Tribune
Columbus Dispatch

Outside of Minnesota, there were also new polls out in Colorado and Ohio. Ohio is very much tracking along the same lines as the national polls. But Colorado, after having a few polls in the wake of the Lehman collapse favor Obama at levels outside the margin of error, has reverted to a margin that, while it still leans toward Obama, is certainly tighter than in some of the other toss up states. And that process continues with the Mason-Dixon poll showing the race in a dead heat.
[Click Map to Enlarge]

But back to Ohio for a moment. The Columbus Dispatch poll, like other recent polls out of the Buckeye state, indicates a mid-single digit lead for Obama. [It should be noted that this poll is a mail in poll which comes with some potential issues, but that figure is in line with some of the other polling that has emerged from Ohio in the last week or two.] Another pretty good Obama result and Ohio still doesn't turn blue? No, but it is really close now. [Hey, weren't you supposed to be making some changes to address this lack of responsiveness?*] But close doesn't count, does it? For now then, Ohio stays in the McCain toss up area and the electoral college vote distribution remains unchanged from yesterday, 278-260 in favor of Obama.

The Electoral College Spectrum*
*Follow the link for a detailed explanation on how to read the Electoral College Spectrum.
**The numbers in the parentheses refer to the number of electoral votes a candidate would have if he won all the states ranked prior to that state. If, for example, McCain won all the states up to and including Pennsylvania (all Obama's toss up states, but Michigan), he would have 299 electoral votes. Both candidates numbers are only totaled through their rival's toss up states. In those cases, Obama's number is on the left and McCain's is on the right in italics.

Colorado is the state where Obama crosses (or McCain would cross) the 270 electoral vote threshold to win the presidential election. That line is referred to as the victory line. It is currently favoring Obama, thus the blue text in that cell.

The new Minnesota poll also has the effect of moving it back on to the Watch List, within a point of moving into the strong Obama category. That polls also vaults the North Star state above New Jersey on the Electoral College Spectrum. In both cases, Minnesota is the only change. Colorado is still at the center of the struggle. In its current position as the victory line, the Centennial state puts Obama over 270 electoral votes and would put McCain over if the Arizona senator was able to hold on to the states in shades of red and pick up Nevada. Even though that tie in the Mason-Dixon poll of Colorado is shows a tie race, if McCain were to win it and not Nevada, that would get the race to a 269-269 tie in the electoral college. And that tiebreaker doesn't look too good if you put any stock in any of the House election projections.

The Watch List*
Iowafrom Obama lean
to Strong Obama
Michiganfrom Toss Up Obama
to Obama lean
Minnesotafrom Obama lean
to Strong Obama
Missourifrom Toss Up McCainto McCain lean
Nevadafrom Toss Up Obama
to Toss Up McCain
North Carolinafrom McCain lean
to Toss Up McCain
Ohiofrom Toss Up McCain
to Toss Up Obama
Oregonfrom Obama lean
to Strong Obama
Pennsylvaniafrom Toss Up Obama
to Obama lean
Texasfrom Strong McCainto McCain lean
Virginiafrom Toss Up McCain
to Toss Up Obama
Washingtonfrom Strong Obama
to Obama lean
*Weighted Average within a fraction of a point of changing categories.

On the Watch List, this is still very much a Nevada, Ohio and Virginia discussion. Those states are the ones closest to switching sides of the partisan line. Of the rest, most states on the list are flirting with moving into or out of the toss up category. And for the most part, most of that potential movement is toward Obama.

*Yes, and I think I've settled on a slightly different methodology that should help there. As I said, it is more of a progressive weighting structure and it better captures polling changes while rooting them in past results. But I'll get into that more in an FAQ-type post later on...after I've got it implemented. Speaking of which, the implementation of the new formula is somewhat tedious. I hope [HOPE] to have it up and ready to go tonight. If I do, I'll update the map and other graphics as if there was no change and post the altered methodology version along side of it for comparison's sake. Again, hopefully that will be tonight, but we'll see.

Recent Posts:
The Electoral College Map (10/5/08)

The Electoral College Map (10/4/08)

The Electoral College Map (10/3/08)


Unknown said...

So will the Virginia polls, move Virginia closer to blue?:

Virginia SurveyUSA Obama 53, McCain 43 Obama +10

Virginia Suffolk Obama 51, McCain 39 Obama +12

Anonymous said...

Yeah, I saw those. They are certainly good polls for Obama. It goes without saying that they will move Virginia closer to the blue. The question is, how much? For that you'll have come back later tonight for the full update.

[I'm such a jerk.]

As an update, I applied the new formula to all of the toss up states and that produced some interesting results. Hopefully I can plow through the remaining states and get the full alteration up as well.

Robert said...

We are waiting on baited breath!

Jack said...

Looking forward to a blue Virginia, and to the new formula that will hopefully produce a bluer map.

The Columbus Dispatch polls make the Zogby Interactive polls look reliable. Since you're not including Zogby Interactive, any thoughts about dropping Columbus Dispatch, too? Yes, I know they produce less noise than Zogby Interactive due to the low number of polls from them, but noise is still noise. I mean, if I decided one day to personally poll, say, Iowa, and sent you the results, you wouldn't include that on the grounds that it wouldn't throw off the average too much, right?

Also, why do your zeroes look funny? What typeface do you use?

Anonymous said...

How do the zeros look? And where do they look strange? If it is the tie in the table for Colorado, I can change that to "tied" if you it would help. I can change the font in the straight Blogger formatting, but the charts are a different tale. But I can mess around with some point.

I knew that someone was going to question the Dispatch polls at some point. They too probably need to be dropped from consideration since they suffer from some of the same problems the Zogby polls do. We aren't lacking for data in Ohio.

And I'm waiting on that Iowa poll, Jack.

Jack said...

The zeroes look ... short and squat. And it's both in the table and in the text. But there's absolutely no need to mess around with the formatting on my account. You're entirely too accomodating.

I'll see what I can do with the Iowa poll. Due to the logistical difficulties, as well as the fact that I'll have to do some research as I completely lack knowledge on how to do this kind of thing, I won't be able start immediately. My goal is to be able to begin polling by November 5 at the earliest.

Robert said...


Starting then will greatly improve your accuracy unless Iowa is next in the Florida--Ohio--??? progression!

Jack said...

And then FiveThirtyEight will include me at the top of their Pollster Ratings!

Robert said...


I think you may need more than just one poll!

Jack said...

Well, on November 5, I could do 51 (don't forget DC!). Surely, that would be enough!