Sunday, October 12, 2008

Forecasting the Electoral College with October Data

As promised in my last post on the September Model I'm be posting a rolling Electoral College forecast (in the panel to the right), based on past election results and October national and state-level trial-heat polls. As with the September model, these state-level forecasts were developed with Jay DeSart. Clicking on to map to the right will take you to Jay's web page, where more details are available. As new polls come in the forecast will change, so please check in on on a regular basis for changes.

The results for today are:
  • Electoral vote: Obama 354, McCain 184
  • National popular vote: Obama 52.85%, 47.15%
The biggest change from the September model is that both Virginia and West Virgina are now in the Obama column.


Josh Putnam said...


I was late to comment on the Missouri/Virginia situation in the previous post, but was curious about North Carolina and to a lesser extent some of those other red states.

North Carolina would likely drift onto the Obama side of the win percentage table if the past elections variable were dropped. It wouldn't be insulated by those four consecutive GOP wins there since 1992. But how would some of those other red states be affected if you dropped that variable or changed it to the last two elections instead of the four? Most of them, I'd suspect, would stay where they are simply because Obama has seemingly reached his ceiling in recently polling (These are reliably red states.), but is there any consequential change at the margins?


S said...

The real issue is not how well Obama or McCain might do state-by-state, but that we shouldn't have battleground states and spectator states in the first place. Every vote in every state should be politically relevant in a presidential election. And, every vote should be equal. We should have a national popular vote for President in which the White House goes to the candidate who gets the most popular votes in all 50 states.

The National Popular Vote bill would guarantee the Presidency to the candidate who receives the most popular votes in all 50 states (and DC). The bill would take effect only when enacted, in identical form, by states possessing a majority of the electoral vote -- that is, enough electoral votes to elect a President (270 of 538). When the bill comes into effect, all the electoral votes from those states would be awarded to the presidential candidate who receives the most popular votes in all 50 states (and DC).

Because of state-by-state enacted rules for winner-take-all awarding of their electoral votes, recent candidates with limited funds have concentrated their attention on a handful of closely divided "battleground" states. In 2004 two-thirds of the visits and money were focused in just six states; 88% on 9 states, and 99% of the money went to just 16 states. Two-thirds of the states and people have been merely spectators to the presidential election.

Another shortcoming of the current system is that a candidate can win the Presidency without winning the most popular votes nationwide.

The National Popular Vote bill has passed 21 state legislative chambers, including one house in Arkansas, Colorado, Maine, North Carolina, and Washington, and both houses in California, Hawaii, Illinois, New Jersey, Maryland, Massachusetts, Rhode Island, and Vermont. The bill has been enacted by Hawaii, Illinois, New Jersey, and Maryland. These four states possess 50 electoral votes-- 19% of the 270 necessary to bring the law into effect.



Jay DeSart said...

It's funny, Tom and I had a little debate about the prior vote variable prior to his posting our forecast here, for precisely this reason it's being discussed now. [BTW, I think I won ;) ]

Our original model, the one we initially developed and have used the past couple of elections, used a prior two elections variable, instead of the prior four. In the paper that we presented at the WPSA in March, we showed that using the prior four variable made the model generally more stable and accurate. So that's why we decided to go with it for this election.

Interestingly, if we'd gone with the prior two elections variable in the model for our September forecast we would have generated a prediction that McCain would win both in the popular vote, 50.79 to 49.21, and in the electoral college, 278 to 260. Why? Because the original model gives much more substantial weight to Bush's performance in states, and not really capture the "true leaning" of the state over the long haul.

Simply dropping the prior vote variable altogether signficantly weakens the model by all measures (Adjusted R-Square, SE y/x, Mean Absolute Error in state predictions, % of States correctly predicted, and average error in the national popular vote and electoral college vote)

But to answer Josh's question more directly, here are the specific differences in the state predictions if we went with a two-prior election variable in model for the September forecast... McCain would be projected to win (some of them by comfortable margins) most of the "tossup" states: Colorado, Florida, Missouri, Nevada, New Hampshire, North Carolina and Virginia. Obama would be predicted to win Michigan, Minnesota, New Mexico, Pennsylvania and Wisconsin. These are states that are currently viewed as states that are firmly in the Obama column, but the prior 2 model would consider them toss-ups that he just barely wins.

In light of the current landscape, it looks as though our decision to go with the prior 4 elections variable might have been the right call. We'll just have to wait and see.

Josh Putnam said...

Thanks Jay.

I sat in on that WPSA panel in San Diego. That may be why I had the prior two elections version of that variable on my mind (at least subconsciously). I definitely agree that the prior four is the better option if only for the balancing effect between two Democratic elections and two Republican elections.

Incidentally, I'm hoping to post the results of the forecasting end of the Gurian and Cann model from that very same WPSA panel on my blog sometime soon. They are working out some kinks at the moment.

Thanks again, Jay. I appreciate the prompt response. I'm looking forward to the updates that you and Tom will have.

Matty Boy said...

Hi, professors. I'm also keeping track over at my blog, Lotsa 'Splainin' 2 Do, using a method I call confidence of victory. If no polls make major moves between now and early tomorrow morning, by current polls I have Obama leading 350 to 188 in expected value, while the numbers as of most recent input from is Obama 352, McCain 166 and 20 too close to call, as the last three polls in Ohio are one for Obama, one for McCain and one even split.

You can take a look at my predictions for the past few weeks at

Anonymous said...

Who knows where to download XRumer 5.0 Palladium?
Help, please. All recommend this program to effectively advertise on the Internet, this is the best program!

Term Paper said...

Thanks for pulling this together and sharing!

Term Papers said...

This is such a great news, it really helps, Your blog is nice and informative. Thanks for the article.

College Research Paper said...

I appreciate the work of all people who share information with others.

Education Agent said...

Thats an interesting post. It was worth visiting your blog and I have bookmarked your blog. Hope to visit again.

Boundless Technologies said...

Hi, I have just visited your site and the info you have covered has been of great interest to me.

Web Hosting

black leather jackets said...

Great blog here, such a wealth of content.
discount leather jackets

celebs clothing said...

Most of the times i visit a blog I see that the construction is poor and the writing bad. On the contrary,I have to say that you have done a good job here.

Leather Jackets said...

Over all blog is fantastic and informative, I like this blog.....

thanks to share,

cctv karachi said...

Pretty good post. I just stumbled upon your blog and wanted to say that I have really enjoyed reading your blog posts. Any way I'll be subscribing to your feed and I hope you post again soon.

dean winchester leather jacket said...

Really very nice post, Thanks for sharing this valuable information with us.
casino royale leather jacket

casino royale leather jacket

Casino Royale Jacket

Term Papers said...

Thank you for the post, really interesting and amazing. Keep it up.

Custom thesis writing services said...

That is an amazing way of calculation and it is quite easy though. Thank you for the post.

david said...

Nice ..

angela seo said...

This blog is very amazing thanks admin sharing this post.
drive scorpion jacket

angela seo said...

Finally I found an blog that has all that i want....
dark knight rises costume

angela smith said...

Thanks for sharing this gossip, I shared your blog news other social platforms, your blog is very interesting & excellent hard work.
prototype game alex mercer jacket

Alexia416 said...

woow, interesting to see this kind of stuff. I ralley appreciate ur stuff, its very helpful for others
mens slim fit jackets

Adrianna102 said...

I found lots of interesting information here. The post was professionally written and I feel like the author has extensive knowledge in this subject.
ferris bueller jacket

Sidra shahid said...

It’s great to see good information being shared and also to see fresh, creative
ideas that have never been done before.

Cctv Karachi
Happy Times

Kierra James said...

this is really informative post about Electoral College forecast i think your all post is good i will read all of your post.

Kierra James said...

you should have to work hard for your blog it seem look more better than ever the post you write is interesting to read and i got much information after reading it.

Stinson Leather said...

Great Post finally Got What i need
Captain America Jacket

Celebs Leather Jackets said...

you can check our leather jackets collection at bane leather vest