Brad's election prediction

Kerry wins by a larger margin than anticipated

With the election tomorrow it's time for predictions. Based on using The Political Machine's issue based predictions and looking at polling data and trends, this is what I'm concluding:

Kerry victory with 306 electoral votes to Bush's 232.  This is a much clearer victory for Kerry than any of the polls and pundits are showing here. But then again, they don't have The Political Machine.

So what do I base this on?

  1. The Iraq war isn't a positive for Bush but in fact a negative.

  2. I think the Democrats are better at mobilizing their base. This has to do with demographics. It's easier to get people to the polls in urban areas than rural/subburban areas. As I explained to a neighbor, there's roughly 200 people in our neighborhood. It would take hours to pick us all up and take us to a precinct to vote. By contrast, there might be 500 voters in a single building in Cleveland.

  3. I think fraudulent voting is going to be a real issue. It's a lot tougher to challenge fraudulent voting than it is to try to preserve the sanctity of the system. It's disappointing to think that my vote might be counter-balanced by someone's cat or "Dick Tracy" but our sensational media has clearly decided that it's better to have 1000 fraudulent voters get through than to prevent a single legitimate voter from being turned away. I disagree. I think there should be a balance because a fraudulent voter is effectively taking away the voting rights of someone else by canceling out their vote.

  4. The "early voting" mechanism is going to be the big story this year. Early voting favors less committed voters which tends to favor Democrats.

  5. That ridiculous "missing explosives" story from last week that looks to have not had any reasonable basis in fact (latest news on that is that a US unit actually removed the explosives at the end of April). But the media ran on that for a full week which may have been enough to tilt the balance in Florida, New Hampshire, Ohio, Wisconsin and Iowa to go for Kerry.

But in the end, if Kerry wins, it will be that first Bush debate that caused Bush to lose. The momentum was in Bush's favor and he so blew that first debate that it shifted momentum that he has not been able to recover from.

So there's my prediction. Not nearly as close as the pundits are predicting.

That said - EVERY informed voter should go out and vote. So even if you're in a blue or red state, go and vote tomorrow. I think the good news out of this election would be that we may have record turn-out which I think would be fantastic regardless of who wins. 

You Republicans who are hoping Bush wins better do what you can to get your friends and neighbors and family out to vote. Because you can be assured that the NEA (teachers) and other unions and other organizations are way ahead of Bush and his allies in organizing get out to vote on election day drives.

This election is going to come down to a handful of votes. Particularly in Florida, Ohio, Wisconsin, and Minnesota.

21,290 views 64 replies
Reply #1 Top
I hate to agree with you Brad, but I believe this is a very accurate scenario.
Reply #2 Top
mmm I hope you are 100% correct in this prediction, Brad...the Political Machine model is actually quite similar to what electoral-vote.com is reporting today -- 298 Kerry vs. 231 Bush with NM and RI tied.
Reply #3 Top
Osama bin Laden lives. I say Churchhill and FDR were right. Fight this murderer until he can hide no more. Never give in to tyranny. Fight him with the most powerful weapon you have, your vote. Vote Red State for Freedom.
Reply #4 Top

Interesting article. If the political machine is this good, I may have to download it after all (and give up playing GalCiv for a while - that may prove to be hard).

Obviously, I hope that your prediction is correct, but more importantly, if it is, then I hope that the two parts of America will settle their differences. It has been a quite scary last fourtnight, with threats and allegations from both sides.

On a side note, I think that Bush's pledge to remove Castro from power may just swing Florida in favor af the Republicans.

Morten
Reply #5 Top
Well, if the Political Machine hits in on the head that accurately, who knows... you might be able to license the engine to the media as a predictive tool for other elections too
Reply #6 Top
This will be an excellent litmus test for the political machine's accuracy. It would be a boon for stardock if your analysis was dead on. Good luck!
Reply #7 Top
Interesting, Political Machine is a great game, but a great tool for election predictions? We'll see tomorrow!
Reply #8 Top
Interesting article Brad! The fun begins tomorrow morning, lawsuits to follow for the next 6 mos
Reply #9 Top
This election is going to come down to a handful of votes. Particularly in Florida, Ohio, Wisconsin, and Minnesota.

I agree with this statement Brad, however I feel Florida and Ohio will go to Bush. The polls are just too close. Also, the game is so much fun! But unfortunately, it cant predict other entagables... such as weather, work, traffic, polling lines etc. I think it will be much closer than you have it here...
Reply #10 Top
NM and RI tied.


Being a RI'er, I nearly died when I read that and had to go check it out--NM and NH are predicted as ties. RI is as blue as it can go!

I'm interested in seeing how close the Political Machine is, Brad. When it was announced in the papers a few weeks (months? I don't remember) ago, my roommate, who works on Capitol Hill, emailed me the story with a comment like: Do you believe this? Some computer guy thinks he can predict the election. I was trying to cover up my blogging addiction (which is still a secret to most) while also letting her know how cool the game is (and avoiding all questions about why I would have any knowledge of it). She remained doubtful, so I'd like to be able to gloat --very mature of me, I know.
Reply #11 Top
It's been my experience that a lot of people vote with their hearts and not with their heads; some with neither. That presents a conundrum for any game engine or for that matter any AI developement. Your average Trekkie knows that! Best of luck with the game, however I don't think it can be relied upon as an accurate barometer. This election has stirred quite a firestorm of emotions. Those come from the heart.
Reply #12 Top

It's going to be really close in a bunch of different states. That's the thing, a few points more in war in Iraq and the situation changes dramatically.

Ohio, Florida, Iowa, New Mexico and Wisconsin are all up in the air. Just moving a slider a few milimeters one way changed the results in some of these states.

Reply #13 Top
It'll be interesting to see how close your model ends up being to the actual results. I'm with Historyishere on this one, you should license it to the media as a visualization/prediction tool. It would certainly rock to see news anchors playing a video game to explain results or to predict outcomes.

Reply #14 Top

On a side note, I think that Bush's pledge to remove Castro from power may just swing Florida in favor af the Republicans.

I did not hear that one.  But unless he has a lot of banana peels, I kind of doubt that!  I am hoping Brad is wrong, but if Bush made that statement, that is desperation.

Got any Banana peels? !

Reply #15 Top
I'm with Historyishere on this one, you should license it to the media as a visualization/prediction tool. It would certainly rock to see news anchors playing a video game to explain results or to predict outcomes.


Yeah, they already use videogames to predict the outcomes of sporting events!
Reply #16 Top
Tough choice, here, Brad -

Good for game, bad for Bush. Bad for game, good for Bush. Hmm...

This having been a year of remarkable surprises in many spheres, perhaps TPM is missing the boat. If I remember correctly, Howard went to the polls either in a dead heat or with a very narrow poll lead in Australia & came out with a rather comfortable win. We can always hope.

Cheers,
Daiwa
Reply #17 Top
Red Sox won the World Series, but Redskins lost, I would wager that all predictability has been lost this year!!

Unpredictable Plinko!!
Reply #18 Top
Looks like you're not the only one to predict that: Link
Reply #19 Top
Interesting map... Kerry takes Wisconsin and Bush takes New Mexico -- both kind of surprising calls.

This also assumes that Kerry takes both Ohio and Iowa, which appear to be dead heats, plus litigation-prone Florida. Seems to me that each of these three is maybe a 55% shot for Kerry, so if I were Mr. Kerry, I wouldn't start the celebration quite yet.

But in the end, if Kerry wins, it will be that first Bush debate that caused Bush to lose.

I agree 100%

I think fraudulent voting is going to be a real issue.
I have not seen much from reliable sources on this. Honestly, I had not perceived this as a serious problem. Any links from reasonably objective sources? (Not to argue, just to read up.)

Reply #20 Top
Hey, Brad, just out of curiosity, how close did Arkansas wind up being? There's a pretty strong Kerry base here and our boy Bill came around yesterday to stump for Kerry (which actually might hurt him in some parts of the state -- plus, our only statewide paper hates Clinton). I'm thinking Bush will probably win Arkansas with about 53% of the vote, and I just wondered what your model had to say on the matter.
Reply #21 Top
I can only hope and pray that there is a teensy weeny little flaw in your AI Brad.
Reply #22 Top
Draginol:

We agree for almost the same reasons. However, one thing that you didn't mention which I think has played a much bigger role in this than anyone ever imagined is the growth of Air America. Now at approximately 40 stations it has given the liberal base a voice to listen to and Al Franken hip-hopping all over the country at either rallies or broadcasts has lifted the democrats from being in a real funk in the spring to being ready, willing and able to report for duty on election day.

I know that you are not on the same page as Franken but you have to give the guy some real credit for highlighting the issues in a way that made Bush the ultimate flip-flopper and Kerry presidential.
Reply #23 Top
I don't think Air America has had any impact.
Reply #24 Top
Interesting prediction from TPM. I hope that whoever wins does so by a large margin so that we don't have all the recounting that delayed the election results 4 years ago.
Reply #25 Top
I don't think Air America has done just a whole lot. Maybe someday that will be different, but its market is way too limited to have done much.