2004 User Predictions - Discussion

[mooster]

I'm new to this web-site.  I'm impressed by Vorlon's almost casual mastery of the electoral map.

Why isn't Penn amongst the toss-up states?

What does voter registration data tell us?

I have a statistics background, and I'm curious about the "model" that you refer to.  Is this a formal mathematical predictive model, and if so, what are the input variables?

[mooster]

Thanks , Vorlon, for the info on the voter registrations.

So about this predictive model...

Siege40 - if Canada had won the war of 1812, we'd now have universal healthcare, one quarter of the present level of gun violence, and we'd currently hold the hockey gold medal, in other words, peace, order, and good government.  Oh well, we'll have to be happy with life, liberty, and the pursuit of happiness.

However, your predictions are, to use the technical term, perquacky.

 

[mooster]

Vorlon,

On the prediction pages, you are listed as I-indep, but on the discussion pages, you are listed as I-california.  Also, I notice that your email address is @shaw.ca, which, when googled, turns up a canadian service provider.

Also, what happens when you plug your own demographics into your model (age, occupation/employer, level of education, etc.)?  Surely it does't return a high probability for a libertarian vote?

[mooster]

Right now, the national polling numbers are 45% to 45%.  Obviously, it all comes down to a few battleground states, and how the vote divides there.  Amongst LIKELY voters (the only kind that count), its 48-44 in Michigan, 45-41 in Ohio, 45-44 in Penn, 48-45 in Florida, 47-44 in Minn, and 51-41 in Iowa,  ALL IN FAVOUR OF KERRY!  (I know, its a bit shocking).  If the election were held today, Kerry would actually win the electoral vote.  Now, naturally, the margins are within the polling error in many states, but the trend can't be denied.