Web site predicts electoral votes

[Nicholas_J]

11/1/2004 1:54 PM
By: Associated Press

A lot of people who don't want to wait until election night to see which are the red states and which are the blue states have been visiting Andrew Tanenbaum's Web site.

The site, Electoral-vote.com, features a tally of which way the electoral college vote will go for each state if the most recent polls are correct.

States on the site's map appear in shades of blue, red, pink and white -- depending on whether they're strongly or weakly leaning toward President Bush or John Kerry, or tied.

Tanenbaum had kept his identity secret until Monday, revealing on the Web site that he's an American computer science professor living in Amsterdam.

http://rdu.news14.com/content/headlines/?ArID=58023&SecID=2


Since Mr Tannenbaums site is currently under seige, becasue he has predicted a Kerry winI thought I would publish this here on DU.

[cheezus]

1. time for the micro vs monolithic kernel debate

[lapfog_1]

2. No pure examples of either exist, anymore?

Right? Linux and Winblows have KLMs, as do all current unices...
which only leave old mainframe OSes like VM and MVS, both of which
have some micro kernel features...

Oh well. Back to 24 x 2 election coverage.

[endnote]

3. Bullshit. This site uses no MOE info, therefore it's mostly crap.

[Nicholas_J]

4. This site far more accurate than sites that Use an MoE

As it sets criteria for determining how states are awarded by a factor that is higher than the MoE of the majority of polls Itdoes not give a state to a candidate for sure unless a candidate is polling more than 5 point above the other candidate. This is more accurate than giving a state to a candidate who is ahead in a poll that has an MoE of 3 percent, and the candidate has a 3.5 percent lead. Secondly, you cannot get an MoE for an average of polls. Averagingthe MoE's for ten polls does not give you the actual MoE of those polls as a group. Using a fixed figure that is higher than the MoE of all polls is a more certain way of determinnig if the state is actually going for the candidate. The argument that he does not use an MoE thus is an invalid argument. His method of determining a states leaning is a far more certain method.