What are these poll analysts smoking?? Stats wizards come debunk this...

[ALago1]

I don't know how these people come to the conclusions they come to, but they have a probability analysis in which B/C have an 80% probability of winning while K/E have a 17.9% probabiliby.

They claim to be independent but they seem like crackpots. Check it out:

http://dataseers.com/041027/tep.htm

[John Q. Citizen]

1. I hope their methodology is a poor

as their web design abilities.

[skylarmae]

2. divining the essence? ... WTF ...n/t

[NJCher]

3. bookies have it the other way around

Cher

[cubsfan forever]

4. Who cares what a bunch of RW jerks say?

There are three types of lies: Lies, Damned Lies, and Statistics.
These are just a bunch of wingut number-crunchers trying to appear
intelligent, to whit: "seers." Forget this nonsense. GOTV

Professor 2

[Awsi Dooger]

5. There are mathematical models all over the internet

I have looked at literally dozens of them, in 2000 and 2004. Borrowed/stolen some of the ideas. Unfortunately, some of the best ones from 2000 have not been updated this time.

I generally dismiss the models that rely too heavily on state polls. This one is an example of that.

[salib]

6. This is BOGUS

THere is NO transparency here. No names, no background. Even if you look up the ip name, there is no info whatsoever. They are a stealth operation and that means lies.

[ALago1]

7. Hey, you know what, you're absolutely right

Any self identification is very vague and general.

"We are a small group of dedicated computer professionals that has worked together for years."

No names, no affiliation, nothing at all.

It probably is a stealth op. A sucky one at that...

[TruthIsAll]

8. I have taken a quick look at it. Very strange.

Seems like they are trying to rip off The Election Model, at least conceptually. Their wording is suspiciously similar to mine. They appear to be shilling for consulting work in Microsoft Office.

Finally, they give themselves away when they talk about running millions of simulations, where 5000 is sufficient. Election trial envy. My simulation is bigger than yours.