[pburgh01]

Here is something to chew on Gold. I was polled today by the Romney campaign..in person, by a real person, outside my local Giant Eagle grocery store. Pretty young, Penn State student holding a clip board. She did not disclose who she worked for until I pressed her. This is how campaigns poll, in addition to phone polls. Rasmussen, Zogby..none of them do this, the take a small sample and use mathematical rules to extrapolate. A SWAG as we say in software speak..Scientific Wild Ass* Guess. So yes watch the polls but watch the intangibles and the climate and what the enemy is doing. I smell a rout coming in 3 days, big rout. We are taking the Senate and throw in another 5-10 seats in the House. Harry R will have a time just maintaining his caucus and doing any filibustering.

[goldstategop]

The math is just bad. If Ras shows the race tied again today, I’m writing it off. A D+4, D+5, D+6 turnout model simply isn’t believable.

The conventional wisdom is its 2008 all over again. Its simply wrong. That’s what happens when the media and the pollsters follow the herd.

If your assumption is wrong, then the data isn’t a good fit for it. GIGO.

[newnhdad]

back in the 90’s, I used to work with a guy that was a student of theoretical mathematics and he was taking internships at companies that allowed him to just “think” about mathematical models and applications of advanced statistical methods. He was a total dork, but thinking on such a high level that we had no idea what he was talking about. He would volunteer to go out and get us lunch if we would sit in the conference room and listen to his ideas.

In referencing both 1996 and 1998 election models, he cited Michael Barone.