Two New Senate Polls

Unfortunatly for Mason Dixon the Demographics are way the hell off (4.00 / 1)

What is your current official party voter registration? Are you registered to vote as a:

DEMOCRAT 40%

REPUBLICAN 32%

INDEPENDENT 27%

NJ has about 2 million nonaffiliated Voters, 1.3 million DEMs & 850k repubs. For a total of about 4.15 million voters. Turnout for the "R" & "D" might be as high as 50%, Indies turnout, not as much, 40%. But I won't get into that right now. Of the 4.15 million voters in NJ about 30% are DEMS & about 20% are Repubs. In rough numbers thats a ratio of 3/2. Mason Dixon assumes a ratio of 4/3.

DEMOCRAT 40% x 1.12=44.8

REPUBLICAN 32% x 1.31=41.92

The simplfied math above show that the multiplier used is not the same. All of the polls I have seen on the NJ senate race indicate that Indie voters are not breaking to either candidate by large numbers, 2% one way or the other.

Now I will apply the Mason Dixon Multipliers to the real voter rgistration numbers.

"D" 30% x 1.12=33.6-
"R" 20% x 1.31=26.2

Thats about a 7.4pt lead for Bob. Over the last 4 weeks the breakdowns I have seen, favor Bob by maybe 2-3% at most. That would just pad Bobs lead..... lets make it 8 or 9%. Which would be in line with other recent polls like the Gannett/Monmouth poll which had Bob up by 9%.

Now, I am not a math wiz by any means, I will try an averaged multiplier of 1.21 (which I am not sure is right)

"D" 30% x 1.21=36.3
"R" 20% x 1.21=24.2

Here, Bob is up by 12%.

How to identify a bogus poll:

NJ voter reg as noted above:

Dems/R/Ind
30/20/50

Mason Dixon says they used:
Dems/R/Ind
40/32/27

Generous Est. turnout
"D" 50%=501,500
"R" 50%=425,000
Indie 40%=800,000

If all affiliated things are equal, DEMs have a 76k vote advantage. Indie voters would have to be breaking to Kean by nearly 60/40 for the race to be even. I'm guessing 57/43 to Kean if Bob has a 3pt lead. ANd no-where have we seen any poll during the campaign that had Kean up by more than 2-3pts with Indie Voters.

Check out my 3 paragraph primer on Polywell Fusion.

---

by: FogerRox @ Wed Oct 25, 2006 at 18:20:57 PM UTC