Odd pattern in NH. I'm just reporting the numbers here.

[creativelcro]

Sorry, I posted this in another thread that got buried. I'd like to know what you think (serious discussion only, please).



I've crunched a few numbers, just to see... I created a vector of towns with a 1 for towns that have optical scan machines and a 0 for standard paper ballots. I created another vector with a 1 if Kerry's numbers were higher than Dean's numbers (not even taking into account the magnitude of win) and 0 otherwise.

There are 237 data points. The correlation is about 0.36, highly significant (p<.0001) http://faculty.vassar.edu/lowry/ch4apx.html . This means, that towns with optical scan machines had a much higher chance to go Kerry than Dean. It may be a coincidence (e.g., there may be other factors that correlate with voting method distribution).
Perhaps worth looking into...

What do you think ?

[pnziii]

1. Sounds interesting

Worth another look especially with all the black box voting debate going on.

[displacedtexan]

If there hadn't been detailed exit polling...

I would say you might be onto something. Human exit polling is what convinced us that Gore had been robbed in Florida.

[stickdog]

65. The exit polling was full of shit. (nt)

[EstimatedProphet]

2. Any way to examine Iowa too?

That will bring in more data, and, since it was a different state, act as a seperate "treatment"...

[seaglass]

6. You know there were no optical scan voting machines in

the Iowa caucuses, right?

[EstimatedProphet]

17. D'OH!!!

Forgive me. Not much sleep last night...

[EstimatedProphet]

22. I see now where I got confused...

Iowa used DRE electronic voting machines, not optical scan machines.
One question though: are they any more reliable?

http://www.whatreallyhappened.com/04votefraud.html

[seaglass]

24. Ok, now I'm confused.

I watched the caucuses on TV and I thought they were only taking voice votes (and counting bodies where groups were larger). Then I saw them call in the votes. I didn't see any use of machines except calculators.

[lyonn]

28. you got that right

No machines only raised hands and telephones, oh, and a calculator.

[drfemoe]

45. Democrats spent $750,000 for new computerized

technology for the IA caucas according to a screen roll I read that night. Very little additional information at this time, but parts of it were *new* and *computerized*. States are implementing the new regs from legislation the "Help America Vote Act".

I'm not into statistics, so this is not my perch. Just adding what I know to the discussion.

[elfwitch]

3. I think it is VERY interesting...

I posted in another thread something similar. I find it odd that Dean would lose that much of a lead in such a short time to finished so far behind. He lost something like 20-30 percentage points in two weeks.

Skull and Bones. That's all I'm saying.

[bryant69]

4. It is interesting

But personally I'd need to know a bit more before making up my mind.

This story is more believable if you think that Dean is the best candidate and that President Bush is afraid of him. But many would have the opinion that it's more the opposite, Kerry is the scarier candidate to the Bushs.

Still, It's hard to know exactly what the President and his men are thinking.

Bryant
Check it out --> http://politicalcomment.blogspot.com

[pnziii]

7. Wait a second

Most of the people saying Kerry is scarier are Bush supporters.
Why would they tell us to nominate someone that Bush would have a hard time beating?
Then they go overboard telling us to nominate Dean because Bush would kick his ass. Why would they show their hand like that?
If they really wanted to win and they really thought Dean was weak why didn't they keep their mouth shut and let Dean get the nomination?

[bryant69]

10. Well, for one thing

I"m not necesarily buying what Bush supporters are saying. I think Kerry is a bit more electable, but that's based on my own analysis of the two men, not Bush Supporters.

I also don't know what you mean by most people, either because I suspect most Kerry Supporters (and apparently there are plenty of them) probably aren't also Bush Supporters. And I suspect if you asked a Kerry Supporter he'd say that Kerry was a better candidate than Dean.

But I could be wrong.

[stickdog]

67. What makes Kerry more electable other that the fact that the media

hasn't attacked his electabilty yet?

[lyonn]

29. who's the scariest

You really need to watch the republican analyst. Kerry was way behind just a few months ago. Amazing, now he is the "one". I've noticed even the demo. analyst don't speak well of Dean. I keep wondering what is wrong with me, being a good demo., that I can't see the big problem with Dean. Again, the media holds all the cards.

[stinkeefresh]

8. I used to wonder about that...

until I saw the differnce in media treatment between Dean and Kerry. The difference is so intense, I've come to the conclusion that the media must really want to kill the Dean candidacy. They have never been so nice to a NE liberal as they have to JK. Why?

Personaly, I think the truth is that Rove would rather see a predictable, Washington-style cross between Gore and Dukakis (isn't he though?) than an unpredictable firebrand.

[revolve]

5. Thats interesting

Or perhaps the poorer towns use the paper ballots while the richer towns use the optical scan and lower income people like Dean and richer people like Kerry. You can make stats lie any way you want and there arnt conspiracies everywhere.

[creativelcro]

9. I agree with your point.

There may be other factors that correlate with voting system type (such as income) and with political tendencies. Unfortunately, this is all correlational stuff (so, impossible to know what causes what)...

For what it's worth, Dean lost by 12.88% (on average) in optical scan towns and by 0.7% in paper ballot towns.
(%diff. was calculated as 100*(Dean votes-Kerry votes)/total votes for town)

Iowa did not have any machines, right ?

[nothingshocksmeanymore]

13. But isn't that becase opti scan is more accurate?

[stickdog]

69. Thank you, James Baker. (nt)

[HereSince1628]

31. Well taken, and that could be controlled for using partial correlation

put the scores for potential confounders in first and the psuedo-variable last...

[stickdog]

68. Perhaps. Or perhaps the election was far closer. (nt)

[theboss]

11. Where were the machines?

The easisest way to view this race is that Dean won the part of the state that borders Vermot and Kerry the part that borders Massachusetts. Unfortunately for Dean, a lot more people live on the border with Massachusetts. (I also believe those are the places more likely to have optical scan).

[seaglass]

12. You're not actually reporting the numbers, you're

reporting your analysis of whatever numbers you're using.

[creativelcro]

14. Numbers are here

http://www.boston.com/news/special/politics/2004_results/
http://www.electionline.org/interactiveMap_result.jsp?state=NH&stateText=New%20Hampshire&topicText=Voting%20System%20Used&topic_string=22:votingsystemtypemain

Screen-dump is here. last number is 1 if optical (nothing for paper).


City or town TP PR Clark Dean Edwards Kerry Kucinich Lieberman Sharpton Others Optical(1)
Belknap
Alton 1.00 1.00 103.00 185.00 87.00 213.00 4.00 68.00 1.00 4.00 1.00
Barnstead 1.00 1.00 87.00 162.00 109.00 244.00 10.00 50.00 0.00 2.00 1.00
Belmont 1.00 1.00 106.00 168.00 119.00 348.00 8.00 79.00 0.00 8.00 1.00
Center Harbor 1.00 1.00 26.00 53.00 28.00 62.00 0.00 15.00 0.00 2.00
Gilford 1.00 1.00 170.00 299.00 137.00 453.00 17.00 115.00 0.00 8.00 1.00
Gilmanton 1.00 1.00 51.00 174.00 68.00 221.00 5.00 41.00 1.00 4.00 1.00
Laconia 6.00 6.00 289.00 520.00 298.00 916.00 21.00 265.00 4.00 6.00 1.00
Meredith 1.00 1.00 148.00 243.00 148.00 415.00 5.00 96.00 2.00 2.00 1.00
New Hampton 1.00 1.00 52.00 116.00 41.00 130.00 3.00 26.00 0.00 3.00 1.00
Sanbornton 1.00 1.00 49.00 156.00 56.00 224.00 26.00 45.00 1.00 2.00 1.00
Tilton 1.00 1.00 93.00 160.00 97.00 230.00 12.00 46.00 1.00 8.00 1.00
County Totals 16.00 16.00 1174.00 2236.00 1188.00 3456.00 111.00 846.00 10.00 49.00

Carroll
Albany 1.00 1.00 16.00 41.00 13.00 50.00 8.00 4.00 0.00 5.00
Bartlett 1.00 1.00 75.00 186.00 69.00 240.00 9.00 30.00 1.00 3.00
Brookfield 1.00 1.00 11.00 53.00 13.00 34.00 2.00 6.00 1.00 4.00
Chatham 1.00 1.00 3.00 14.00 6.00 19.00 5.00 0.00 0.00 0.00
Conway 1.00 1.00 250.00 445.00 132.00 558.00 28.00 59.00 1.00 7.00 1.00
Eaton 1.00 1.00 12.00 26.00 13.00 39.00 5.00 5.00 0.00 1.00
Effingham 1.00 1.00 26.00 70.00 17.00 89.00 2.00 9.00 1.00 0.00
Freedom 1.00 1.00 34.00 80.00 26.00 109.00 3.00 23.00 0.00 1.00
Harts Location 1.00 1.00 6.00 3.00 2.00 5.00 0.00 0.00 0.00 0.00
Jackson 1.00 1.00 53.00 143.00 35.00 110.00 5.00 16.00 0.00 2.00
Madison 1.00 1.00 55.00 121.00 38.00 169.00 2.00 22.00 1.00 4.00
Moltnboro 1.00 1.00 108.00 170.00 73.00 263.00 4.00 84.00 0.00 2.00
Ossipee 1.00 1.00 56.00 131.00 64.00 180.00 7.00 35.00 1.00 4.00
Sandwich 1.00 1.00 49.00 180.00 31.00 100.00 9.00 15.00 0.00 2.00
Tamworth 1.00 1.00 94.00 110.00 39.00 212.00 20.00 15.00 1.00 4.00
Tuftonboro 1.00 1.00 43.00 93.00 50.00 138.00 4.00 44.00 0.00 2.00
Wakefield 1.00 1.00 82.00 145.00 85.00 212.00 8.00 46.00 0.00 3.00 1.00
Wolfeboro 1.00 1.00 123.00 380.00 109.00 401.00 23.00 73.00 1.00 6.00 1.00
County Totals 18.00 18.00 1096.00 2391.00 815.00 2928.00 144.00 486.00 8.00 50.00

Cheshire
Alstead 1.00 1.00 64.00 188.00 42.00 92.00 23.00 21.00 2.00 3.00
Chesterfield 1.00 1.00 118.00 307.00 94.00 217.00 14.00 26.00 0.00 0.00
Dublin 1.00 1.00 56.00 132.00 45.00 149.00 18.00 22.00 0.00 1.00
Fitzwilliam 1.00 1.00 63.00 129.00 49.00 147.00 12.00 40.00 1.00 9.00
Gilsum 1.00 1.00 24.00 45.00 10.00 29.00 15.00 4.00 1.00 3.00
Harrisville 1.00 1.00 43.00 162.00 45.00 112.00 17.00 10.00 0.00 0.00
Hinsdale 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Jaffrey 1.00 1.00 144.00 253.00 107.00 415.00 22.00 65.00 0.00 4.00
Keene 5.00 5.00 739.00 1717.00 575.00 1484.00 216.00 214.00 7.00 12.00 1.00
Marlborough 1.00 1.00 100.00 174.00 67.00 154.00 13.00 20.00 1.00 1.00
Marlow 1.00 1.00 23.00 70.00 15.00 45.00 4.00 14.00 0.00 2.00
Nelson 1.00 1.00 28.00 73.00 22.00 52.00 17.00 13.00 1.00 0.00
Richmond 1.00 1.00 29.00 74.00 31.00 66.00 6.00 13.00 0.00 1.00
Rindge 1.00 1.00 95.00 219.00 68.00 304.00 43.00 74.00 2.00 9.00
Roxbury 1.00 1.00 9.00 22.00 7.00 20.00 5.00 2.00 0.00 1.00
Stoddard 1.00 1.00 33.00 47.00 40.00 72.00 11.00 18.00 0.00 2.00
Sullivan 1.00 1.00 25.00 48.00 19.00 41.00 4.00 8.00 0.00 2.00
Surry 1.00 1.00 26.00 39.00 23.00 50.00 7.00 22.00 1.00 0.00
Swanzey 1.00 1.00 199.00 336.00 123.00 364.00 18.00 52.00 2.00 3.00 1.00
Troy 1.00 1.00 59.00 86.00 46.00 145.00 6.00 15.00 0.00 4.00
Walpole 1.00 1.00 93.00 359.00 108.00 233.00 6.00 44.00 0.00 4.00
Westmoreland 1.00 1.00 61.00 172.00 52.00 105.00 10.00 23.00 1.00 1.00
Winchester 1.00 1.00 78.00 172.00 90.00 203.00 10.00 27.00 1.00 3.00 1.00
County Totals 27.00 26.00 2109.00 4824.00 1678.00 4499.00 497.00 747.00 20.00 65.00

Coos
Berlin 4.00 4.00 343.00 394.00 245.00 1149.00 6.00 143.00 2.00 18.00 1.00
Carroll 1.00 1.00 26.00 36.00 25.00 61.00 0.00 10.00 0.00 0.00
Clarksville 1.00 1.00 4.00 12.00 2.00 7.00 0.00 1.00 0.00 0.00
Colebrook 1.00 1.00 65.00 61.00 23.00 71.00 0.00 12.00 2.00 0.00
Columbia 1.00 1.00 4.00 24.00 11.00 21.00 0.00 1.00 0.00 2.00
Dalton 1.00 1.00 13.00 38.00 9.00 38.00 1.00 13.00 0.00 0.00
Dixville 1.00 1.00 8.00 1.00 2.00 3.00 0.00 1.00 0.00 0.00
Dummer 1.00 1.00 7.00 13.00 6.00 14.00 1.00 2.00 0.00 1.00
Errol 1.00 1.00 7.00 9.00 6.00 18.00 0.00 2.00 0.00 0.00 1.00
Gorham 1.00 1.00 101.00 125.00 94.00 320.00 3.00 22.00 0.00 28.00 1.00
Jefferson 1.00 1.00 9.00 43.00 9.00 43.00 1.00 11.00 0.00 0.00
Lancaster 1.00 1.00 38.00 158.00 60.00 180.00 3.00 18.00 2.00 2.00
Milan 1.00 1.00 34.00 46.00 24.00 100.00 0.00 8.00 0.00 2.00 1.00
Millsfield 1.00 1.00 1.00 1.00 1.00 1.00 0.00 0.00 0.00 1.00
Northumberland 1.00 1.00 76.00 81.00 40.00 139.00 2.00 10.00 0.00 0.00
Pittsburg 1.00 1.00 11.00 5.00 4.00 26.00 0.00 7.00 0.00 2.00
Randolph 1.00 1.00 4.00 56.00 8.00 42.00 3.00 9.00 0.00 0.00
Sharon 1.00 1.00 12.00 39.00 7.00 32.00 3.00 5.00 0.00 0.00
Shelburne 1.00 1.00 6.00 26.00 14.00 28.00 1.00 5.00 0.00 0.00
Stark 1.00 1.00 9.00 17.00 10.00 34.00 0.00 2.00 0.00 1.00
Stewartstown 1.00 1.00 10.00 17.00 7.00 30.00 0.00 2.00 1.00 1.00
Stratford 1.00 1.00 16.00 25.00 7.00 35.00 0.00 0.00 0.00 0.00
Wentworth Location 1.00 1.00 1.00 0.00 0.00 3.00 1.00 0.00 0.00 0.00
Whitefield 1.00 1.00 46.00 86.00 49.00 106.00 4.00 22.00 1.00 1.00
County Totals 27.00 27.00 851.00 1313.00 663.00 2501.00 29.00 306.00 8.00 59.00

Grafton
Alexandria 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Ashland 1.00 1.00 37.00 91.00 32.00 119.00 5.00 24.00 2.00 2.00 1.00
Bath 1.00 1.00 4.00 49.00 8.00 37.00 1.00 7.00 0.00 0.00
Benton 1.00 1.00 3.00 14.00 4.00 8.00 0.00 0.00 0.00 0.00
Bethlehem 1.00 1.00 55.00 201.00 41.00 141.00 13.00 22.00 2.00 0.00
Bridgewater 1.00 1.00 21.00 44.00 11.00 73.00 5.00 8.00 0.00 0.00
Bristol 1.00 1.00 89.00 102.00 64.00 165.00 6.00 41.00 0.00 7.00
Campton 1.00 1.00 73.00 170.00 43.00 167.00 3.00 47.00 3.00 1.00
Canaan 1.00 1.00 54.00 240.00 51.00 193.00 7.00 20.00 0.00 3.00 1.00
Dorchester 1.00 1.00 2.00 22.00 5.00 9.00 2.00 4.00 0.00 0.00
Easton 1.00 1.00 5.00 44.00 10.00 27.00 0.00 4.00 1.00 0.00
Ellsworth 1.00 1.00 7.00 5.00 3.00 7.00 0.00 0.00 0.00 0.00
Enfield 1.00 1.00 68.00 335.00 82.00 257.00 7.00 26.00 1.00 6.00
Franconia 1.00 1.00 26.00 123.00 28.00 84.00 5.00 12.00 1.00 0.00
Grafton 1.00 1.00 22.00 90.00 19.00 57.00 0.00 4.00 0.00 2.00 1.00
Groton 1.00 1.00 7.00 26.00 4.00 18.00 2.00 9.00 0.00 0.00
Hanover 1.00 1.00 453.00 1499.00 302.00 1010.00 41.00 116.00 11.00 4.00 1.00
Haverhill 1.00 1.00 41.00 149.00 26.00 129.00 3.00 5.00 0.00 5.00
Hebron 1.00 1.00 14.00 36.00 13.00 33.00 0.00 7.00 0.00 2.00
Holderness 1.00 1.00 71.00 166.00 42.00 148.00 7.00 33.00 1.00 4.00
Landaff 1.00 1.00 4.00 18.00 5.00 18.00 0.00 1.00 0.00 0.00
Lebanon 3.00 3.00 233.00 1192.00 232.00 810.00 44.00 103.00 4.00 9.00 1.00
Lincoln 1.00 1.00 36.00 51.00 37.00 100.00 4.00 23.00 0.00 3.00
Lisbon 1.00 1.00 14.00 29.00 24.00 54.00 1.00 0.00 0.00 2.00
Littleton 1.00 1.00 83.00 239.00 89.00 272.00 6.00 42.00 1.00 6.00 1.00
Lyman 1.00 1.00 5.00 34.00 2.00 24.00 1.00 4.00 1.00 1.00
Lyme 1.00 1.00 49.00 333.00 37.00 139.00 5.00 19.00 2.00 0.00
Monroe 1.00 1.00 12.00 41.00 7.00 29.00 2.00 6.00 0.00 2.00
Orange 1.00 1.00 6.00 31.00 6.00 20.00 0.00 1.00 0.00 0.00
Orford 1.00 1.00 17.00 139.00 18.00 59.00 4.00 9.00 0.00 1.00
Piermont 1.00 1.00 12.00 82.00 11.00 32.00 6.00 6.00 0.00 0.00
Plymouth 1.00 1.00 87.00 383.00 100.00 274.00 21.00 56.00 4.00 1.00
Rumney 1.00 1.00 29.00 109.00 26.00 67.00 7.00 24.00 0.00 1.00
Sugar Hill 1.00 1.00 30.00 67.00 15.00 37.00 2.00 6.00 0.00 0.00
Thornton 1.00 1.00 88.00 119.00 49.00 135.00 5.00 36.00 0.00 2.00
Warren 1.00 1.00 13.00 37.00 18.00 25.00 4.00 4.00 0.00 4.00
Waterville Valley 1.00 1.00 9.00 29.00 11.00 25.00 1.00 10.00 0.00 0.00
Wentworth 1.00 1.00 11.00 40.00 4.00 35.00 1.00 3.00 0.00 1.00
Woodstock 1.00 1.00 29.00 55.00 27.00 75.00 0.00 17.00 0.00 1.00
County Totals 41.00 40.00 1819.00 6434.00 1506.00 4912.00 221.00 759.00 34.00 70.00

Hillsborough
Amherst 1.00 1.00 283.00 611.00 258.00 802.00 20.00 191.00 3.00 3.00 1.00
Antrim 1.00 1.00 64.00 170.00 66.00 142.00 9.00 27.00 2.00 11.00
Bedford 1.00 1.00 355.00 638.00 344.00 1113.00 15.00 478.00 4.00 16.00 1.00
Bennington 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Brookline 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00
Deering 1.00 1.00 53.00 79.00 42.00 106.00 6.00 27.00 0.00 3.00
Francestown 1.00 1.00 44.00 151.00 31.00 104.00 4.00 32.00 1.00 0.00
Goffstown 1.00 1.00 332.00 478.00 335.00 1018.00 26.00 343.00 5.00 27.00 1.00
Greenfield 1.00 1.00 32.00 86.00 39.00 101.00 7.00 11.00 0.00 1.00
Greenville 1.00 1.00 37.00 73.00 39.00 124.00 6.00 16.00 0.00 2.00
Hancock 1.00 1.00 41.00 219.00 58.00 158.00 20.00 31.00 0.00 0.00
Hillboro 1.00 1.00 111.00 188.00 136.00 281.00 11.00 60.00 0.00 8.00
Hollis 1.00 1.00 200.00 449.00 178.00 556.00 21.00 158.00 2.00 16.00 1.00
Hudson 1.00 1.00 412.00 639.00 374.00 1388.00 21.00 318.00 6.00 15.00 1.00
Litchfield 1.00 1.00 140.00 215.00 129.00 458.00 3.00 123.00 3.00 3.00 1.00
Lyndborough 1.00 1.00 40.00 114.00 48.00 101.00 9.00 24.00 0.00 3.00
Manchester 12.00 12.00 1982.00 3086.00 2154.00 6365.00 172.00 2317.00 30.00 153.00 1.00
Mason 1.00 1.00 30.00 91.00 34.00 86.00 8.00 25.00 0.00 1.00
Merrimack 1.00 1.00 594.00 797.00 560.00 1774.00 27.00 445.00 1.00 27.00 1.00
Milford 1.00 1.00 268.00 497.00 260.00 894.00 23.00 174.00 4.00 7.00 1.00
Mont Vernon 1.00 1.00 68.00 112.00 65.00 176.00 11.00 30.00 2.00 2.00
Nashua 9.00 9.00 1499.00 2694.00 1741.00 6178.00 106.00 1338.00 26.00 57.00 1.00
New Boston 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
New Ipswich 1.00 1.00 72.00 122.00 64.00 227.00 4.00 60.00 1.00 4.00 1.00
Pelham 1.00 1.00 187.00 317.00 176.00 816.00 11.00 175.00 5.00 10.00 1.00
Peterborough 1.00 1.00 229.00 580.00 141.00 509.00 52.00 90.00 2.00 8.00 1.00
Temple 1.00 1.00 38.00 151.00 35.00 87.00 8.00 20.00 1.00 0.00
Weare 1.00 1.00 165.00 316.00 132.00 416.00 14.00 107.00 1.00 6.00 1.00
Wilton 1.00 1.00 103.00 257.00 88.00 260.00 67.00 69.00 0.00 8.00
Windsor 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
County Totals 49.00 45.00 7379.00 13130.00 7527.00 24240.00 681.00 6689.00 99.00 391.00

Merrimack
Allenstown 1.00 1.00 116.00 137.00 104.00 341.00 4.00 71.00 0.00 9.00 1.00
Andover 1.00 1.00 55.00 180.00 81.00 170.00 5.00 23.00 0.00 4.00
Boscawen 1.00 1.00 90.00 194.00 82.00 163.00 6.00 50.00 0.00 1.00
Bow 1.00 1.00 173.00 382.00 234.00 606.00 17.00 158.00 2.00 5.00 1.00
Bradford 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
Canterbury 1.00 1.00 68.00 247.00 63.00 214.00 17.00 44.00 1.00 3.00 1.00
Chichester 1.00 1.00 44.00 128.00 59.00 192.00 10.00 31.00 1.00 0.00
Concord 10.00 10.00 1005.00 2445.00 1142.00 3439.00 119.00 549.00 10.00 50.00 1.00
Danbury 1.00 1.00 19.00 74.00 16.00 49.00 2.00 10.00 2.00 3.00
Dunbarton 1.00 1.00 67.00 140.00 63.00 148.00 12.00 37.00 0.00 2.00
Epsom 1.00 1.00 101.00 135.00 98.00 284.00 7.00 75.00 1.00 3.00 1.00
Franklin 3.00 3.00 172.00 280.00 180.00 527.00 14.00 105.00 0.00 9.00
Henniker 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00
Hill 1.00 1.00 23.00 41.00 27.00 76.00 0.00 10.00 0.00 2.00
Hooksett 1.00 1.00 224.00 332.00 222.00 685.00 15.00 219.00 3.00 4.00 1.00
Hopkinton 1.00 1.00 174.00 484.00 149.00 569.00 26.00 113.00 2.00 1.00 1.00
Loudon 1.00 1.00 108.00 159.00 106.00 272.00 11.00 66.00 1.00 8.00 1.00
Newbury 1.00 1.00 51.00 138.00 47.00 154.00 5.00 34.00 0.00 3.00 1.00
New London 1.00 1.00 112.00 383.00 107.00 310.00 16.00 61.00 0.00 4.00
Northfield 1.00 1.00 102.00 154.00 89.00 276.00 7.00 64.00 0.00 4.00
Pembroke 1.00 1.00 155.00 264.00 170.00 482.00 10.00 90.00 1.00 4.00 1.00
Pittsfield 1.00 1.00 58.00 93.00 68.00 208.00 2.00 54.00 1.00 4.00 1.00
Salisbury 1.00 1.00 31.00 55.00 41.00 110.00 0.00 18.00 1.00 3.00
Sutton 1.00 1.00 50.00 154.00 54.00 150.00 6.00 45.00 0.00 3.00
Warner 1.00 1.00 77.00 268.00 54.00 230.00 12.00 40.00 0.00 2.00
Webster 1.00 1.00 46.00 83.00 56.00 130.00 7.00 44.00 3.00 0.00
Wilmot 1.00 1.00 42.00 136.00 29.00 105.00 12.00 23.00 1.00 0.00
County Totals 38.00 36.00 3163.00 7086.00 3341.00 9890.00 342.00 2034.00 30.00 131.00

Rockingham
Atkinson 1.00 1.00 112.00 233.00 125.00 573.00 4.00 153.00 1.00 4.00 1.00
Auburn 1.00 1.00 126.00 159.00 90.00 302.00 7.00 122.00 3.00 5.00
Brentwood 1.00 1.00 62.00 138.00 66.00 225.00 8.00 62.00 0.00 5.00
Candia 1.00 1.00 103.00 169.00 78.00 235.00 4.00 87.00 1.00 2.00 1.00
Chester 1.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.00
Danville 1.00 1.00 63.00 94.00 53.00 257.00 6.00 61.00 1.00 10.00 1.00
Deerfield 1.00 1.00 120.00 204.00 66.00 258.00 17.00 71.00 1.00 2.00 1.00
Derry 1.00 1.00 442.00 757.00 524.00 1702.00 29.00 413.00 8.00 14.00 1.00
East Kingston 1.00 1.00 38.00 58.00 31.00 152.00 3.00 24.00 0.00 2.00 1.00
Epping 1.00 1.00 198.00 217.00 107.00 335.00 8.00 81.00 2.00 3.00 1.00
Exeter 1.00 1.00 364.00 843.00 322.00 1153.00 22.00 210.00 2.00 13.00 1.00
Fremont 1.00 1.00 54.00 95.00 50.00 206.00 3.00 54.00 1.00 5.00
Greenland 1.00 1.00 74.00 149.00 84.00 273.00 7.00 67.00 0.00 2.00 1.00
Hampstead 1.00 1.00 150.00 287.00 158.00 575.00 6.00 156.00 7.00 6.00 1.00
Hampton 1.00 1.00 299.00 646.00 397.00 1554.00 17.00 297.00 6.00 19.00
Hampton Falls 1.00 1.00 38.00 96.00 34.00 129.00 5.00 46.00 1.00 0.00
Kensington 1.00 1.00 38.00 136.00 48.00 155.00 8.00 38.00 0.00 6.00
Kingston 1.00 1.00 91.00 156.00 96.00 364.00 10.00 71.00 2.00 8.00 1.00
Londonderry 1.00 1.00 382.00 762.00 402.00 1477.00 20.00 399.00 6.00 5.00
New Castle 1.00 1.00 37.00 77.00 33.00 97.00 1.00 36.00 0.00 3.00
Newfields 1.00 1.00 65.00 107.00 38.00 127.00 3.00 32.00 0.00 3.00 1.00
Newington 1.00 1.00 14.00 55.00 22.00 47.00 3.00 11.00 2.00 2.00 1.00
Newmarket 1.00 1.00 247.00 585.00 297.00 696.00 43.00 142.00 4.00 4.00 1.00
Newton 1.00 1.00 82.00 113.00 78.00 306.00 4.00 61.00 4.00 4.00
North Hampton 1.00 1.00 114.00 204.00 127.00 384.00 9.00 85.00 2.00 5.00 1.00
Northwood 1.00 1.00 102.00 197.00 94.00 219.00 21.00 60.00 0.00 3.00 1.00
Nottingham 1.00 1.00 121.00 182.00 105.00 281.00 14.00 58.00 3.00 5.00
Plaistow 1.00 1.00 106.00 248.00 123.00 472.00 5.00 139.00 1.00 10.00 1.00
Portsmouth 5.00 5.00 713.00 1601.00 647.00 1995.00 114.00 311.00 14.00 26.00 1.00
Raymond 1.00 1.00 206.00 270.00 158.00 473.00 10.00 127.00 7.00 2.00 1.00
Rye 1.00 1.00 148.00 324.00 143.00 437.00 17.00 89.00 6.00 13.00 1.00
Salem 1.00 1.00 435.00 798.00 416.00 2009.00 14.00 498.00 3.00 32.00 1.00
Sandown 1.00 1.00 88.00 175.00 87.00 282.00 5.00 84.00 1.00 4.00 1.00
Seabrook 1.00 1.00 98.00 184.00 121.00 549.00 7.00 68.00 0.00 0.00 1.00
South Hampton 1.00 1.00 29.00 51.00 20.00 76.00 6.00 20.00 0.00 1.00
Stratham 1.00 1.00 135.00 301.00 188.00 491.00 19.00 144.00 5.00 10.00 1.00
Windham 1.00 1.00 215.00 437.00 174.00 781.00 5.00 229.00 6.00 4.00 1.00
County Totals 41.00 40.00 5709.00 11108.00 5602.00 19647.00 484.00 4606.00 100.00 242.00

Strafford
Barrington 1.00 1.00 208.00 356.00 174.00 500.00 26.00 108.00 1.00 12.00 1.00
Dover 6.00 6.00 690.00 1350.00 679.00 2393.00 122.00 409.00 7.00 24.00
Durham 1.00 1.00 362.00 837.00 240.00 757.00 99.00 134.00 3.00 12.00 1.00
Farmington 1.00 1.00 135.00 159.00 107.00 288.00 10.00 58.00 1.00 5.00 1.00
Lee 1.00 1.00 128.00 318.00 123.00 406.00 31.00 59.00 2.00 3.00 1.00
Madbury 1.00 1.00 74.00 110.00 53.00 145.00 18.00 34.00 0.00 1.00
Middleton 1.00 1.00 28.00 37.00 37.00 89.00 3.00 17.00 1.00 1.00
Milton 1.00 1.00 63.00 149.00 102.00 202.00 5.00 45.00 2.00 4.00 1.00
New Durham 1.00 1.00 43.00 98.00 43.00 151.00 2.00 39.00 0.00 4.00
Rochester 6.00 6.00 548.00 714.00 596.00 1648.00 27.00 391.00 7.00 25.00 1.00
Rollinsford 1.00 1.00 53.00 140.00 76.00 200.00 9.00 37.00 0.00 4.00
Somersworth 5.00 5.00 188.00 359.00 291.00 1007.00 23.00 152.00 1.00 26.00 1.00
Strafford 1.00 1.00 103.00 196.00 103.00 275.00 9.00 70.00 1.00 6.00
County Totals 27.00 27.00 2623.00 4823.00 2624.00 8061.00 384.00 1553.00 26.00 127.00

Sullivan
Acworth 1.00 1.00 14.00 97.00 26.00 48.00 19.00 12.00 0.00 2.00
Charlestown 1.00 1.00 58.00 334.00 97.00 307.00 5.00 37.00 0.00 3.00
Claremont 3.00 3.00 161.00 740.00 296.00 813.00 19.00 95.00 2.00 34.00
Cornish 1.00 1.00 30.00 195.00 38.00 112.00 13.00 9.00 1.00 0.00
Croydon 1.00 1.00 10.00 48.00 8.00 36.00 0.00 0.00 0.00 6.00
Goshen 1.00 1.00 11.00 39.00 11.00 37.00 4.00 0.00 0.00 3.00
Grantham 1.00 1.00 68.00 266.00 72.00 213.00 3.00 46.00 0.00 2.00 1.00
Langdon 1.00 1.00 8.00 53.00 18.00 40.00 2.00 7.00 0.00 1.00
Lempster 1.00 1.00 19.00 63.00 16.00 34.00 4.00 5.00 0.00 2.00
Newport 1.00 1.00 100.00 293.00 115.00 288.00 9.00 44.00 0.00 2.00
Plainfield 1.00 1.00 37.00 365.00 63.00 148.00 14.00 31.00 1.00 3.00
Sprngfield 1.00 1.00 21.00 96.00 20.00 64.00 3.00 12.00 0.00 1.00
Sunapee 1.00 1.00 51.00 265.00 71.00 206.00 8.00 43.00 2.00 4.00 1.00
Unity 1.00 1.00 23.00 89.00 28.00 67.00 15.00 9.00 1.00 0.00
Washington 1.00 1.00 20.00 65.00 26.00 47.00 4.00 16.00 0.00 0.00
County Totals 17.00 17.00 631.00 3008.00 905.00 2460.00 122.00 366.00 7.00 63.00

State Totals 301.00 292.00 26554.00 56353.00 25849.00 82594.00 3015.00 18392.00 342.00 1247.00

[Ravy]

21. You should also figure in the average income of the precinct...

I think you will probably find that the higher average income of the precinct you will find the optical equipment. This equipment is more expensive and therefore usually used in more affluent jurisdictions.

If there is the correlation there, I would think it is more likely that the more affluent voted for Kerry and the less affluent voted for Dean.

Without analyzing other correlating factors it appears that you are indicating that the system used in voting was somehow swaying the outcome. I would look for other, more probable correlations.

You may be entirely right. Eliminating the other factors will help you prove your point.

Ravy

[seaglass]

35. The other thing to look at is regions.

Dean won in high population areas with optical scan machines on the Vermont border but these areas aren't as highly populated as the MA. border region that Kerry won.

[creativelcro]

36. True: what killed Dean were two towns bordering MA that

had many registered voters (relatively speaking): Hillsborough and Rockingham. Given that the polls were so close to the actual results, it is hard to believe that the noted correlation indicated cheating (unless the pollsters are in it too :). I posted a link to the Excel dataset I put together in this thrad.

[LisaM]

15. I thought optical screen ballots were pretty accurate

You can review them once you're finished (though I suppose you could rig the software counting them.) I thought it was the touch screen machines that posed the problem, because there is no paper and no way to recount.

[creativelcro]

16. Both optical scanners and TS can be a problem

With optical scanners you can have recounts, but in practice it is almost impossible....

[Eloriel]

32. Not necessarily

There were a lot of anomalies in FL 2000 that we never heard about for all the "hanging and pregnant chad" business (which IMO, looking back, may very well have been classic distraction). Plus, remember how James Baker & Co. were hysterically fighting recounts?

Here's the deal. Optical scans are no more "safe" than DREs except that they do inherently provide a paper ballot which records the voter's actual preferences. They run on the same software as touchscreens in Diebold's case, and are also just as vulnerable to errors in programming, malicious code, and realtime fraud via phone connections or wireless capabilities (if present).

Further, judging from the Diebold tech manuals, they require a lot of programming (locally, as a rule) to mesh with state law. For example, one that I recall because I could see so much room for mischief, requires the user to select what to do about undervotes and overvotes -- whether they are ignored as unread, whether the ballot's other races are counted, whether they are unread and alerted so they can be re-scanned or manually counted. There were reports in FL2000 that some of the heavily Dem counties or precincts automatically rejected these WITHOUT alerting the user so those ballots could be read.

These are definitely the machines I prefer simply because they START with a voter-verified paper ballot, but they are NOT "safe" by any means from either programming error or potential tampering and fraud.

It's also my understanding that the machines in use in NH are Accuvote, which I believe are Diebold. If that doesn't send shivers down your spine, it should. California's recent audit of all its voting systems found that NONE of the Diebold machines (DREs or optical scan) had certified software on them during the recall election (not that being "certified" means anything, but it is a violation of state law that UNcertified software was loaded on the machines).

Also, it was Diebold software on an optical scanner that accounted for that infamous MINUS 16,000 votes for Gore in Volusia County, I think it was.

Lots of ways things can go wrong -- accidentally or on purpose.

[progressivebydesign]

18. Something has been bothering me about this..

It's been bothering me that Kerry seems to be getting a light treatment, compared to what Dean has gotten. But then again, I guess all the candidates have been treated well.. except for Dean.

One of my lame theories, which may prove to be true, is that Hannity and others have proclaimed a "bombshell" about Kerry that will be revealed after he gets the nomination. I have a feeling it's something that won't matter.. seems that the people here on DU are informed enough that someone would have uncovered it by now. However, If the republicans think they have the goods on Kerry, and they think the media will beat the Americans over the head with it (like Gore's "lies", and Dean's "anger" both nonexistant!) then it would make sense to want him to run against Bush. Perhaps, because Kerry is more known to them, and part of their insider culture, they will have an easier time painting him with whatever image they want to, in contrast with Bush.

I think the GOP is afraid of Dean.. not necessarily because they think he can beat Bush when the votes are "counted", but because he represents a return to 60's style independent thinking. He represents, by his true grassroots campaign and people-powered popularity, a rejection of the republicans drive to take over the country. We are fighting for the soul of America, and they know it. Dean has the audacity to say that we can take back the American flag from the likes of John Ashcroft, Rush Limbaugh, and Jerry Falwell. They are scared because he is has a huge voice, and he's rejecting the core of their doctrine. He isn't just talking issues.. he's talking real ideaology.. and it scares them. Outsiders scare them. That's how Clinton kicked Bush's ass.. and that's the appeal of Bush (even though we know that his outsider status was a lie)

[stickdog]

73. What? That Kerry had plastic surgery? That he dyes his hair? That

he's a gold digger? That he threw away someone else's medals in a famous Vietnam protest?

That he tried to get his first marriage annulled? That he cheated on his first wife for years? That he's pulled all sorts of electoral dirty tricks?

That he's anti-guns? That he's a boring, patrician MA tax & spend liberal?

[4dog]

19. Please say more about your data to help wider understanding

1) Good stuff, but write something also for people who don't know what a vector is.

2) On the stat side, did you try the correlation using percentages? Also give your data sources for people who may want to crunch numbers themselves. Perhaps you can park this in a slower forum or if you have new data you could re-post here.

3) Exit polls had Dean and Kerry neck and neck until the evening tallies suddenly jumped Kerry by about 13%. We have seen similar patterns before in other elections. Let us encourage those with expertise and concerned about voting accuracy to dig into this (anybody here?).

[creativelcro]

26. Ok, I'll look into it later today. Perhaps post an Excel file somewhere.

[veggiemama]

20. Have you compared these results to the Shaheen/Sununu race

in 2002? Shaheen not only polled ahead of Sununu, but she was projected to win. I think you might find similar statistics there. IRRC, the towns with the optical scanners put Sununu in the Senate. Might be worth a look.

[creativelcro]

25. I'll look into this more later today. Gotta finish some other work....

So far, I like the suggestion of looking at income as another predictor, in a multiple regression type of analysis (the simple linear regression I did was something quicky, just to see if there was anything at all).

[Eloriel]

33. Please also look, if you can, at some of the other numbers --

votes going to the bottom finishers. If they're rigging an election (and I'm not saying they are), they have to keep the same number of votes (rather than engage in ballot stuffing) to match the number of voters -- so those votes have to go somewhere. I'd be VERY interested in knowing if there's any unusual figures there.

Good work, Creativelcro!

[4dog]

23. Will try to send this to BBV.org

See what they can make of it. It definitely looks skewed.

[ieoeja]

27. Assuming urban more likely to have optical.

Then you're saying Kerry does better in the urban areas the Democratic party will probably carry regardless of our candidate while Dean does better in the rural areas where we do not do so well.

[DFLforever]

30. You may also want to consider

whether in these optical scan districts, any third candidate has gotten a significantly greater number of votes than he did elsewhere. Taking votes from one candidate and dumping them to another could be the most efficient way to fix the outcome in a multi-candidate race.

[creativelcro]

34. Ok, placed the complete raw data Excel worksheet here:

Feel free to analyze this to death. I'm still recovering from the flu and I'm not sure exactly how much more I'll be working on this this week. Please report mistakes, if you notice some.

http://www.geocities.com/mrigurl

[4dog]

37. Thanks for the spreadsheet

Will try to do a little tinkering with it.

By the way, your morphing image on the same page is making me a little queasy. I had to hide it.

[lostnfound]

56. Yes, your morph is too much on an empty stomach! nt

[creativelcro]

59. Crap. Sorry, did that a while back... Just a site for temporary files

to share...

[On the Road]

40. Eyeballed the Spreadsheet

Just downloaded, compulted the percentages by precinct, and it looked surprisingly consistent. The biggest outlier was Wesley Clark winning 53% of one small rural precinct. That was it.

Frankly, I expected that there would be geographic pockets of support for different candidates, but in general, it was just different shades of gray. Nothing jumped out. Try it yourself.

[creativelcro]

41. Sorry, not sure what you did, but there are major geographic pockets.

Dean won big in Hanover, Lebanon and the West. Kerry the South/east. Clark won in one small precint north (near some Academy, i think, not sure). What did you plot ?

[On the Road]

63. Sure, There are Geographical Pockets of Support

but there were no precincts that looked like the vote had been manipulated.

OK, Clark wins one small precinct with 53%. Dean is strong in some areas, Kerry in others. Nothing unusual there.

Let me give a couple of examples of what would look strange: (1) In the California recall vote, there were precincts in which a number of minor candidates got ten times the support as they did statewide. (2) In the 1972 Ohio primary, there were urban black precincts which supported Humphrey over McGovern by 99%. Hunter Thompson was reporting from the McGovern campaign and they felt it was fraud.

(3) In the Florida recount, one northeastern county (Volusia?) reported no double votes, but over 10,000 triple votes. (4) The original results in Florida from a southeastern county had Al Gore with negative 16,000 votes and minor candidates with artificially high results. After Gore's concession speech, the results were rebroadcast with the same total but different distribution.

Those patterns all suggest either fraud or some kind of horrible error. I saw nothing in the NH results reminiscent of those situations.

[stickdog]

71. Suppose in all opscan areas, 2% of Dean's vote was "awarded" to

Lieberman and 4% of Dean's vote was "awarded" to Kerry?

[cthrumatrix]

38. maybe they can "recount a few" to validate the numbers..(see if all is ok)

[bigtree]

39. You brought this up before to insinuate voter fraud

This is the worst political tactic in the book. It makes you look bitter and whiny. Get over it and get busy for your candidate.

[creativelcro]

42. bigtree, I posted the numbers. Voter fraud is always a possibility.

So, there is nothing unethical about looking at the final numbers. Everybody should look at them and see if there are weird patterns...

[bigtree]

43. All kinds of things happen in elections

one sure pattern: High turnout from energy started by Howard Dean and his supporters. Benefits all the candidates. All of the party. Other campaigns got in and caught the buzz. Good for the party. Way to go Howard!

[harmonyguy]

44. Online Sample Ballots?

There are numerous ways to 'fudge' optical scanned votes, and it doesn't have to be voter fraud. The key thing would be to do a manual count of the original scanned ballots, in precincts with questionable results. The recount cannot be based on the so-called 'ballot images' held by the system.

Are there online sample ballots (pdfs) available anywhere for the NH primary? I've looked but haven't found any so far.

HG

[theboss]

46. Voter fraud does not result in a 13-point spread

Not even Tammany Hall cheated like that.

[creativelcro]

47. Ok, makes sense. But the Gore "glitch" in FL initially

costed Gore 16K votes... With software, flipping one vote is not less effortful than flipping 10k votes... Nobody is saying there is voter fraud in this case, or not of this type... The numbers are out there, for the record... That's all...

[cthrumatrix]

48. we are talking about "fraud and perception" it effects voters

[creativelcro]

49. Well, I think that some random recounts should ALWAYS

be conducted. To keep it honest, there should be a random selection conducted AFTER the election and votes from those precints recounted.

[4dog]

50. The issue of "random" recounts is being discussed

by Bev Harris and her band at blackboxvoting.org. It appears that the academics or other experts (e.g., accountants, auditors) do not yet have a good solution in hand to recommend to our legislators, but Bev feels that more than a small number of random recounts will be necessary to prevent fraud or error. It's an important issue (one of many) to keep in mind as the struggle continues to define and legislate how to enable accurate voting.

[creativelcro]

51. I totally agree... See? Now we are wondering if there was something fishy

The only way the losers can accept that for good is if the system can be PROVEN to be honest.

[stickdog]

72. Wasn't this Kerry's advice about Gore's win in Florida as well?

Very telling, wouldn't you agree?

[Polemonium]

52. nice p-value

but p value alone just tells me it's worth looking at. Need more of the story. Correlation ain't causation etc. Next obvious step aside from clear reporting of methods and results would be to check if the locations of machines are correlated to traits found in Kerry supporters.

[creativelcro]

61. I agree with you. Also, those values were calculated when things

were 97% or so. But the worksheet has the full dataset at 100% with the rest of the info.

[Stuckinthebush]

70. PPMCCs p values are very sensitive to sample size

You have a large sample here, so it is very likely that smaller correlations will show very significant p-values.

Just a thought.

[John_H]

53. If I were Rove's BBV Op, i'd be setting those machines to Dean (n/t)

[stickdog]

74. Isn't it obvious by now that Rove is most afraid of Dean?

Otherwise, while are they still piling on?

[Dolphyn]

54. Percentages

According to my calculations, here are the percentages for each type of district. I find it "interesting" that Kucinich got 1.83% with the non-optical method but only 1.24% with the optical machines.

Non-opt Optical Candidate
11.98 12.50 Clark
30.56 24.60 Dean
11.80 12.21 Edwards
35.93 39.55 Kerry
1.83 1.24 Kucinich
7.17 9.17 Lieberman
0.15 0.16 Sharpton
0.59 0.57 Others

100.00 100.00 Total

[nbsmom]

55. The non-opt column is closer to the exit polls...

I think that's significant...

[Dolphyn]

57. ... but the majority of votes were optical

About 73% of the votes were "optical," so we would expect the exit polls to be more closely aligned with the "optical" votes.

If the exit polls match the "non-optical" votes more closely than the "optical" ones, that's very spooky indeed. Unless, of course, they just happened to poll more "non-optical" precincts somehow.

[creativelcro]

58. Yup, southern towns have many more voters... And they happened

to have optical... Should also check out how many registered voters there are and how many actually voted.
if you find problems with the worksheet, please report and I'll fix them and repost it. Thanks!

[stickdog]

78. No. Thank you.

I'm pretty sure that you are on to something here.

[stickdog]

75. The OPSCAN counters gave 4% of Dean's vote to Kerry.

and 2% of Dean's to Lieberman.

[THUNDER HANDS]

60. all the scan machines have paper ballots

So please take your conspiracy theories elsewhere.

[stickdog]

76. This means nothing unless we manually recount these ballots.

But you knew that.

[sangh0]

62. optical scan machines tend to be in higher income areas

and higher-income voters voted for Kerry

[stickdog]

77. They also tend to be in areas where people are more educated.

And the more educated voters voted for Dean.

[Snow]

64. Friday afternoon entertainment....

Okay, I took your spreadsheet from the page with the weird graphic & plugged it into my favorite statistical package, STATA v.7. I don't recall which voting district was coded which, so I guess I did a blind analysis. First, it seems all the candidates gathered more votes in type '0' districts than in type '1'. Fair enough, there're more voters in type 0 than type 1. So, the question is, is that difference more pronounced for some candidates than for others? To get at that, everything needs to be on the same scale, so I standardized the count data - that is, transformed them into z scores. When looked at that way, only one candidate, Clark, looks much different from the others, and that's not by much. I included each income indicator in the regression (actually ANCOVA) model to check for that effect, and there was none. The income variables were not associated with the vote counts for any candidate, not were did they change the association of Voting Type with vote counts. So, the difference between type 0 and type 1 districts were the same for each candidate, and those differences were not due to income levels.

[stickdog]

66. Please include the margin. And include Lieberman's percentage.

Places in which Kerry beat Dean by more than 20% strongly correlate with places where Lieberman received a higher than normal percentage of the vote.

[ramblin_dave]

79. So what's the verdict, if any?

Are the NH results odd by coincidence or is anyone looking further? At any rate I salute the effort. All election results need to be looked at for any abnormalities.

[zulchzulu]

80. File under: "Sore Loser Department"

Deal with it.

[Carl Spackler]

81. I think if you've got the time to do this

you might also run correlations on race, income, and median age of the counties and towns to see if that would not only explain this but also shed light on who is bringing in what kinds of voters.

Or maybe Kerry's a blood-brother of GWB. Did I read that Kathryn Harris recently vacationed in New Hampshire?