Well, what do you know?

Surprise, surprise: There seem to be some inaccuracies in New Hampshire's vote count after all. Brad Friedman has the update:

Numbers are now being posted from both the Democratic and Republican hand-counts in the NH Primary Election contest. So far, only wards in Manchester (Hillsborough County) have been hand-counted, and disparities between the original counts from the Diebold optical-scan machine and the hand inspections seem to be occurring in many wards, and for many candidates.

Here is the SoS Recount page with the totals, that I haven't yet been able to review in full.

While sources on the ground at the counting today have told me that officials were not announcing the originally counted results at the counting room, the SoS web page lists what they claim are the original counts --- previously verified by nobody --- versus the recounted numbers.

The disparities, as I've quickly been able to review them, are small, but consistent, in ward after ward, across almost all of the candidates. I'm told that the manufacturers of the optical-scan machines (in this case, Diebold) have estimated an expected error rate of 1% on this type of tallying device which, as noted by one of our contacts in NH, is ridiculous, if you consider that most states and counties only kick in "automatic recounts" when the margin between the two leading candidates is less than .5% or so.

ADDITIONALLY...Public records requests are being made on the spot, for errors and malfunctions at various voting precincts. An early review of the error forms turned over from the public record request made by Election Integrity experts overseeing the counting, has revealed that in Stratham there were some 550 ballots that were not read by the op-scan at all. They were seen as blank ballots. Officials there noticed the problem, and then hand-counted some 3000 ballots after the error was discovered.

Apparently, as we've seen elsewhere, voters were given the wrong pen to use and the op-scanners did not "see" this particular type of ink.

[snip]

LHS, apparently, is the one responsible for tracking (or not) and reporting (or not) any such errors, rather than the Secretary of State or local election officials, it would seem. That tracks with previous BRAD BLOG reporting on LHS, and how they operate in Connecticut, where there are similar concerns for whether or not the SoS even knows what the error rates are for the system they use, since problem reports are given to LHS instead of to public officials.

The BRAD BLOG has reported within the past few days machine problems during the election in a number of towns. In fact, of the first four towns we called that used the Diebold machines, all four reported machine failures of one type or another.

No one's saying at this point that there was deliberate tampering with the machines, or that the recount will change the outcome. But if you look at the recount report so far, you see changes in the vote totals just about across the board.

If it's YOUR vote that isn't counted because Diebold regards a 1% error rate as acceptable, that's disenfranchisement the same as if it's 100 votes that aren't counted. If the whole point of going to electronic voting machines was to ensure a more accurate count, and the counts are even LESS accurate than hand-counted votes, then you don't have to be a genius to do the math and see that these machines are just not an acceptable voting mechanism.

If you haven't already bookmarked Bradblog, I highly recommend you do so now. Because this is only one primary, there are 22 of them on Super Duper Hi Test Tuesday, and there's much more of this to come.