If you’re reading this early, come back later.

It is midnight, as late as some…uhhh, shall we just say, ‘veteran’ observers are willing to stay awake.

Maybe as late as they are able to stay awake.

According to the unofficial tally from WCAX-TV (which had far and away the best election night coverage in the state), Peter Shumlin is now ahead of Brian Dubie by 1,003 votes with almost a quarter of the precincts yet to report.

Shumlin said he was going to win. Could be, but the prudent observer will wait for some more precincts to report.

At this point, then, there is nothing to say. And while it is not a policy universally practiced in the journalism dodge, the preference here has always been that if there is nothing to say…don’t say anything.

So for you early birds, check back later. Maybe 10 AM.

And for all of you, there will be an extra posting tomorrow, Thursday.

Whew, it’s almost over. Don’t forget to vote, and would you like to know who’s going to win?

So would we all.

There is a new poll. The survey by Rasmussen Reports shows Democrat Peter Shumlin ahead of Republican Brian Dubie by 50 to 45 percent, with three percent undecided and one percent supporting minor candidates.

The poll was taken of a randomly selected sample of 750 likely voters and has a margin of error of plus or minus four percentage points.

Do not despair, Dubie-ites. Just taking the poll at face value, it does not “prove” that Shumlin is going to win. Remember, another poll taken two weeks earlier, by anther firm, gave Dubie a one-point lead. Only the count of actual ballots tomorrow night will reveal the winner (and if the race is as close as was the Democratic Primary, more counting may be needed).

What is often forgotten is that a poll’s margin of error applies not to the “spread,” but to each candidate’s percentage of support. All this poll says is that had all 250,000 or so likely voters been polled, between 54 and 46 percent would have chosen Shumlin, between 49 and 41 percent would have opted for Dubie.

So if on Tuesday night Dubie ends up with 47 percent of the vote to Shumlin’s 46 percent, the poll would not have been wrong. The result would have been within the poll’s margin of error.

(And would require the State Legislature to elect a governor, a matter to be discussed here only if it happens, which it probably won’t).

But don’t be of too much good cheer, either, Dubie-ites. A poll’s reported results represent the midpoint of its findings, which is far more likely to reflect actual public opinion than either extreme. It’s as likely that Shumlin is really ahead 53-42 as that Dubie is ahead 47-46.

Now we come to the question of whether this poll should be taken at face value. The answer will offer no comfort to Dubie and the Republicans, but the question should be explored anyway.

Rasmussen polls have a slight Republican bias. This is not because proprietor Scott Rasmussen has a slight Republican bias. Rasmussen in fact has a colossal Republican bias. But he’s also, on the basis of the evidence at hand, a perceptive businessman (he helped found ESPN), and for the sake of his business,  a pollster tries to make his polls accurate so that people take them seriously and continue to hire his firm. A series of inaccurate polls can put a pollster out of business.

But when a voter answers his or her phone because the Rasmussen folks are calling, an actual folk is not at the other end of the line. Instead, a recorded voice asks the respondent a question to be answered by pushing ‘one’ for candidate A, ‘two’ for candidate B, etc.

This ‘robocall’ technique is frowned on by the polling establishment at the American Association of Public Opinion Research . and scorned by several quality news organizations including the Washington Post and the New York Times. In fact, Rutgers University political scientist Cliff Zukin, in a paper he wrote last year for the AAPOR, argued that “these types of surveys have little claim of scientific validity and probably should not be reported.”

One reason for doubts about robocall polls is that they can not include random sampling within the household. A person on the phone can ask to speak to a woman or a man, or the person with the most recent birthday, or use some other method to try to get a representative sample.

But as Zukin noted, “If interviewers simply spoke with whoever answered the phone, the resulting samples would be older and more female than the population as a whole.”

That could be why robocall polls have been found to tilt about two percentage points more Republican than other surveys, or than the final election results.

On the other hand, Rasmussen polls don’t have a bad track record for accuracy. In the Massachusetts Senate race won by Republican Scott Brown earlier this year, Rasmussen polls picked up Brown’s growing strength earlier than other surveys.

With all the caveats, then, it’s fairly likely that Shumlin is ahead.

Which is not the same thing as predicting he will win. It might rain (though right now the Weather Channel is predicting pretty nice weather for both Burlington and Rutland on election day) and bad weather usually keeps more Democrats than Republicans away from the polls. The Republicans might have a better get-out-the-vote operation this year (though they usually don’t). And the poll could just be wrong. Sometimes, a poll is simply wrong. Polls, after all, are based on the laws of probability, not the laws of certainty.

Rasmussen is not the only outfit polling the race. So are both campaigns. “The campaign has maintained a consistent lead in our internal polling and as the campaigns (sic) momentum grows with each day we expect to win on Tuesday night,” said Dubie campaign manager Corry Bliss after the Rasmussen poll came out.

In the same statement, Bliss tried to argue that the Rasmussen poll showing Dubie behind by five points actually predicted an eight-point victory, because polls in 2002 had understated Vermont Republican strength by 13 points.

As responses go, this one was more inventive than persuasive.

Neither Bliss nor the Shumlin campaign revealed any numbers from their internal polling, requiring the curious to engage in tea-leaf reading to try to figure out what the polls might show.

Until the weekend, there were some signs that Dubie might be ahead. Conventional political strategy is to shift, in the final week or two, from television advertisements attacking the opposition to more positive ads, often featuring the candidate talking directly to the camera.

Last week the Shumlin campaign and its allies were still running anti-Dubie ads, while the Dubie campaign was featuring a spot called, “I’m your man” in which a blue-shirted, smiling, Dubie looked directly at the camera and promised to focus on jobs.

It’s a very effective ad even though its main point – “Vermont’s economy is in crisis” – is wrong. By all the usual criteria – economic growth (or, these days, shrinkage), unemployment rate, poverty rate, foreclosure rate – Vermont’s  economy is less in crisis than that of most other states.

This would be the downside of a Dubie governorship—the constant (and, worse, inaccurate)  griping about Vermont’s economy (there would be upsides, to be discussed later regardless of the outcome).

By the weekend, however, the Dubie campaign was back on the attack, including one ad – falsely accusing Shumlin of wanting to let prisoners out of jail – that has been widely criticized for being not just inaccurate but dishonest.

Considering that even many Republicans have been speculating that it is ads like this that seem to have put Dubie behind in the race, the decision to keep running it could be a sign of some desparation.

Maybe the Dubie campaign’s internal polls weren’t that different from the Rasmussen numbers.

WARNING: Turns out the News Guy is not the only Vermonter receiving  regular if not incessant emails claiming to be from the Electronic Federal Tax Payment System screaming: “Your Federal Tax Payment…has been rejected.”

You probably know this, but just in case, the Internal Revenue Service does not communicate via emails. These emails are attempted scams. What the IRS would like you to do is forward them to: phishing@irs.gov.

Or just ignore them.

CLARIFICATION: It turns out that the fellow who read through the August 15 Rutland Herald on the News Guy’s behalf didn’t read it closely enough. He was, kindly,  on his own time and on request from here, trying to find evidence for Brian Dubie’s claim that an article in the paper on that date supported Dubie’ss contention that Peter Shumlin has proposed releasing some prisoners before their terms expire. The article, Dubie said, reported that Shumlin “wanted to empty the prisons of 780 nonviolent offenders.”

As reported Wednesday (scroll down) the volunteer reader from Rutland could find neither hide nor hair of any such statement, though he did find a profile of Shumlin in that day’s paper.

Actually, way down in the third from the last paragraph of that profile, there was such a hide and/or hair.

At least so says – and there’s no reason to doubt her – Dubie spokesperson Kate Duffy, who emails that the story contains the following: “Emptying the prisons of non-violent offenders, he says, will cover the nearly $50 million price tag attached to his early childhood education plan.”

So Dubie was not making something out of nothing.

Only something out of little. The words “Emptying the prisons of non-violent offenders” were not Shumlin’s, but the reporter’s, There remains not a scintilla of evidence that Shumlin ever proposed releasing a single convict before his or her sentence has run its course, and Dubie’s continued efforts to argue that his opponent has made that proposal is pitiful at best.

(Though perhaps here it should be noted that Shumlin’s office still has not replied to an email asking how and why he misstated the numbers about the decline in the number of Vermont dairy farms).

NON-WARNING: If anyone is really concerned about either voter fraud or voter intimidation in Vermont next Tuesday: stop. By all discernible evidence and all rational analysis, neither will occur. That the subject is even under discussion seems to be the result of various confusions plus perhaps a little opportunism.

Here’s what happened: A small item in Monday’s Burlington Free Press reported that  members of Vermont’s “Tea Party” movement (also known as the Green Mountain Patriots) would hold a “poll watching training session” in Rutland Wednesday evening.\

The next day, Democratic Secretary of State candidate Jim Condos pounced, suggesting that the Tea Partiers were planning voter “intimidation,” and claiming that his Republican opponent, Jason Gibbs, was associated with the Tea Partiers “and other national groups that are planning to interfere with the voting process.”

In reply, Vermont Tea Party coordinator Jon Wallace of West Rutland, in a telephone interview Thursday, called Condos “shameless” and “disgusting,” as he insisted that “poll watching is part of the responsibility we have as citizens,” and that the Tea Party “wants every eligible voter to cast a ballot.”

Maybe everybody should calm down. First of all, in Vermont, the Tea Party represents a small if not quite insignificant minority which couldn’t intimidate more than a handful of voters if it wanted to. Second, there is no reason to suspect that it wants to. Condos may have been making much out of little as he saw an opportunity to link Gibbs with the Tea Partiers.

But considering the national context (to which he referred) Condos didn’t make much out of nothing. This is all part of the continuing debate between Republicans who suspect voter fraud and Democrats who claim Republicans are using “ballot security” concerns to try to scare minority (meaning mostly Democratic) voters away from the polls.

The Democrats have by far the better case. There is almost no voter fraud in America, and more than a tinge of racism in the allegations of it, most of which focus on minority polling places. Most of the few federal convictions over the last decade or so have been technical or individual (somebody voting in another town to vote for his brother-in-law; or maybe to vote against his brother-in-law). Large-scale voting fraud would make no sense. It would too complicated, too expensive, too easy to get caught. In fact, today’s only credible vote fraud accusations – fake absentee ballots cast in a Troy, N.Y., city election – make the case. There were – if the allegations are true – all of 36 fraudulent ballots cast.

(Even history’s most celebrated voter fraud, Richard J. Daley’s Democratic organization supposedly stealing votes for John F. Kennedy in Chicago in1960, may never have happened. It’s the stuff of legend, but there’s no convincing evidence demonstrating that any votes were stolen).

On the other hand, there is real evidence of attempted – and sometimes successful – intimidation of black and Hispanic voters at the polls. Armed private guards at polling places in Newark; leaflets falsely warning prospective voters that they’d be checked for unpaid traffic tickets or utility bills at the polls in Baltimore; false information about when and where to vote in Cleveland.

But that’s never happened in Vermont, where Democrats are neither as identifiable nor as easily intimidated.  Besides, Wallace is so open about what the Tea Partiers intend that he emailed a copy of the instruction sheet passed out to the potential Tea Party poll watchers. It essentially tells them to obey the law and not be disruptive.

Not that Wallace has any evidence of voter fraud in Vermont. He said he had heard from “college kids that some students they know boast that they vote several times.” That doesn’t really qualify as evidence.

Vigilance is always in order, and there’s nothing wrong with politically involved folks being on the lookout for intimidation or fraud. In this state, neither is likely, and whoever wins Tuesday will almost surely win fair and square.

UPDATE: The News Guy did not get to Woodstock yesterday for the evidentiary hearing in the case of Galloway versus Town of Hartford (scroll down to Wednesday’s post). Allen Gilbert of the Vermont chapter of the American Civil Liberties Union reported that Judge Katherine Hayes indicated she would rule in a week or two in journalist Anne Galloway’s request for official records relating to the seizure of a man in his own home last May. The ACLU is supporting Galloway,