Down East 2013 ©
Elections are the ultimate performance evaluation. Next week, the merit of years of strategy and months of hard campaigning from each of eleven gubernatorial candidates will be decided in one day of voting.
What do you think will happen?
Or, perhaps a more important question: Do you think you’re smarter than Rep. Sean Flaherty?
That’s right, the June 2010 Maine Election Prediction Pool is now open for business . The Honorable Mr. Flaherty eked out a win  last November, but this year the field is wide open and anything can happen. Enter your predictions here  for both the gubernatorial primaries and the five ballot questions. The top three winners will each receive a piece of Maine campaign memorabilia.
Feel free to use all of your wisdom and political acumen in making your predictions, but don’t expect to get any help from the polls. Or perhaps that should be the poll. Only one independent survey has been released of the Maine governor’s race, the Pan Atlantic SMS omnibus .
The Pan Atlantic poll, which made news yesterday  and this morning , shows Mitchell in the lead for the Democrats and Otten ahead among the Republicans, but the differences between all the candidates are small and there are a large number of undecided on both sides. Additionally, the small sample size, number of candidates, likely turnout and other factors make the results less than reliable.
Here are the numbers for the Democratic candidates:
And here are the Republicans:
What the poll also can’t fully measure is which of these candidates has done the best job identifying their supporters and has the best infrastructure in place to get them to the polls, a very important part of winning a low-turnout primary.
One interesting result from this survey is the position of Paul LePage. If these numbers are close to the truth, he has amassed a significant supporter base despite having raised very little money compared to the other candidates and despite being absent from the airwaves until last week. It also hints at a strong grassroots operation.
The individual numbers for each of these candidates probably aren’t too far off the mark. The poll has a relatively high margin of error (5.7 percent at a 95 percent confidence interval) but individual results within the poll actually have a lower margin. If I remember my statistics correctly, based on the sample size, LePage’s total support is likely to be within 3.4 percentage points of the amount reported (10.3 percent) 95 times out of 100. Those margins quickly increase once you begin comparing candidates, however, and a tight race means most horse race comparisons aren't very useful.
The major feature of this poll is the huge number of undecided voters. The truth about that category is that a large number of them won’t vote and many of the rest will vote in the same proportions as their peers. There are only a few things that could swing a large group of undecideds into one column or another, mainly a “June surprise” (some kind of public event in the next few days that elevates or damages a candidate), strong field operations, or some undetected current of public opinion that will rise to the surface on election day.
Historically, studies show that undecideds have often ended up siding with the challenger. With no incumbent in the race, perhaps this will mean an election day surge for the candidates least identified with the political establishment. On the Democratic side, this would be good news for Rosa Scarcelli.
All of this is just reading tea leaves. Real answers will only come on Tuesday. Until then, make your predictions here .
One last addendum: Peter Mills announced today that his campaign will be releasing the results of polling that they commissioned from the Maine Center for Public Opinion (a company that didn’t exist until a few months ago and has released no public polls) with a sample size of 900 [Correction: I first wrote 50,000 here, when I misinterpreted a Mills campaign email and erroniously believed that they were releasing voter ID info]. Their results could be informative, but should be taken with a large grain of salt.