Election day predictions for Georgia . . .

Regular readers know that I’m a poll watcher.

And there has been plenty of polling of the big Georgia races over the last few weeks. The weight of that polling has suggested — consistently suggested — that Republicans will have a pretty good day. Looking at the polls, one would assume that Governor Nathan Deal (R) will win a narrow victory over Jason Carter (D) and that David Perdue (R) will have an even narrower win over Michelle Nunn (D) in the race for the open U.S. Senate seat.

In theory, the sheer weight of the polling should even out discrepancies between the polls. In other words, if we consider the numbers from all the polls together, the margin for error should be exceedingly small.

But if you’ve been following the analysis and comments at Peach Pundit (I’m an occasional contributor), you already know that there have been serious methodological questions raised about almost every poll. For example, most analysts who are especially well-versed in Georgia demographics and politics are expecting a black turnout of approximately 28 percent, but many of the polls pegged that number at 26 percent or lower.

In early polling, with perhaps as many as 1/3rd of total ballots already cast, the black turnout is over 33 percent.

It also seems that pollsters’ likely voter screens have weeded out too many voters, and some polls have likely also understated the percentage of women in today’s electorate.

In other words, it looks to me like there has been a pretty broad, but quite small, bias toward the Republicans in Georgia’s big races.

Also, before I share my final guesses, I should add that we really have no idea how the particular dynamics of regionality (will Perdue get out those Kingston voters?) and of family legacy (Carter is Jimmy Carter’s grandson, Nunn is Sam Nunn’s daughter, Perdue is Sonny Perdue’s cousin) will play out.

I should also add that I’m assuming that Nunn is fundamentally a stronger candidate than Carter and that the incumbent Deal is a stronger candidate than Perdue.

I think both of these major races will see the winner get just 50 percent — thus avoiding the January runoff that many are predicting. Here’s my guess:

Nunn 50.8, Perdue 47.8
Deal 50.5, Carter 48.0

Given the power of incumbency and the likelihood that many voters will ignore the down ballot statewide contests, I expect all of those to go Republican, although perhaps by very narrow margins.