Voter models show challenge for Reid if he loses independents by big margin, for Angle if she hemorrhages base

Sunday, Oct. 24, 2010 | 3:46 p.m.

I decided to have some fun with voter modeling in the Senate race after the first week of early voting, which probably represents about a quarter (or slightly more) of the entire turnout.

The actual turnout of just over 160,000 early voters has this breakdown:

Democrats: 42.5 percent

Republicans: 42.5 percent

Others: 15 percent

So let’s predict what might have happened that first week, based on certain assumptions. I should note that these are based only on early votes – I understand the Democrats have a slight edge in absentee ballots returned and we need to assume these are hardcore partisans. That also brings the total early vote to about a third of the total expected after Election Day.

So remember, you probably can add a fraction toward Harry Reid in the analysis below.

Also, I work under the assumption that Reid loses little off his base (although I took 15 percent) while Sharron Angle could lose more. If Reid loses more than Angle off his base, he will almost surely lose. But less likely Democrats will vote for Angle than Republicans for Reid. That is generally agreed upon.

Suppose both candidates held 85 percent of their bases through that first week, and each base voted 2 percent for other candidates or “none of the above.” Then, suppose Sharron Angle wins independents by 5 percentage points. The results:

Reid: 48.2 percent

Angle: 48.9 percent

Suppose Angle lost 2 percent off her base – not to Reid but to “None of the above” and this happens:

Reid: 48.2 percent

Angle: 48.1 percent

Now suppose Reid lost independents by 10 percentage points during the first week, and both hold 85 percent of their bases:

Reid: 47.9 percent

Angle: 49.4 percent

Suppose Angle lost 2 percent off her base – not to Reid but to “None of the above” and this happens with her 10-point advantage with independents-- she still wins:

Reid: 47.9 percent

Angle: 48.6 percent

If Reid lost independents by 10 points in early voting, he would need (assuming he holds 85 percent of his base, which seems conservative), Angle to retain 82 percent of her base or less to be ahead of her.

Bottom line:

If these percentages hold through Election Day – and no telling if they will until we get more data this week -- this race is likely to be agonizingly close unless one of two things happen:

1. Reid gets crushed by independents (even the GOP-leaning Mason-Dixon had him down 8 in its last survey)

2. Angle retains less than 80 percent of her base (hard to believe but not impossible)