Thursday, November 6, 2008

How Frank Kratovil Wins, Pt. IV

It's looking good. So far, with 7 of the counties completed their counting,  Frank Kratovil has extended his lead by another 1,000 or so votes. The current total includes Anne Arundel County and Wicomico County, which isn't done, bit its current totals are reflected in that number.

It is usually impossible for a Democrat to pick up votes during absentee counting - it almost always trends to the Republican. But a few things are at play here:

1) The majority of the requests came from the Eastern Shore - more than 15,000 of the 25,000 completed absentee votes came from the 8 Shore counties. That left only 9,400 from Harford, Anne Arundel and Baltimore counties. 

2) Rather than favor Republicans, absentee voter requests (and received ballots) were evenly split. There were 12,555 Dem completed ballots returned, as well as 12,266 Republican ballots and 3,743 other/unaffiliated. 

It's not over. Depending on which two jurisdictions are left, Andy Harris has a mathematical chance of overtaking Kratovil. But given what we know, and what is reasonable to predict, I'd much rather be Frank Kratovil right now.

The graph below is what I predicted would happen once we knew where all the absentee ballots were coming from. (Really, I did, you can ask Frank's guy in Salisbury). So far, it looks good, but there's a chance he's actually under performing his Eastern Shore numbers (which should be a +2,000 margin for him) and under performing on the western shore, where he should lose about 2,000 votes of whatever lead he gained in the Shore counting. Since this +1,000 number includes one western county, he may be doing just fine. 

But maybe not. As long as Frank does no worse than a net loss of 1,000 on the western shore and does at least +1,200 on the Eastern Shore, he should be OK. You've only got to win by one.


UPDATE: According to TwoSentz, Frank has won Anne Arundel County absentee votes by 16. Wow. I think I hear a large woman practicing vocal scales...

No comments: