Related to this post:

1) Even with the pro-Rossi/pro-Republican tilt of later ballots, Gregoire should still will win the primary, and her lead should open up a bit again thanks to the estimated 80,000 ballots left to count in King County. But, her margin should remain notably less than it was on election night.

2) Related to such estimates, one real problem that needs more discussion in public circles is the wretched "system" we have of counties "estimating" the remaining ballots they have yet to count.

Anyone who has dealt with large counties in tracking post-Election Day returns - especially for a close race - knows exactly what I'm talking about. The "estimates" the counties give out are either only a count of the ballots already processed and ready for counting (sometimes a very different number than the actual number they have physically received) or a rough estimate of the number of ballots sitting in bins and trays that have come in via mail or from collection centers/poll locations. Sometimes, the "estimate" is a combination of those two methods.

Either way, they're a terrible gauge for the number of ballots left to count. If turnout starts out modest then spikes at the end - like we saw with this primary - then county estimates end up being way off. The number of ballots publicly estimated as being left to count by King County has actually grown this week, even as tens of thousands of ballots have been counted.

I've dealt with the same issue in Snohomish County in past years as I've worked with the county assessing what is left to count in races on which I was involved. There is simply now way of accurately assessing how many ballots are in the mail system or how many ballots might end up being deposited at drop-off locations. And trust me, when you're concerned about a close race, the last thing you usually want to hear is "oh, we were wrong, there's actually X thousand more ballots than we expected left to count based on what came in via the mail system."

Thus, it is highly probable that on Election Night in November, we will have no earthly idea how many ballots are actually still out there to be counted...in part because the counties themselves have no real way of knowing - let alone publicly reporting accurate figures. And that doesn't even touch on the fact that large counties are going to be swamped by the demands of counting the raw number of ballots cast if turnout is near the 80% number currently expected.

Meaning that on Election Night, in our splendid system, we really have no way of knowing what to expect if a race is close...perhaps for days on end.

Sound familiar?