The razor-thin margins in the Florida U.S. Senate election this week brought me back 18 years.

It was eight days before the presidential election of 2000, the one that would eventually take George W. Bush to the White House by virtue of a 537-vote cushion in Florida, a state that had cast nearly 6 million ballots.

One of the more prophetic statements made before that election came from Washington-based Democratic lawyer-lobbyist Richard Goodstein.

PHOTOS: Election Day in Palm Beach County.

Goodstein predicted that Florida might be the decisive state in that election, and he said that the vote in Palm Beach County may very well tell the story of the state, and ultimately the national election.

Palm Beach County would certainly give the Democrats a vote cushion. But for each party, declaring “victory” in the county came down to the size of that cushion.

Democrats shot for 65 percent. Republicans shot for 40 percent. You can see the problem with this: If you add 65 and 40, you get 105 percent — which is impossible.

It’s that overlapping 5 percent of aspirations that determines the battle for Palm Beach County, and as Goodstein predicted that year, maybe a lot more.

“The difference between 62 percent and 66 percent, that’s the whole election right there,” Goodstein predicted.

As it worked out, Gore ended up with 62.3 percent of the Palm Beach County vote. If he won a fraction of a percent higher in the county, he would have been the 43rd president of the United States. (For example, those couple of thousand votes cast by mistake for Pat Buchanan in Palm Beach County would have been enough.)

I was thinking about the Palm Beach County vote as a key to winning the state this week as the votes started coming in for the U.S. Senate race between Rick Scott and Bill Nelson.

Nelson was first elected to the U.S. Senate during that presidential election in 2000. Nelson beat Republican Congressman Bill McCollum that year, and in Palm Beach County, he held McCollum to just 36.3 percent of the vote.

When Nelson was re-elected in 2006, it was a wave election for the Democrats and he easily beat Republican candidate Katherine Harris, the former Florida secretary of state.

Nelson got 72.5 percent of the vote in Palm Beach County that year, easily shattering the 65 percent threshold.

Six years later in 2012, while running against Republican Congressman Connie Mack IV, Nelson won again. That time he got 63.7 percent of the Palm Beach County vote, closer to the 65 percent goal than the 60 percent mark of peril.

What happened Tuesday?

By the end of the night, Nelson got only 58.26 percent of the Palm Beach County vote.

He had slipped below the 60-percent floor, while Scott got 41.74 percent of the Palm Beach County vote — nearly 2 percent higher than the Republican high-water mark of success.

The Republicans won the turnout competition in Palm Beach County, and that made all the difference.

To show how important that was in the statewide result, it helps to look at the raw numbers.

Nelson’s tally after Tuesday night in Palm Beach County was 330,294 votes out of a total of 566,902 ballots cast.

If Palm Beach County had been performing optimally for Democrats, and Nelson would have gotten 65 percent of the county vote, that would calculates to be 368,486 votes.

Now, if you subtract the votes Nelson actually got in Palm Beach County (330,294) from the votes he needed to get at 65 percent of the total (368,486), you get the size of the under-performing Democrat vote total in Palm Beach County.

And that come to 38,192 votes.

Those are enough votes to change the outcome of the election, with Scott’s statewide margin of victory so far at about 34,500 votes.

Sure, you can come up with plenty of other variables to predict an outcome. But if you're looking for a reliable variable, the fate of Florida as a blue or a red state may be reduced to examining the voting percentages in this county.