Last night as I sat down to write this post I thought to myself "I haven't been this excited the night before an election since November 1, 2004."

The following day didn't turn out very well back then, but this is different.  In 2004 I was supporting a candidate out of necessity.  I was supporting a candidate I had not voted for in the primary. 

On November 2, 2004 I cast a defensive vote against George W. Bush.  Not a hopeful vote for John Kerry.

Although I can not vote tomorrow my hope is that from my post here in The Haight-Ashbury in San Francisco I have moved at least a few voters to vote for Barack Obama in Pennsylvania's primary.  Every one of them will be important as the stakes for today's vote are the highest yet.

In the event of a Clinton loss this campaign is over.  There will be nothing Clinton can do to take Obama's delegate lead, popular vote lead or perceived lead away from him.  The 2008 primary season will come to an end and the general election campaign will begin. 

2008 will officially be Obama vs. McCain.

In the event of a Clinton win...well that's more complicated.  So I'll break this down into different types of Clinton wins.  It is worth noting that delegate count is not going to come into play in any of the following scenarios. 

By most estimations, even if Clinton has a huge victory tomorrow night in The Keystone State she will pick up about 20 delegates.  As it stands now, before votes and delegates are counted in Pennsylvania, Obama has about a lead of somewhere between 140 and 160 delegates.  Plus any gain in that count in Pennsylvania for Clinton is likely to be completely erased by a Barack Obama victory in North Carolina on May 6^th.

So the only thing that really matters at this point in PA is the popular vote, and here are the four different potential types of victory for Clinton tonight:

Clinton Wins by 10%-Plus: This is the least likely scenario tonight, but if it does occur there is no chance we'll see Hillary drop out until at least Indiana and North Carolina are finished on May 6^th.  A victory of this size will have the pundits in the mainstream press worked up into a wild frenzy.  We will hear an annoyingly large number of sports comeback analogies and the phrase "buyer's remorse" will grate on your ears every one of the thousands of times it is uttered by the Clinton Campaign.

Chances of this happening: 5%

Clinton Wins by 5-9%: Clinton will be very happy if this scenario occurs.  Conversely it will be a very tough loss for Obama.  Even though Obama will have battled back from being 20-points behind a little more than a month ago, it will fall short of the expectations of now.  The momentum has been swinging Obama's way and it seems that he has closed the gap between himself and Sen. Clinton in recent days.  This scenario will swing the momentum back to Clinton as we go into the now very crucial May 6^th primaries.

Chances of this happening: 25%

Clinton Wins by 1-4%: This is the most likely scenario tonight.  Polling trends are suggesting roughly a 6% edge for Clinton right now, but most polling data has a lag of a day or two.  So if the trends continue, the slow crawl of Obama's comeback will prove to be too little too late and he will come close, but come up short.  The good news is that essentially this is a tie.  Isn't politics fun?

Even though this scenario will be viewed as a tie and Clinton will not immediately drop out the race, the calls for her exit will become louder.  The superdelegates will understand that Obama outperformed expectations and has basically as much support as Clinton in Pennsylvania.  Any fears they have about Obama's ability to carry a state like Pennsylvania against McCain will vanish and the slow trickle of superdelegates to Barack will become a steady stream.

Rep. Chaka Fattah of Pennsylvania, an Obama supporter, agrees with me.  On Hardball yesterday he said "As long as we hold it close here, under 5 points, I'll be very happy. 

Chances of this happening: 40%

Clinton Wins a Florida-Style "Too Close to Call" Election:  This might happen.  We might not go to bed until very late tonight.  Or at least I won't.  Some of you normal people will just choose to wait until Wednesday morning to find out what happens.

With the consistent tightening of this race and the lag in the polls, Obama may in fact be closer to being even with Sen. Clinton that we think at this point.  Also, it looks as if it will be mild and partly cloudy with no rain in Philadelphia today, good news for an area of the state where Obama needs high turnout.

If this scenario occurs it is a victory for Obama.  Just like a real Obama win, the perceived victory will all but doom Clinton's chances of catching Obama in any way, shape or form before the convention.  The calls for her exit might not come as quickly as they would after a genuine Obama victory, but they will be heard before May 6^th.

Chances of this happening: 5%

Conclusion:  As you may have noticed, the possibilities of the Clinton victory scenarios only add up to 75%.  That means that I am leaving Obama with a 25% chance of victory in Pennsylvania.

Although polling suggests otherwise, this is a very real possibility tonight.  There are a few quirky little details in this election which could give Barack Obama a small edge.

One is cell phones.  A major problem for pollsters these days is that it is very difficult to get in touch with people who do not have a landline.  Voters who choose to go completely wireless never get counted until Election Day.

So far in this campaign voters above the age of 45 have favored Clinton, while those under 45 are strongly in favor of Barack Obama.  Who is more likely to be wireless?  The younger voters are of course.  Since many of them have not been counted in the polling data yet, there is a good chance that a solid chunk of pro-Obama voters been muted until today's vote.

Potentially adding another small chunk to Obama's total are the 306,918 new registered Democrats in Pennsylvania this year, about 48% of whom are first time voters.  The other 52% were either Independents or members of another political party prior to this election.

The Rush Limbaugh-switch-parties-to-bloody-up-Barack fiasco seems to have died down since we came to learn that encouraging this behavior is not just immoral, but also illegal. 

Conventional wisdom suggests that if these new Democrats are not Republican trouble makers, the majority will be Obama voters.  If all 300,000-plus new Democrats break for Barack at a generous to Clinton 60%-40% it will equal roughly a 1.5% bump for Barack Obama.  At 70%-30% it is almost 3%.

Finally tonight, a quote that Barack Obama made yesterday leads me to believe that the Obama Campaign is seeing something we are not.  All of the campaigns do their own internal polling and Obama's internal numbers have been very solid throughout this primary season.  Barack Obama is a smart politician and he understands the expectations game as well as anyone.  So he would not raise the bar for himself in Pennsylvania unless he had a very good reason to believe he might be able to clear it.

In a radio interview today Obama said "I'm not predicting a win.  I'm predicting it's gonna be close and that...we are gonna do a lot better than people expected."

It is difficult to imagine him saying this unless he believed it. 

My final prediction for tonight's final score is a tough one to make.  So I will give each candidate a range, rather than an exact number:

Clinton has a range of 49% - 53%.

Obama has a range of 47% - 51%.

Happy poll watching everyone!