In June of this year I became increasingly frustrated by the political discussion in blogsphere.  It was all about the National Polls, and how Clinton could not be beaten.  This disappointed me - it was really no different than what you would read in the Washington Post or Time.  It also angered me, because in a way the national polling was being used as a weapon against dissent within the Democratic Party.  

But what really made me mad was that is completely wrong.  I have been involved in Democratic Politics for nearly 30 years, and the one thing I do know is that National Polls are meaningless before New Hampshire.  And yet these polls were being used to create the impression that resistance to Clinton was futile.

The point of the diary is not to say I told you so.  OK, so it isn't the ENTIRE point of this diary.  I am a strong believer that history can provide guidance about how primaries affect each other.  

Here is what history teaches about the current state of the race:  it is far from over.  After the diary about Iowa and New Hampshire, I wrote another diary about how New Hampshire changed National Polling.  On average, it found that when a front-runner loses New Hampshire, there is a 33-point swing in National Polling.  Since that time I have done more detailed analysis that shows when you include GOP contests since 1980, the average is 34.  

But the data also shows something else: This bounce is reversible.  In at least three instances (1984, 1992 and 1996) the bounce from New Hampshire substantially receded about 3 weeks after New Hampshire.  The reasons for this are complex, and I will write more about this tomorrow.  

But here is the key point I want to make: This race will not be over when Obama wins NH tomorrow.  The people who are writing that are the same people who told you this summer that Clinton had the race sewn up.  In short, this is another example of a press that does NOT UNDERSTAND THE PRIMARY PROCESS.

Back to my diary on Iowa and New Hampshire. Before I review the evidence, I would note that Mark Penn, the Clinton chief strategist, argued that there would be little or no bounce from Iowa to New Hampshire. He was wrong, and I was right.

I wrote the diary on Iowa and New Hampshire in June, entitled how Iowa effects New Hampshire.  the key part of that diary is this table, which summarizes the bounce from Iowa to New Hampshire.

How is my prediction fairing thus far?  The table below summarizes the results (I am using the realclearpolitics average for the post Iowa number, and the average of three polls taken right before Iowa - Zogby, ARG and Suffolk).

The net swing between Clinton and Obama right now is 16.46 (my model predicted a swing of 19).  Pretty Good.  BTW, in the June diary I predicted the following if Obama won, Edwards finished second, and Clinton finished third:
Obama 35, Clinton 30.7, Edwards 20.2.
The current realclearpolitics average is Obama 36.9, Clinton 29.1, Edwards 18.8

As I mentioned above, I have conducted additional research how National Polling is effected by Iowa.  The updated table, which I will discuss tomorrow, is below:

This table summarizes how various finishes impact National Polling

This table summarizes the impact depending on whether the front-runner won or lost New Hampshire

As I said, I will walk through the evidence tomorrow, but I find it interesting that the addition of GOP data did almost nothing to my model.