Larry Sabato is a very smart man, having come closer than anyone I’ve seen in predicting the outcomes of recent elections. However I’m unclear when he made the switch from prognosticator to campaign adviser, offering this unsolicited advice to John Edwards:

Edwards can play a critical role, though. He shows every sign of continuing his campaign, which is certainly his right. But he may have the ability to determine the Democratic nomination by ending what will very likely be a losing campaign and endorsing either Obama or Clinton. (One assumes it would be Obama, given the antipathy that exists between Clinton and Edwards, but with politicians you never know anything for sure until they do it.) Usually candidates never withdraw until they run out of money or energy. Edwards may have sufficient quantities of both to last through February 5th, but he may discard his ability to play kingmaker if he waits that long. Maybe he doesn’t mind: Most presidential candidates really don’t care much who is President if it’s not them.

I don’t know about you, but I find that fairly insulting, both to the integrity of John Edwards and to my intelligence.

Actually writing publicly what I confided privately a month ago is not something I take lightly, but Sabato and the punditocracy is right to a point. Barring something catastrophic, John Edwards will not be the Democratic Parties nominee because everything hinged on Iowa.

Before Iowa, in the face of polling data that was less than inspiring, I knew that Edwards could not win without a 1st or very narrow 2nd place in Iowa. He needed to tie Obama not Clinton there and take that momentum into New Hampshire. Barring an outright win in Iowa where a 2nd place in NH would have been acceptable, he probably would have got the Culinary Union endorsement in Nevada, won NV, and improved his standing in South Carolina from respectable to inevitable.

Right now, Edwards doesn’t even look viable in South Carolina. A good showing in South Carolina was the ticket to being competitive on Tsunami Tuesday, a locomotive that would steadily pick up steam and get enough free media attention to counter the gross domestic product of Micronesia that the other candidates will pour into the major US media markets in the months to come. All of that hinged on Iowa, the key domino in what was always an unlikely Quixotic journey.

We all knew that, and if we didn’t do the math, knowing that Edwards practically lived in Iowa the last three years should have told us something pivotal would happen on caucus night. I know that’s the reason I’ve learned more about the absurdity of the political dynamics of that little state than I ever wanted to.

Under the fold, I’ll let you in on some of the math that bears this out, and why Larry Sabato’s cynical statement is all wet. As long as he keeps going, Edwards can write his own ticket.

Every day Edwards stays in the race messes with Larry Sabato’s crystal ball (which is clearer than most I’ve seen) and makes it more difficult to do what he does — predict elections with uncanny accuracy.

Nobody will come out of Tsunami Tuesday with 2000 delegates. There’s 1698 up for grabs that day and it’s doubtful even in a 2-person race that any one candidate gets over 1000 of them since these things aren’t “winner take all” — and you need 2025 to win.

From wiki:

A candidate needs a simple majority of the combined delegate and superdelegate votes to secure the nomination. Democratic delegates from state caucuses and primaries number 3,248. This means that the total number of votes is 4,049. The total number of delegate votes needed to win the nomination is 2,025.[2] Superdelegates account for approximately one fifth (19.6%) of all votes at the convention. Delegates chosen in the Democratic caucuses and primaries account for approximately four fifths (80.4%) of the Democratic convention delegates.

Looking at the numbers, Sabato must be thinking about momentum, not actual delegate count.

As long as Hillary keeps her superdelegates in line (which was what all Bill’s phone-calling Monday was about), she’s dominating the delegate count in the early going. When you talk about the Clinton machine, being able to call up favors is no small part of their power.

To date, here’s were they stand with 277 days until the Convention, including the Superdelegates:

Now there are three ways to predict the future to see if Edwards is more relevant now when momentum matters, or later when actual delegates count.

There’s the fast, easy way by simply dividing all the outstanding delegates by the most current National Poll percentages. Or you could match the most recent state polls with the delegates presumed to be awarded if all things stay equal. Finally you can do a study of each state one at a time, factor in trends to the recent polling data per state, analyze some demographic data and try and narrow the unknown. Right now, I’m going to do the cheap analysis. In a later post I’ll get into even more detail and see where we are.

The simple method naturally leaves a lot in the “unknowable” column, doesn’t factor in “momentum” and/or media hype, and neglects the voters’ learning curve as their particular primary nears. We’re talking about at least 15% of the voters out there who have yet to even think about who they prefer.

No rocket science is involved in plugging in Hillary as the favorite doing it the easy way. She’s ahead nationally, therefore there’s no reason to actually vote since the pollsters are telling us she’s still going to win. Right?

Nope, it’s actually quite fascinating how it would work out even doing it the quick and cheap way.

The Real Clear Politics average of Rasmussen, Pew and Gallup have Clinton up by 7 points: Clinton 37.7%, Obama 30.7% and Edwards 16.3, leaving 15.3% undecided. If the perception he’s d.o.a. leaves Edwards with none of the undecided vote which break between the frontrunners along the same lines as their relative support nationally, of the 3761 still left to be decided before the convention, combined with their current delegate count, we have the following highly intriguing totals:

Wiki tells us there are 314 Superdelegates, which means that if Hillary gets all the remaining 91, she’d still be short of the magic number 2025 needed to win, only being able to count on the old guard to the tune of 1826 delegates walking into the convention.

But, and here’s where Edwards stakes out the kingmaker status, if Edwards waits until the convention, drawing the same mediocre support he got in New Hampshire, his 613 delegates added to Obama winning his presumed 1413 delegates based on his current 30.7% support equals exactly 2026 — a cushion of one delegate over what is needed to win.

Yeah, I was surprised it worked out that way too.

Let me repeat that. If Edwards stays in until the convention then pledges all his delegates to Obama, Obama wins by one vote.

Likewise, Hillary needs Edwards to commit to her or drop out so some of his supporters go to her camp, or she cannot win the nomination.

Unconvinced? Good. What’s left is a bit more challenging, gives less deference to the national polls and is more of a State-by-State analysis. In the medium (undergraduate) analysis, you plug in all the numbers from the latest available state polls, match that on your spreadsheet with the delegates selected by each state from those polls, and divide the undecided vote per their relative support.

The doctoral level analysis requires you do the painstaking study of each state poll and weigh the undecided voters as breaking consistently with the preferences that state is currently trending and factor in weighted demographic data that will include the women’s vote preferring Hillary countered by the African-American preference for Obama.

When I finish that, you’ll have it.