In the last week, John McCain has pulled most of his resources out of the get-out-the-vote ground game and poured it all into television ads:

The decision to finance a final advertising push is forcing McCain to curtail spending on Election Day ground forces to help usher his supporters to the polls, according to Republican consultants familiar with McCain’s strategy.

The vaunted, 72-hour plan that President Bush used to mobilize voters in 2000 and 2004 has been scaled back for McCain. He has spent half as much as Obama on staffing and has opened far fewer field offices. This week, a number of veteran GOP operatives who orchestrate door-to-door efforts to get voters to the polls were told they should not expect to receive plane tickets, rental cars or hotel rooms from the campaign.

But television ads and robocalls don’t get people to the polls.  Cars get people to the polls.  Volunteers offering to help people get to the polls gets people to the polls.

McCain’s move isn’t completely crazy.  As Nate Silver at 538 has pointed out several times, McCain needs the polls in every state to tighten about five points to even have a shot at picking off enough battlegrounds to win the electoral college.  Television ads are more likely than get-out-the-vote efforts to have this kind of nationwide effect.  The problem is, McCain’s television ads aren’t moving the numbers at all, and he’s still down about 5 and 11 points in every national poll and tracker.  Which means that McCain is working from at least a 5-point deficit, AND he’s going to be turning out less voters.

Think of it like this.  There’s 100 voters in a room; 53 supporting Barack Obama, 47 supporting John McCain.  For McCain to win a vote from the people in that room, he has to either convince 3 or more Obama supporters to switch sides, or he has to make sure that all 47 of his people actually vote, while maybe 6 or 7 Obama supporters flake out and stay home.  McCain’s problem is two-fold: 1) In the last few weeks, he hasn’t been able to convince any of those 53 Obama supporters to switch sides; and 2) Because he put all his resources into the side-switching tactics, he now has very few resources left to get the 47 people who are on his side to actually get out and vote.  So in the end, Obama is going to use his superior ground game resources to turn out something like 50 of his supporters (a 95% turnout rate), while McCain is only going to get 42 of his supporters to show up and vote (90% turnout).  That’s an 8-point loss in a room McCain only trailed in the polls by 5 in.

I’m pretty confident that with increased African-American turnout that has amounted to a self-propelled get-out-the-vote advantage, plus the enormous ground game advantage Obama has (more volunteers, more offices, much more voter contact), that things state-by-state are going to look a lot like the hypothetical of that room.

If you need evidence of how much stronger Obama’s ground efforts are, Ben Smith is posting a lot of good anecdotal stories today from reader e-mails.  538′s “Road to 270″ series also has a lot of really good on-the-ground evidence from correspondent Sean Quinn.  And Josh Marshall posted this video yesterday of Politico’s Roger Simon on MSNBC, discussing the sheer magnitude of Obama’s GOTV advantage:

Electoral vote tally: Barack Obama 375, John McCain 163.

Just a few notes:

– Every poll confirms that Pennsylvania is a done deal. I know McCain sees it as his last stand, and he definitely needs to win it to have any real chance at the presidency. But if it’s such a swing state, why hasn’t Obama been there in a week?

– Ohio and Florida seem to be the two battlegrounds getting the most attention from both campaigns in the final days. Ohio is one of the states hit hardest by the economic downturn, especially in the housing market, and I think that puts it over the top for Obama. The theme there, even from white voters skeptical of a black candidate, is “Won’t Get Fooled Again.” And if Ohio falls, Indiana will also get swept in, by virtue of being sandwiched between the Buckeye State and Obama’s Illinois (where he’ll win by 20% or more). Florida is all ground game, and McCain is getting trounced by Obama in that respect. Kerry lost Florida by 5%; Obama will make that up in increased African-American turnout alone. Obama’s superior get out the vote effort will push it over the top.

– Early voting has locked up Nevada and North Carolina. It’ll also put Obama within three points in Georgia, though not quite close enough to win there.

– McCain’s team fumbled the ball badly in Virginia. Not only did they build no ground game, but they spent the last few weeks insulting the northern half of the state. It’ll cost them 13 electoral votes.

– Missouri in the Obama column is the prediction I’m most worried about.  I think the vote there might look like it did on Super Tuesday, when the state was originally called for Hillary Clinton but was eeked out at the end by Obama.

– If McCain is going to pull off a surprise anywhere, it’ll be the midwest states, Ohio and Indiana. Under my map, that still leaves him 76 electoral votes short.

– If you want to know when on election night you can start celebrating, here’s a guide. If Obama wins just two of the following four states, it’s over: Pennsylvania, Ohio, Virginia, or Florida. If Obama wins just one, he can still be saved by holding serve in the west, winning Colorado, New Mexico, and Nevada, or by winning just one of those states plus North Carolina. If Obama somehow loses all four– Pennsylvania, Ohio, Virginia, and Florida– hold onto your hat. He would then need the three Western states– Colorado, New Mexico, and Nevada– plus North Carolina (which could easily happen), or the Western states plus Indiana and Missouri (a scenario that is unlikely if he’s already lost Midwestern voters in Pennsylvania and Ohio).

A Decision Analyst poll of Texas voters released today shows a 14-point lead for Barack Obama over Hillary Clinton.  It’s an outlier, to be sure, but I’d wager it’ll come closer to the final result than any poll released thus far.  At any rate, no matter which poll you look at– and keeping in mind that we’re still 11 days away from the primary– the trendlines in Texas don’t look good for Clinton:

But I don’t think it will come to any of that.  I think we’ll see in Texas and Ohio the same thing we’ve been seeing for the past two weeks.  Wherever the poll numbers are a week before the primary, they’ll swing a good 15 to 20 points in Obama’s direction by election day, thanks to unaccounted for surges in African-American, Independent, and first-time voters.  Ohio may be close, but I’ll be surprised if Obama doesn’t win Texas by double digits.

(Full disclosure: Truthfully, the entire purpose of this post is so I could write the headline “Houston, she has a problem.”)

UPDATE: Commenter warner is all over the links, and provides a great one to a Houston Chronicle article that shows early voter turnout in Texas up 2 to 3 times what it was in 2004 in Clinton-friendly areas such as El Paso and Corpus Christi, but up between 5 and 10 times the 2004 numbers in Obama-friendly areas such as Dallas and Austin.  Their prediction based on early voter turnout: “The same factors that carried the last 10 states for Obama will carry Texas.”

A quick look at exit polls from Virginia bears this out.  While beating Clinton by 29 points in the overall vote, Obama won Independents by 39 points, voters under 45 years of age by 43 points, and African-American voters by 80 points.  In the past, Clinton has been able to withstand these types of disparities by holding her own among Latinos, older voters, and registered Democrats– often by enough of a margin to win, as in California.  But in Virginia, as we saw in other states over the past weekend, Obama managed to win Latinos by 8 points, voters 65 or older by 10 points, and registered Democrats by 24 points.  And this is just the exit polls from the state he won by 29 points.  No exit polls were released for D.C. by any major news organization, but imagine what they would look like.

Fortunately, we here at TrainWreck Politics conducted our own exit poll yesterday of voters in all three Potomac primary contests.  It’s a fairly complex analysis and may be difficult to follow for some readers, but bear with me.  I believe it gives as much information, if not more, than any exit poll conducted by the larger news organizations, and will give you a better idea of exactly what happened yesterday.

As you can see, this chart illustrates exactly how voters lined up in each state and in each key demographic.  Not only that, but many demographics that the major organizations did not cover can be broken down and analyzed here.  How did Obama do among black voters in D.C. yesterday?  Obama won.  Which way did women voters in Maryland lean?  Obama won.  How did Clinton fare among older voters overall?  Obama won.  What was the preference of gay voters age 35-44 who work full-time but spend at least an hour a day shopping online at work?  Obama won.  Among Latino voters from Bethesda who believe the economy is on the wrong track, which genre of music did they say provides them the most comfort?  Obama won.  How did Clinton do among male voters age 55 or older who have been institutionalized in the past for believing that they were in fact the real Hillary Clinton?  Obama won.

What’s clear from yesterday’s results is that Barack Obama has built a historic coalition of new voters that is now being shored up by a Democratic base that once belonged to Hillary Clinton.  In a general election– as we have seen thus far in the primaries– this combination of old and new, past and future, that reaches deep into every imaginable demographic will be difficult if not impossible for even a moderate like John McCain to match.  And unless Clinton can find some way not just to stop, but also to dramatically reverse Obama’s momentum in the next three weeks, both parties may very soon have their nominee.