Let's say Obama loses OH and FL. Let's also assume that he loses Bush states where Poblano has identified that he has a less than 40% at winning.
Finally us assume that he wins all the Kerry states except where Poblano has identified that McCain has a greater than 40% chance at winning (MI and NH). The remaining swing states are: MI, IA, NV, CO, NM and NH.
There are still 5 ways he could win:
Win: MI, IA, NV, CO, NM, Lose: NH (274 EV)
Win: MI, IA, NV, CO, NM, NH, Lose: None (278 EV)
Win: MI, CO, NV, NM, NH, Lose: IA (271 EV)
Win: MI, IA, NV, CO, NH, Lose: NM (273 EV)
Win: MI, IA, CO, NM, NH, Lose: NV (273 EV)
Horay! 5 ways!
Unfortunately, there are 6 states here, and a total of 2^6 = 64 combinations, and I have just listed all scenarios that produce an outright win for Obama. There are also 3 scenarios which produce an outright tie (win all of the above but (NH, NM) (NH, NV) (CO).
That leaves 56 combinations. 5 ways to win, 3 ways to tie, 56 ways to lose.
In short, Obama's chances aren't that great without Ohio and Florida. But wait-- aren't Ohio and Florida the old swing state strategy that Obama's candidacy was supposed to try and avoid? Doesn't his performance in the swing region of southeast Ohio stink? Isn't his performance among Hispanics, older voters and Jewish voters that are key in Florida subpar? Aren't those states the places where Hillary Clinton was polling better?
All well.
Missed opportunities.
So if Obama does the Western state strategy without OH and FL, he has, among states where his probabilities are between 40% and 60%, 56/64 losing combinations and only 5/64 winning combinations. But putting his eggs in the OH/FL basket isn't what his strategy wants to be either.
What else can he do? One thing he can try is the landslide strategy: Win big. He can win so big, that the whole concept of swing states doesn't matter-- now he brings states like Virginia (13), Indiana (11), Missouri (11), Montana (3), and North Dakota (3) into play, as well as Nebraska's CD 1 and 2.
Yet this plan has problems too: it raises the bar higher. He is currently polling nationally at only 46-46, compared to McCain. He has to not only bring in Clinton's voters but get a lot of swing voters too-- and that means emphasizing his appeal to the center, something he cannot do while in a Democratic primary, or while in a Democratic primary mentality.
In short, Obama has three paths to victory:
(1) The Map-Changing Strategy by giving up on the traditional battlegrounds of Ohio and Florida, or "forcing McCain to spend money" there but not really trying to win them. He gives up these states to "change the map." The problem with this is, there's only 5 ways to win, and 56 ways to lose. If you're an Obama supporter and counting on this strategy, you're fucking batsh*t insane.
(2) The Swing-State Strategy where he tries to win Ohio and Florida along with the other swing states. The problem with this is, Hillary Clinton was a better suited candidate to do this, he is weak among key demographics in both states. This is a bad idea too, as everyone is aware of.
(3) The Win Big Strategy where he tries to get around the problem by winning so big, he won't have to count which states he wins, and he'll put states like Virginia, Missouri and Indiana into play. The problem with this is, he's not likely to actually win that big. Bad idea.
All of these strategies are bad ideas. None of them are well thought out.
What's Obama's real strategy going to be?
(1) He can't ignore Ohio and Florida.
(2) He can't focus only on Ohio and Florida.
(3) He can't assume he'll win massive because the polling is not backing him up on it.
His strategy will probably be a combination of the three. Contest as many states as possible, and win as big as possible. But he'll have to seriously try to win some of the following:
OH
FL
MO
VA
He can't afford to give up on any of them.
Update: I have changed the wording to replace 'permuations' with 'combinations'.
Update 2: I am NOT trying to say Obama cannot win OH, FL, MO or VA. I am saying that he must make it a goal to win at least 1 of those, or his odds are long. This is in direct contradiction to some conventional wisdom I've seen here which says he can take a loss in OH and VA but win elsewhere.