Here’s Obama’s odds of winning the states necessary to become the next President:

State - Percentage - Electoral Votes

HI - 100% - 4
DC - 100% - 3
VT - 100% - 3
IL - 99% - 21
CT - 97% - 7
MD - 94% - 10
CA - 92% - 55
NY - 91% - 31
DE - 91% - 3
WA - 90% - 11
______________
Subtotal 148 electoral votes

NJ - 89% - 15
IA - 86% - 7
RI - 86% - 4
MN - 83% - 10
MA - 82% -12
OR - 82% - 7
______________
Subtotal 203 electoral votes

ME - 79% - 4
______________
Subtotal 207 electoral votes

WI - 62% - 10
NV - 60% - 5
______________
Subtotal 222 electoral votes

MI - 56% - 17
CO - 56% - 9
PA - 55% - 21
NM - 55% - 5
______________
Subtotal 274 electoral votes

These 22 states plus DC push Obama over the 270 votes needed to win in the Electoral College. But by these metrics, he has to campaign hard in the bottom six: WI, NV, MI, CO, PA, and NM. Those are the most vital swing states.

And does he near 50% in any others? Where he has at least a 1 in 3 chance?

OH - 41% - 20
IN - 37% - 11
ND - 32% - 3

So 25 states and DC are likely to be the most he can win short of a major McCain meltdown, and that would yield 308 electoral votes.

Had Clinton been the nominee, some in that group would be improbable, but there’s also a few states where she’d do much better. Consider these:

AR - 71% - 6
WV - 67% - 5
FL - 63% - 27
MO - 40% - 11

Don’t get excited, this is not meant to advance Clinton’s electability claims, as she garners 261 electoral votes where her odds are 50% or better, compared to Obama’s 274. But notice the information we can glean from those final four:

1) FL and AR are the only Southern states where either Democrat has a 1 in 3 chance of winning.
2) MO is the ultimate bellwether state. It’s only picked the losing presidential candidate ONCE since 1900.

It would be incredibly risky for Obama to confine his campaign to the 23 places that barely give him the win. At the least, he has to add OH, IN and ND. But historically, IN and ND have been pretty untouchable by Dems while AR, WV, FL and MO have gone both ways, so adding those four makes sense, too. That provides a 30 state campaign (if DC counts as a state) , with most of the campaigning focus to be directed at the bottom 13.

So let’s use MO as the benchmark, because of its historical record of success at selecting presidents, and because Obama won it despite the greater odds for Clinton. What if we add every state with odds higher than MO? Here’s what we’d get:

NH - 27% - 4
AK - 26% - 3
VA - 25% - 13
NE - 24% - 5
FL - 21% - 27
MT - 21% - 3
NC - 18% - 15
SC - 16% - 8
MO - 15% - 11

This suggests AR and WV don’t make the list. 34 states plus DC do.

The 16 states Obama is least likely to win, where active appearances will likely be close to nil?

UT – 14% - 5
SD – 14% - 3
TX – 9% - 34
KS – 6% - 6
GA – 5% - 15
LA – 3% - 9
AZ – 2% - 10
ID – 1% - 4
WV – 0% - 5
TN – 0% - 11
AL – 0% - 9
KY – 0% - 8
OK – 0% - 7
AR – 0% - 6
MS – 0% - 6
WY – 0% - 3

Those represent the 141 electoral votes McCain will consider to be his base.

Obama should pursue a 34 to 36 state strategy (UT and SD are only 1% below MO). But as the first 17 states, from 79% to 100%, are fairly safe for Obama, most of his campaign will be directed at 17 to 20 states, half of them essential to a November win.

And if we set aside the knowledge of red state/blue state voting patterns as a rationale and ask if there’s any characteristics common to the reddest – those where Obama has less than a 10% chance of winning – I found these commonalities:

The old Confederacy plus the border states of KY and Bloody Kansas – there’s only 4 of the 5 Atlantic coast Confederacy states as the exceptions (VA, NC, SC, FL).

The home of the Aryan Brotherhood, ID, can be added to that group.

The coal and oil states – TX, LA, OK, WI, WV – with only AK and PA as exceptions.

McCain’s home state, AZ.

The odds against Obama remain long even in the exceptions, with only PA (55%) above 50%. AK is the next, at 26%.

Race and fossil fuels – the latter the villain in global warming and a driver of Middle East conflicts – remain the biggest factors working against Obama’s candidacy. But rather than the oft-asked question of ‘what’s wrong with Obama?’ the question should rightly be ‘what’s wrong with the majority of the voters in states where old hatreds and old technologies carry so much power over our foreign policy, environmental policy and political system?’

Note: the percentages cited for each of these states came from the anonymous author of the site FiveThirtyEight.com. In the past couple of days, that site’s drawn plaudits from Mark Blumenthal of Pollster.com and several top election expert bloggers for creating a demographic-based model for predicting the primaries. That model produced the most accurate in predicting the outcomes in IN and NC, besting all of the paid pollsters. I echo their praise and encourage readers to add the site to their group of Election Analysis Essentials.