Obama won't get 38% in Oklahoma.
It's entirely possible that he could actually. That's not the surprising part here. The surprising part is that McCain is only at 52%. I expect by election day that'll he be over 60% but Obama getting 38% is hardly completely unreasonable.
Careful. I'm sure that AHDuke will try to pervert what you said into a quote about Obama being able to win Oklahoma. 
Nice try at trying to pervert my statements, buddy. Obama can get 38% in Oklahoma if he wins by 6-8% nationally. I don't think he'll outperform Kerry if the margin is less than 6% on either side.
What is the basis for this idea?
A look at recent history might be helpful here...
Kerry in 2004: 34%
Gore in 2000: 38%
Clinton in 1996: 41%
Clinton in 1992: 34%
Dukakis in 1988: 41%
Just for fun, let's do a back-of-the-napkin calculation of Obama's expected share of the vote in OK given the historical vote totals:
Relative to the 2-candidate vote, the historical % Dem. vote becomes:
Year - Dem. %
2004 - 34.43%
2000 - 38.92%
1996 - 45.60%
1992 - 44.37%
1988 - 41.61%
Relative to the national, 2-candidate margin (), this becomes:
Year - Dem. %
2004 - NPV-14.33% (2.48%)
2000 - NPV-11.34% (0.53%)
1996 - NPV-9.13% (9.46%)
1992 - NPV-9.09% (6.91%)
1988 - NPV-12.29% (7.80%)
Assuming a 4% margin swing (i.e., 2% of voters swinging) due to a neighboring son effect, this becomes:
Year - Dem. %
2004 - NPV-12.33%
2000 - NPV-9.34%
1996 - NPV-9.13% (AR roughly cancels KS)
1992 - NPV-9.09% (AR roughly cancels TX)
1988 - NPV-10.29%
which gives an average of NPV-10.04%
So a 52.5%-47.5% 2-candidate NPV win by a generic Democrat against McCain (i.e., a 51%-48% NPV win by Obama, assuming that he under-performs the generic Dem. by a 2% margin) would yield 52.5%-10.04% = 42.46%, which corresponds to 42.04% of the total vote (assuming 1% of the vote goes to 3rd Parties). Assuming a net racially motivated faction amounting to ~13% of Democrats = ~6.5% of voters (as per the exit polls from the OK Dem. Primary, i.e., assuming no Bradley Effect on top of the openly admitted racially motivated voting) swinging away from Obama as compared to the generic Dem. leads to Obama getting 35.54% of the vote, which corresponds to McCain getting 63.46% of the vote. This is very close to my current model's projection (based on polls, and not historical data) of 63.01%-35.99% (having taken this poll into account).
