Even as Hillary Clinton racked up victories in Pennsylvania, Indiana, West Virginia, Kentucky, and Puerto Rico, political analysts (echoed by cable news anchors) forecasted the final presidential primary contests on June 3rd (South Dakota and Montana) as “Obama states.”

This hasty analysis by political commentators and anchors was driven by several variables. First, there has been an undercurrent in the media to celebrate the new (Obama) and downplay the old (Clinton) at nearly every turn, even when Clinton was defeating Obama in a majority of state contests held after February 19th (Clinton has won 7 of 12). Secondly, analysts failed to realize that Obama’s dominance in Western states earlier in the year was largely driven by the caucus format—giving him victories in Alaska, Colorado, Idaho, Kansas, North Dakota, Washington, Nebraska, and Wyoming. Clinton only won two Western caucuses—Nevada and New Mexico.

But Smart Politics has stated repeatedly during the past three months that Clinton was likely to perform well in South Dakota and Montana (e.g. March 12, 2008: “Why Clinton Should Stay in the Race Through South Dakota”; April 2, 2008: “Clinton Dominates in Remaining Contests”).

Why? For one reason, these are primary contests and Clinton has won twice as many primaries in Western and Southwestern states (California, Arizona, Oklahoma, and Texas) than has Obama (Utah and Oregon).

And now a brand new poll by American Research Group of 600 likely voters in South Dakota (conducted May 31—June 1) finds Clinton with a 60 to 34 percent lead—numbers that clearly do not gel with the coverage of these Western states by the national news media. (Note: ARG’s final surveys gave Clinton a 36-point advantage in Kentucky (where she won by 36 points) and a 43-point lead in West Virginia (where she won by 41 points). ARG underestimated Obama’s 18-point victory in Oregon by 13 points).

ARG also finds Clinton within 4 points of Obama in Montana, trailing 48 to 44 percent.

These sparsely populated states may pose a problem for Clinton, as she did not get the turnout in Puerto Rico that she expected to overtake Obama in the popular vote count. Still, excluding Michigan and the four caucus states that have not released official vote counts, Obama leads Clinton by just over 24,000 votes.

In South Dakota, where 388,215 residents voted in the 2004 presidential election, a 26-point Clinton victory (if the ARG poll numbers hold) would net the New York Senator approximately 40,000 votes—according to a model of 40 percent turnout of the 2004 general election vote. Smart Politics, however, projects a higher turnout in South Dakota than 40 percent, though Clinton will not achieve a Kentucky- or West Virginia-esque margin of victory.

In Montana, where 450,445 residents voted in the 2004 presidential election, a 4-point Obama victory at 40 percent turnout of the 2004 presidential vote would net the Illinois Senator approximately 7,000 votes. The Clinton campaign is not expecting a victory in Montana—they are simply hoping to keep it close enough to not distill their hopeful big victory margin in South Dakota.

A 30,000+ net vote victory on Tuesday night for Clinton would elevate her ahead of Obama in the overall popular vote count—a dramatic finish to a Democratic primary for the history books. However, even if Obama loses the popular vote and is seen as “backing into the playoffs,” it seems unlikely that Clinton will change the momentum of superdelegates who clearly seem to be falling into Obama’s camp.

Smart Politics will blog live Tuesday night, beginning at 8 p.m. when polls close in South Dakota.