Historically, there has been no case of a candidate winning the presidential election while losing all three top bellwethers: Missouri, Nevada, and Ohio. Nev. was the last of the three to join the union, and get the vote, in 1864.
While Republicans and Democrats have squred off since 1856, the late-1800s saw some winners prevail in only one of the three states (GOP William McKinley won Ohio in 1896 and 1900; no Republican president has ever won election without the Buckeye State). Since 1908, the last 25 elections saw Mo. back the loser in 1956, Nev. going for the 1908 and 1976 runners-up, and Ohio did not picking the winners of 1944 and 1960. Agreeing as a trio in 20 of the last 25 elections—for 80 percent—Mo. has been right 96 percent, and Nev. and Ohio both 92 percent.
Mo., Nev., and Ohio are each included in the states that I predict in this year's election—one that I believe will show Sen. Barack Obama (D-Illinois) prevailing in the Electoral College.
Since Sam credits Obama with Nevada, I just wanted to present this for anyone else finding it worth consideration.
These arguments are cute, but they are born out of coincidence, not logic.
You can find millions of those little things. Remember Bush proving the "No president has won if his approval rating is below..." comments wrong? Or one of my favorites "No incumbent President has won when Redskins take home loss in game preceeding election."
If you just take, say, ~27 elections going back to 1900 or so, think about how many frickin' variables there are! There are literally an infinite amount of variables. You can find "No president with an E in his last name has won without Delaware" sorts of things all the time. It's not surprising at all that three states that have stayed relatively close to the center have been a component of every victory, but it's meaningless in terms of predictive capability.
