I'm seeing another interesting and easy to grasp pattern in Nate's projected election simulation data at 538.
This one's quick and simple, but I think it illuminates some important electoral issues for both the Obama and McCain campaigns.
Starting from Nate's state-by-state simulation results, divide them up into categories: Those states a candidate wins 90% or more of the time, those he wins in the 80-percents, those in the 70s, 60s, 50s, and 40s. What you will find is the following:
90+: Obama 167 EV, McCain 64 EV
80s: Obama 40 EV (running total 207 EV), McCain 87 EV (running total 151 EV)
70s: Obama 45 (252), McCain 6 (157)
60s: Obama 0 (252), McCain 48 (205)
50s: Obama 41 (293), McCain 40 (245)
40s: Obama 40 (333), McCain 41 (286)
Note that for both Obama and McCain there is a hole where they each gain few to no additional electoral votes, thereby dividing their categories into those above and below the hole. For Obama, there are currently no states in which he wins between 60% and 69% of the simulated elections. Above that he's got a fairly solid 252 EV in the 70s, 80s, and 90s. Obama's best chances to win the election are to collect more than 18 EV from his 50s states: NH (4), MI (17), OH (20).
McCain's hole falls in the 70s, where he collects additional EVs only from SD and AK. He's got a solid 151 EV in the 80s and 90s, but he's got to win almost all of his 40s states to win the election.
All in all, the projections are looking good for Obama. They will look great if he can move his 50s states up into at least the 60s, and move his 40s state up into the 50s. Then the November election will start to look like something less stressful than a nail-biter. A 400+ EV landslide continues to look unlikely, however, since McCain is unlikely to lose his hardcore 151 EVs. That puts a cap on Obama's possible EVs at 387.