If I extend the theme of my maps I can create a table showing the base EV at each level of likelihood. For example the 95% likely map from 538 gives Obama 149 and Romney 152. Here's how that changes with likelihood:
| Obama | Romney |
95% | 149 | 152 |
90% | 207 | 155 |
85% | 210 | 166 |
80% | 237 | 181 |
75% | 247 | 191 |
70% | 251 | 206 |
65% | 257 | 206 |
60% | 276 | 235 |
55% | 285 | 235 |
50% | 303 | 235 |
Clearly Obama has a lot more EV leaning his way than Romney does. Even so, Obama only breaks 300 EV when all states are allocated (ie 50%). Romney needs to get the states that are less than 80% likely for Obama to break 300.