Tracking presidential elections state-by-state with the power of statistics since 2004.
| Victory probability map | Obama lead over time |
|---|---|
 |
 |
A revised writeup of the site's methodology is now available at this PDF page. It is intended to require little statistical knowledge, but that doesn't mean it's an easy read.
I'll also make a permanent link in the sidebar.
posted by Benjamin Schak at 12:03 AM
| Place | # of EV | Vote Est. | Win Prob. |
|---|---|---|---|
| National | |||
| US-Popular | - | 53 | 100 |
| US-Electoral | [538] | 366 | 100 |
| No Swing States | |||
| New England | |||
| ME | 2 | 58 | 100 |
| ME1 | 1 | 59 | 100 |
| ME2 | 1 | 56 | 99 |
| NH | 4 | 54 | 100 |
| VT | 3 | 65 | 100 |
| MA | 12 | 60 | 100 |
| RI | 4 | 63 | 100 |
| CT | 7 | 58 | 100 |
| Mid Atlantic | |||
| NY | 31 | 61 | 100 |
| NJ | 15 | 57 | 100 |
| PA | 21 | 55 | 100 |
| DE | 3 | 59 | 100 |
| MD | 10 | 61 | 100 |
| DC | 3 | 91 | 100 |
| Eastern South | |||
| VA | 13 | 53 | 100 |
| WV | 5 | 47 | 1 |
| NC | 15 | 51 | 78 |
| SC | 8 | 45 | 0 |
| GA | 15 | 47 | 0 |
| FL | 27 | 51 | 95 |
| Western South | |||
| KY | 8 | 44 | 0 |
| TN | 11 | 43 | 0 |
| AL | 9 | 41 | 0 |
| MS | 6 | 46 | 0 |
| AR | 6 | 45 | 0 |
| LA | 9 | 46 | 0 |
| TX | 34 | 46 | 0 |
| Great Lakes | |||
| OH | 20 | 52 | 100 |
| MI | 17 | 56 | 100 |
| IN | 11 | 51 | 65 |
| IL | 21 | 62 | 100 |
| WI | 10 | 55 | 100 |
| MN | 10 | 56 | 100 |
| Plains | |||
| IA | 7 | 56 | 100 |
| MO | 11 | 51 | 76 |
| ND | 3 | 50 | 45 |
| SD | 3 | 46 | 1 |
| NE | 2 | 43 | 0 |
| NE1 | 1 | 46 | 4 |
| NE2 | 1 | 48 | 10 |
| NE3 | 1 | 35 | 0 |
| KS | 6 | 44 | 0 |
| OK | 7 | 37 | 0 |
| Mountain West | |||
| MT | 3 | 49 | 12 |
| ID | 4 | 38 | 0 |
| WY | 3 | 39 | 0 |
| CO | 9 | 54 | 100 |
| UT | 5 | 37 | 0 |
| NM | 5 | 55 | 100 |
| AZ | 10 | 47 | 0 |
| NV | 5 | 52 | 99 |
| AK | 3 | 44 | 0 |
| Pacific | |||
| WA | 11 | 57 | 100 |
| OR | 7 | 57 | 100 |
| CA | 55 | 61 | 100 |
| HI | 4 | 69 | 100 |
| Obama electoral win (269+) | 100.0 |
| Obama electoral landslide (369+) | 53.9 |
| Obama electoral avalanche (469+) | 0.0 |
| Electoral tie (269-269) | 0.0 |
| McCain electoral win (270+) | 0.0 |
| McCain electoral landslide (369+) | 0.0 |
| McCain electoral avalanche (469+) | 0.0 |
| Popular/electoral split | 0.0 |
| Obama popular, McCain electoral | 0.0 |
| McCain popular, Obama electoral | 0.0 |
| Likeliest result (378-160 Obama) | 20.5 |
| Mean EV | 365.7 |
| Median EV | 375 |
| Mode EV | 375 |
| 95% confidence range for EV | 311-382 |
| Mean PV | 53.3 |
| 95% confidence range for PV | 51.8-54.9 |
5 Comments:
This post has been removed by the author.
Benjamin,
I just happened upon your site after reading a comment you posted in the thread for Marc Ambinder's inexplicably skewed "state of the race" post a few days ago. Your analysis needs a lot more exposure than it gets, and I can see that you realize that it's because of the way your present your results rather than the content of the analysis itself. Some unsolicited ideas:
(1) Become another Poblano. Somewhat different methodology, but the same engaging 1-5 paragraph posts. Most people who read 538 understand very little of the methodology--the posts are what make or break the site. If you shifted your priorities away from the analysis just slightly and gave yourself an extra 20 min/day to compose solid posts like you sometimes but not too frequently do, then you might be able to attract the same sort of people that 538 does at greater rates than you do now.
(2) Don't become another Poblano, but rather become something that his readers and commenters would also appreciate. Maybe less emphasis on the play-by-play and more on longer-range, semi-theoretical issues. Or maybe add interactive elements.
(3) The list of states and current numbers on the right: make it graphical.
(4) Maybe, just maybe, use data from the primaries to estimate where Clinton and Romney would be right now if they were the nominees. Probably a bad idea.
(5) Team up with someone who understands your methodology but is keener to write the sort of posts that will attract readers.
In any case, props to you, and I'm glad I discovered your site.
Thanks for the ideas.
(1) I'd love to, but just don't have the time. (I work long hours, and spend maybe 15-20 min/day total on this.)
(2) Yes, those sound like good ideas.
(3) Good idea -- probably feasible too.
(4) I have no idea how to did this. (Not that it couldn't be done, more that I lack the talent.)
(5) If there are any volunteers, I'd love a few guest posters.
Hey Benjamin -- just stumbled across your site -- and, echoing Jason, I think the work you've done is pretty awesome! I haven't had a chance to read the whole methodology pdf but I will say, I appreciate the use of LateX :)
This is stunningly good analysis.
Jimbo
Post a Comment
<< Home