Tuesday
(1)Unlike our Dale Amon, I am not going to endorse a candidate - in fact, I am rooting for a 269-269 electoral tie, just for the sake of making history - but I still find the horse race intriguing. I was overwhelmed with requests (okay, two people asked) to run one last version of the election monte carlo that I offered last week. Apart from updating the probabilities, I did a few things differently this time:
- if the price was greater than 90 or less than 10, I changed it to 100 or 0, so that only the swing states impact the model.
- I kept track of which states were most likely to end up in the winners' column; I wanted to know which states were the kingmakers. (Well, we already knew which states, but I wanted a way to quantify it.)
- I ran a few different scenarios, taking different swing states off the table (i.e. setting their probabilities to 100 or 0.)
Scenario I: every swing state up for grabs
BUSH: 5972 wins, avg. 275.82 electoral votes KERRY: 3843 wins, avg. 262.18 electoral votes TIE: 185
Florida ends up in the winner's column 7578 of the 9815 scenarios where there is a winner. After that, the most 'decisive' swing states are Ohio (6515), Wisconsin (5636), New Mexico (5606) and Iowa (5521.)
Scenario II: Bush wins FL, everything else is up for grabs
BUSH: 8227 wins, avg. 287.70 electoral votes KERRY: 1586 wins, avg. 250.30 electoral votes TIE: 187
So basically, Kerry almost has to have Florida at this point.
Scenario III: Kerry wins FL, everything else is up for grabs
BUSH: 3083 wins, avg. 260.46 electoral votes KERRY: 6692 wins, avg. 277.54 electoral votes TIE: 225
Bush has more ways to win without getting Florida than Kerry does. Let's try one more ...
Scenario IV: Bush takes OH and WI; FL and other states are contested
BUSH: 8313 wins, avg. 291.05 electoral votes KERRY: 1515 wins, avg. 246.95 electoral votes TIE: 172
If Bush can take these two Midwestern states, he becomes a prohibitive favorite.
A few other desultory remarks:
- who says the country is more divided than ever? My favorite political story of the week: South Dakota, except for the Indian reservations, is a conservative state, and it is tough for a Democrat to win. So Democratic Congresswoman Stephanie Herseth, in the heat of a tough reelection battle, has pledged that, should the election end in a 269-269 tie, she will vote for Bush when the House of Representatives has to choose the president.
- Since I'm rooting for the 269-269 tie, here's one way it could happen:
- Finally, Megan McArdle, guestblogging for Glenn Reynolds, offers the best election day advice of all: use the electronic political markets to hedge, just like a farmer would use the grain futures markets to hedge against the possibility of low selling prices at harvest time. If you don't want Kerry to win, bet a few bucks that he DOES win, so you can at least drown your sorrows with some hard-earned beer.
If there was an electoral college tie, the newly elected House of Representatives would decide the Presidency. In all likelihood they would favour President Bush but this isn't completely certain because the House is also being elected today.
But the Senate chooses the Vice President independently. So in theory we could end up with a Democrat Vice President and a Republican President. I forsee an assassination attempt if the political culture remains this charged.
The latest calculations predict a 3.25% chance of a tie, much higher than in previous elections.
Still highly unlikely but it would certainly make for an interesting four years.
Posted by cessair at November 2, 2004 11:09 AM
Out of curiosity -- what's the source of the 3.25% stat, cess? In the runs I did, we were usually around 1.8% to 2.2%. I'm not the only person twisted enough to calculate the probability of an electoral vote tie?
Posted by Chris at November 2, 2004 02:11 PM
ALthough it might be fun to watch from overseas, any result that is within the "margin of lawyer" (in Mark Steyn's delightful phrase) will be hugely damaging to American society.
Posted by R C Dean at November 2, 2004 10:23 PM
Nathan Ritchey(Link), a Youngstown State University mathematician, calculated the 3.25 percent chance of a 269-269 tie in the electoral college. The odds have increased since late summer when chances for a tie were only 1.4 percent.
"Looking at the 10 closest states, there are 17 ways this can occur -- 17 out of 1,024 possible outcomes," said Ritchey, who has been tracking statistics in this year's contest between Bush and Democratic Sen. John Kerry of Massachusetts.
Posted by cessair at November 3, 2004 12:32 AM
