Many have been asking: who won our famous Presidential Prediction Contest? The task, you will remember, was to predict how the popular vote would be distributed between the two major candidates, throwing away third-party votes and ignoring the electoral college altogether, expressed as a percentage of votes for Obama. So if the total number of votes for Obama is VO , and the total number for McCain is VM , the number you were predicting is
We’ve been delaying the announcement of the results, as the entries were tightly bunched and it takes time for votes to trickle in. Indeed, Alaska still seems to be problematic, but patience is thin and it’s time to declare a winner! Visit here to be reminded of who had staked out which bits of territory. Here are the vote totals as of today:
Barack Obama: 66,679,600
John McCain: 58,227,508
which implies
f = 53.38.
The relevant entries, courtesy of wqz, are
-
( 52.81689, 53.10869): Tim
( 53.10869, 53.32282): Elliot
( 53.32282, 53.47922): Anonymous Snowoboarder
( 53.47922, 53.74739): Gabe
( 53.74739, 53.78569): joulesm
And so the winner is:
Who, I am guessing, may have mistyped their name. But when you have mad prediction skills like that, who cares about typographical irregularities?
Here were the distribution of predictions near the right value:
It’s interesting to note that the contest was announced in late June, just when Obama was hitting his summer peak of popularity (which was not as pronounced as his fall peak of popularity). I wonder how the predictions would have gone had we done the contest in September?
November 15th, 2008 at 1:21 pm
/pedant mode on
How do those vote totals *imply* that f = 53.38?
Given that f is defined as it is, 53.38 simply *is* the value of f at this time.
/pedant mode off
(sorry, couldn’t resist)
November 15th, 2008 at 1:34 pm
If one aspires to be a pedant, it’s important to be right.
Logical implication.
November 15th, 2008 at 1:58 pm
Indeed.
However, “implies” is a weak (read: non-specific) statement of relation, whereas “equals” is a strong one, and it is generally considered poor form for a pedant to use the former when the latter is applicable.
BUT
You win the pedant war just for bringing up the formal logic definition of “implies” in the first place. I, alas, was merely referring to the common usage of the word.
Oh, the agony of defeat!
November 15th, 2008 at 2:01 pm
While we’re being pedantic, if you want to include the 100 term in that first formula at all, it needs to be in the numerator. If you do include it, the second formula is out by 100: f = 53.38 rather than f = 53.38% .
November 15th, 2008 at 2:05 pm
See, that’s an example of being right! Thanks, Jan.
November 15th, 2008 at 2:33 pm
I don’t know if Blake’s being serious or not, but on the off chance that he is:
Dude, I’ve been reading this blog for a long time now and it’s never even been primarily about physics. It’s mostly cultural commentary with a liberal, godless bent and a place for the authors to post their itineraries.
The latter being physics related, I suppose, though most of those posts are just mountains of platitudinous text describing how wonderful everyone at conference X was (but what do you expect? These people are blogging with their real names).
Not that I mind, or anything (in fact, I particularly enjoy the cultural commentary with a godless bent), it’s just that people need to get real about what these blogs actually are: Blogs. Not second jobs.
November 15th, 2008 at 3:47 pm
Can you plot the predictions in a more narrow window? For example, between 53.08 and 53.68? As it is there is no resolution to see the relative weights of the best predictions (in fact, I can’t even see what color won!).
November 15th, 2008 at 3:59 pm
Never have I been happier to lose.
November 15th, 2008 at 4:23 pm
Zoomed plots are here. More usefully, Mathematica code is here. Plot to your heart’s content!
November 15th, 2008 at 5:40 pm
Yes, that is me.. Bee at Backreaction can attest to that
Wow… shocking! I’m up here in VT for a few days (doing a tax appeal on my winter condo) and just sat down to take a spin thru the blogs and saw the post.
I’ve posted my website (hence I guess I’m no longer anonymous if anyone really
wants to dig around). I was involved in the Ron Paul campaign last year and have always been a politcal junkie, polls in particular. Ironically, I dragged the assistant leader of my RP group is here too.. (why are you shouting!?!). Here is an example of my past poll geekery
http://www.gedankenexperiment.dk/pivot/entry.php?id=453&w=rant
Sadly (?) my blogging has dropped off quite a bit the past year, I think largely out of burnout from the RP effort and subsequent events. Being an ‘R’ and seeing the party
go down the tubes was painful – I even gave up trying to talk any sense into the Redstaters as it became a depressing exercise in frustration.
Cheers,
semi-anonymous Snowboarder
November 15th, 2008 at 6:01 pm
AS, don’t despair. With scientific proof of your prognostication skills, you can surely make a lot of cash as a consultant.
Let us know your T-shirt size, if you’d like to collect a prize.
November 15th, 2008 at 6:27 pm
Aww, I was so close. Also, if we use Obama’s percentage counting third parties, 52.7, it is also very very close to my prediction.
Next time!
November 16th, 2008 at 6:13 am
Congratulations Snowboarder,
I thought I had a shot but as the votes kept coming in and crossed 53.3xxx, I knew I was out. But it was fun to be sure. As some consolation, I did win an election night electoral vote contest where I went 364 for Obama.
As an early Obama supporter, I sincerely hope that he will bring the country together.
Thanks to CV for doing this. Isn’t it about time for a College Basketball prediction pool to fill the void? Maybe pick the NCAA final four NOW or something like that?
e.
November 17th, 2008 at 9:00 am
Wooo!! How exciting
November 17th, 2008 at 11:29 am
Looks from the curves that Reverend Wright was in effect for a month, but Sarah Palin… for what? A week? A few days?
PS: somebody != Somebody.
November 18th, 2008 at 6:29 pm
A t-shirt! this is getting better all the time… Sean I sent you a private email
to the address listed for you on the website.
Cheers
November 18th, 2008 at 6:33 pm
Wait, the winner is only part of it. Shouldn’t you rank everyone else in order so we can see who came in 2nd, 3rd, as so on? How will we determine bragging rights otherwise?
November 27th, 2008 at 3:16 pm
Ah, cool! Looks like I was 3rd.
Not bad for a random guess with little prior thought.