Comments on: What is the probability your vote will make a difference? http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/ Why Americans Vote the Way they Do Tue, 15 Dec 2009 10:59:42 +0000 http://wordpress.org/?v=abc hourly 1 By: Nick http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-1040 Nick Thu, 06 Nov 2008 07:41:55 +0000 http://redbluerichpoor.com/blog/?p=161#comment-1040 Andrew: I had read the paper, though not as thoroughly as I perhaps should have. Clearly you didn't actually use the independent binomial model but I didn't understand why it should give such wildly different results. I'm guessing that the answer to that, though, is wrapped up inside the simulations which aren't described in this paper, hence the lack of enlightenment on my part. (Reading around a bit, it looks to me like perhaps a lot of the difference comes down to putting too much confidence in the polls.) Andrew: I had read the paper, though not as thoroughly as I perhaps should have. Clearly you didn’t actually use the independent binomial model but I didn’t understand why it should give such wildly different results. I’m guessing that the answer to that, though, is wrapped up inside the simulations which aren’t described in this paper, hence the lack of enlightenment on my part. (Reading around a bit, it looks to me like perhaps a lot of the difference comes down to putting too much confidence in the polls.)

]]> By: Needed: A Little Cost-Benefit Analysis « Organizations and Markets http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-953 Needed: A Little Cost-Benefit Analysis « Organizations and Markets Mon, 03 Nov 2008 03:16:46 +0000 http://redbluerichpoor.com/blog/?p=161#comment-953 [...] Andrew Gelman, Nate Silver, and Aaron Edlin estimate the the probability of any given vote being decisive in Tuesday’s US election is about one in 60 million. Residents of a few states, like New Mexico, Virginia, New Hampshire, and Colorado, are especially important: their chance is one in 10 million. Oh, did I mention that your vote doesn’t matter? [...] [...] Andrew Gelman, Nate Silver, and Aaron Edlin estimate the the probability of any given vote being decisive in Tuesday’s US election is about one in 60 million. Residents of a few states, like New Mexico, Virginia, New Hampshire, and Colorado, are especially important: their chance is one in 10 million. Oh, did I mention that your vote doesn’t matter? [...]

]]> By: Johnsenclan » Blog Archive » links 11/1/08 http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-946 Johnsenclan » Blog Archive » links 11/1/08 Sat, 01 Nov 2008 18:34:30 +0000 http://redbluerichpoor.com/blog/?p=161#comment-946 [...] likely is it your vote will count?  Depends where you live.  In New Mexico 1 in 10 million.  In Utah 1 in 100 million. i don’t care, I still felt [...] [...] likely is it your vote will count?  Depends where you live.  In New Mexico 1 in 10 million.  In Utah 1 in 100 million. i don’t care, I still felt [...]

]]> By: Why We Worry » Blog Archive » My vote is bigger than your vote | New articles published Monday-Friday http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-945 Why We Worry » Blog Archive » My vote is bigger than your vote | New articles published Monday-Friday Sat, 01 Nov 2008 16:55:19 +0000 http://redbluerichpoor.com/blog/?p=161#comment-945 [...] Here are some graphs from Andrew Gelman at Columbia University that illustrate that point: [...] [...] Here are some graphs from Andrew Gelman at Columbia University that illustrate that point: [...]

]]> By: Please Why » Blog Archive » New Mexico voters rejoice! http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-939 Please Why » Blog Archive » New Mexico voters rejoice! Fri, 31 Oct 2008 21:11:51 +0000 http://redbluerichpoor.com/blog/?p=161#comment-939 [...] Noah Using the latest polls, the folks at Red State, Blue State, Rich State, Poor State have calculated the probability that a single vote would decide the outcome of the presidential election, i.e.: [...] [...] Noah Using the latest polls, the folks at Red State, Blue State, Rich State, Poor State have calculated the probability that a single vote would decide the outcome of the presidential election, i.e.: [...]

]]> By: He May Not Have Ice In His Veins, But Espresso Will Do Just As Well « phaidimoi logoi http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-933 He May Not Have Ice In His Veins, But Espresso Will Do Just As Well « phaidimoi logoi Fri, 31 Oct 2008 19:14:51 +0000 http://redbluerichpoor.com/blog/?p=161#comment-933 [...] and I have bonded over coffee): protest/third-party votes aren’t without their purpose. And since your vote counts but doesn’t really, think about it. (Or just go read the article and call it even.) Posted in politics | Tags: [...] [...] and I have bonded over coffee): protest/third-party votes aren’t without their purpose. And since your vote counts but doesn’t really, think about it. (Or just go read the article and call it even.) Posted in politics | Tags: [...]

]]> By: United States of NM, NH, and VA « Yes, Let’s Talk About This http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-931 United States of NM, NH, and VA « Yes, Let’s Talk About This Fri, 31 Oct 2008 16:37:39 +0000 http://redbluerichpoor.com/blog/?p=161#comment-931 [...] United States of NM, NH, and VA Filed under: Politics — Tags: electoral college — Jon @ 12:37 pm There are lots of great reasons to live in a place like California or DC — having a direct say in the government isn’t exactly one of them. [...] [...] United States of NM, NH, and VA Filed under: Politics — Tags: electoral college — Jon @ 12:37 pm There are lots of great reasons to live in a place like California or DC — having a direct say in the government isn’t exactly one of them. [...]

]]> By: Ergotism › And You Say Your Vote Doesn’t Matter http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-930 Ergotism › And You Say Your Vote Doesn’t Matter Fri, 31 Oct 2008 16:17:42 +0000 http://redbluerichpoor.com/blog/?p=161#comment-930 [...] computed the probability that one single vote will swing the election in each state. In Pennsylvania, the odds that your vote will be the single, deciding vote is somewhere around 0.4 [...] [...] computed the probability that one single vote will swing the election in each state. In Pennsylvania, the odds that your vote will be the single, deciding vote is somewhere around 0.4 [...]

]]> By: Andrew http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-927 Andrew Fri, 31 Oct 2008 04:36:05 +0000 http://redbluerichpoor.com/blog/?p=161#comment-927 Abe: See recent blog entry for an explanation. Nick: Read the linked article. Our calculations come directly from the uncertainty distributions derived from Nate's simulations. The binomial distribution is irrelevant to the question. Tom0063: Well, we do mention Nebraska and Maine in a footnote . . . Abe: See recent blog entry for an explanation.

Nick: Read the linked article. Our calculations come directly from the uncertainty distributions derived from Nate’s simulations. The binomial distribution is irrelevant to the question.

Tom0063: Well, we do mention Nebraska and Maine in a footnote . . .

]]> By: tom0063 http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-922 tom0063 Thu, 30 Oct 2008 23:12:19 +0000 http://redbluerichpoor.com/blog/?p=161#comment-922 Again no breakdown of Nebraska by Congressional districts. Omaha - the only city in the USA with its own electoral vote, and a very even split between Obama and McCain, should certainly be up near the top of this chart. We'll split NE's vote and bring Electoral College reform to the rest of America. Again no breakdown of Nebraska by Congressional districts.

Omaha – the only city in the USA with its own electoral vote, and a very even split between Obama and McCain, should certainly be up near the top of this chart.

We’ll split NE’s vote and bring Electoral College reform to the rest of America.

]]> By: James H http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-921 James H Thu, 30 Oct 2008 22:20:20 +0000 http://redbluerichpoor.com/blog/?p=161#comment-921 This gives me a lot to think about between now and voting day. Thanks for putting such a good effort into it. James H. http://serviceafol.blogspot.com/ This gives me a lot to think about between now and voting day. Thanks for putting such a good effort into it.

James H.
http://serviceafol.blogspot.com/

]]> By: Nick http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-920 Nick Thu, 30 Oct 2008 20:42:34 +0000 http://redbluerichpoor.com/blog/?p=161#comment-920 This is kind of an obnoxious question since I haven't really taken the time to understand your methodology but... If I use a simple indpendent binomial model for voting in California, starting with a uniform distribution on the percentage of Obama voters and then conditioning on the last four poll results (giving McCain everyone who didn't respond for Obama), I seem to be getting the probability of a tied vote in California as around 10^-19. This is a bit lower than the roughly 10^-8 you have in your paper. Obviously there might be some problems with the independent binomial model but I think I was fairly generous in my calculation in other ways (e.g. the uniform prior and giving McCain all undecided/third-party votes). Is there a simple explanation for why I should believe that the probability of a tied election in California is closer to 10^-8/10^-9 than to 10^-19? This is kind of an obnoxious question since I haven’t really taken the time to understand your methodology but…

If I use a simple indpendent binomial model for voting in California, starting with a uniform distribution on the percentage of Obama voters and then conditioning on the last four poll results (giving McCain everyone who didn’t respond for Obama), I seem to be getting the probability of a tied vote in California as around 10^-19. This is a bit lower than the roughly 10^-8 you have in your paper. Obviously there might be some problems with the independent binomial model but I think I was fairly generous in my calculation in other ways (e.g. the uniform prior and giving McCain all undecided/third-party votes). Is there a simple explanation for why I should believe that the probability of a tied election in California is closer to 10^-8/10^-9 than to 10^-19?

]]> By: LnGrrrR http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-919 LnGrrrR Thu, 30 Oct 2008 20:27:05 +0000 http://redbluerichpoor.com/blog/?p=161#comment-919 Young, military NH voter here, putting my ballot in for Obama... Young, military NH voter here, putting my ballot in for Obama…

]]> By: Abe http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-918 Abe Thu, 30 Oct 2008 19:27:06 +0000 http://redbluerichpoor.com/blog/?p=161#comment-918 Actually, the chance that your vote would be decisive is zero. Even if the "true" vote were exactly 50/50, limitations in verifying such the count would never allow for a single vote to be decisive. Recount after recount would ensue, and none would be definitive to that degree of accuracy--not to mention the legal battles that would ensue. Actually, the chance that your vote would be decisive is zero. Even if the “true” vote were exactly 50/50, limitations in verifying such the count would never allow for a single vote to be decisive. Recount after recount would ensue, and none would be definitive to that degree of accuracy–not to mention the legal battles that would ensue.

]]> By: Your Vote Might Actually Count | NYU Local http://redbluerichpoor.com/blog/2008/10/what-is-the-probability-your-vote-will-make-a-difference/comment-page-1/#comment-917 Your Vote Might Actually Count | NYU Local Thu, 30 Oct 2008 17:07:03 +0000 http://redbluerichpoor.com/blog/?p=161#comment-917 [...] determined that a single vote in New Mexico, New Hampshire, Virginia, and Colorado has about a 1 in 10 million [...] [...] determined that a single vote in New Mexico, New Hampshire, Virginia, and Colorado has about a 1 in 10 million [...]

]]>