tag:blogger.com,1999:blog-6926413038778731189.post6216078259336621920..comments2023-10-16T07:13:12.123-05:00Comments on A plain blog about politics: Does VA Gov Explain the 1992 Perot Vote?Jonathan Bernsteinhttp://www.blogger.com/profile/15931039630306253241noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-6926413038778731189.post-91493950485817392222013-11-08T23:01:25.266-06:002013-11-08T23:01:25.266-06:00Thing is, though, we're never going to stop ta...Thing is, though, we're never going to stop taking polls (and aggregates of polls can indeed be useful). To that end, we ought to understand them as well as possible.theBitterFignoreply@blogger.comtag:blogger.com,1999:blog-6926413038778731189.post-15460074897808324812013-11-08T16:44:26.347-06:002013-11-08T16:44:26.347-06:00I suspect that there's a 99 percent likelihood...I suspect that there's a 99 percent likelihood that this stuff is just noise.<br /><br />Polls are just not as accurate as everyone in politics wants them to be, because they aren't election returns (and thus are not measuring the same thing) and because statistics make everything look more objective and precise than it really is.<br /><br />The thing is, polls are all that pundits have, so they have to rely on them, and assume they are accurate and more important than they are. Instead of realizing "hey look, sometimes these things are going to be off, and when they are it could be for all sorts of reasons and we shouldn't worry about it too much, we just should remind ourselves not to get too caught up in polling before elections".Dilan Esperhttp://www.twitter.com/dilanespernoreply@blogger.comtag:blogger.com,1999:blog-6926413038778731189.post-15599381984148740972013-11-08T15:05:50.260-06:002013-11-08T15:05:50.260-06:00Steve Kornacki did a thorough debunking of the Per...Steve Kornacki did a <a href="http://www.salon.com/2011/04/04/third_party_myth_easterbrook/" rel="nofollow">thorough debunking</a> of the Perot myth some time back.Kylopodhttps://www.blogger.com/profile/06932528611103718373noreply@blogger.comtag:blogger.com,1999:blog-6926413038778731189.post-43209901507391760792013-11-08T13:28:05.442-06:002013-11-08T13:28:05.442-06:00Maine Gubernatorial 2010 somewhat fits into this, ...Maine Gubernatorial 2010 somewhat fits into this, but it of course has it's own wrinkles.<br /><br />RCP ( http://www.realclearpolitics.com/epolls/2010/governor/me/maine_governor_lepage_vs_mitchell_vs_cutler-1231.html ) had LePage leading by nearly 13 points in their last poll average, with his actual win being less than 2% over Cutler. <br /><br />Of course, the major wrinkle is that Cutler was the 3rd party candidate, finishing as the first runner up, 36.5% to 38.3%, with Democrat Libby Mitchell at 19.1%. Polling further back, in the 2-3 weeks before the election range, had Mitchell polling around 29%, Cutler around 14%, and LePage around 32% (balance being around 25%).<br /><br />For the final results, 6% still went to someone outside the top three, so 19% of the 25% undecided took sides, Libby dropped 10 points, LePage gained 6 points, and Eliot Cuttler a whopping 22.<br /><br />While this does slightly fit the pattern, being a Mainer at the time, it felt like what was going on was that Cutler was rising dramatically at Mitchell's expense. Took folks a while to realise that if they didn't want a Republican, they had to vote 3rd party. Things had been shifting this way for weeks, and if the election were held a week later, it's plausible Cutler would have won. LePage underperformed the two Republican candidates for US House of Representatives in 2010 by just over 5 points, so this seems more like a problem of the center-left coalition not getting itself straightened out, then failing to be first past the post.theBitterFignoreply@blogger.comtag:blogger.com,1999:blog-6926413038778731189.post-18093915097712030812013-11-08T13:15:57.465-06:002013-11-08T13:15:57.465-06:00How does one deal with the turnout differential po...How does one deal with the turnout differential portion of that? Namely, it might be coming home or whatever, but it could also be voting/not depending on perceived closeness. So, I'm seeing a lot of complicated things to disentangle.Matt Jarvisnoreply@blogger.com