|
Title: WI: Other Source: Obama up 15 in WI Post by: Wiz in Wis on October 18, 2008, 07:50:20 PM New Poll: Wisconsin President by Other Source on 2008-10-17 (https://uselectionatlas.org/POLLS/PRESIDENT/2008/pollsa.php?action=indpoll&id=5520081015106)
Summary: D: 51%, R: 36%, I: 0%, U: 12% Poll Source URL: Full Poll Details (http://election08data.blogspot.com/2008/10/wisconsin-survey.html) This poll was conducted under the supervision of Dr. Thomas Holbrook at UW-Milwaukee. Dr Holbrook also has a election projection here (http://research.uvsc.edu/DeSart/forecasting/october.html) Title: Re: WI: Other Source: Obama up 15 in WI Post by: Alcon on October 18, 2008, 08:35:03 PM Ah, the DeSart-Holbrook model
Still not good. Title: Re: WI: Other Source: Obama up 15 in WI Post by: Eraserhead on October 18, 2008, 11:31:23 PM Probably not all that far off anyway though.
Title: Re: WI: Other Source: Obama up 15 in WI Post by: TheGlobalizer on October 19, 2008, 12:42:49 AM I don't buy more than +10.
Title: Re: WI: Other Source: Obama up 15 in WI Post by: Gern Blandsten on October 19, 2008, 09:29:33 PM Ah, the DeSart-Holbrook model Still not good. Not good? It doesn't look much different than everyone else's. ??? The only thing in their prediction that might be suspect is wv, and possibly mo. Everything else looks pretty solid. Title: Re: WI: Other Source: Obama up 15 in WI Post by: they don't love you like i love you on October 19, 2008, 09:31:54 PM Ah, the DeSart-Holbrook model Still not good. Not good? It doesn't look much different than everyone else's. ??? The only thing in their prediction that might be suspect is wv, and possibly mo. Everything else looks pretty solid. Holy sh!t. FANTASTIC name! Title: Re: WI: Other Source: Obama up 15 in WI Post by: Alcon on October 19, 2008, 09:34:19 PM Ah, the DeSart-Holbrook model Still not good. Not good? It doesn't look much different than everyone else's. ??? The only thing in their prediction that might be suspect is wv, and possibly mo. Everything else looks pretty solid. Their model is flawed, whether or not it gets good results. Title: Re: WI: Other Source: Obama up 15 in WI Post by: Gern Blandsten on October 20, 2008, 01:57:40 AM Ah, the DeSart-Holbrook model Still not good. Not good? It doesn't look much different than everyone else's. ??? The only thing in their prediction that might be suspect is wv, and possibly mo. Everything else looks pretty solid. Their model is flawed, whether or not it gets good results. Well, that's a pretty empty statement. By your logic, even if it generates perfect predictions it's still flawed? Title: Re: WI: Other Source: Obama up 15 in WI Post by: TheGlobalizer on October 20, 2008, 01:20:12 PM Ah, the DeSart-Holbrook model Still not good. Not good? It doesn't look much different than everyone else's. ??? The only thing in their prediction that might be suspect is wv, and possibly mo. Everything else looks pretty solid. Their model is flawed, whether or not it gets good results. Well, that's a pretty empty statement. By your logic, even if it generates perfect predictions it's still flawed? A random number generator is accurate predicting the current time once in a while. Doesn't make it comparable to a clock. Title: Re: WI: Other Source: Obama up 15 in WI Post by: Alcon on October 20, 2008, 01:21:31 PM Well, that's a pretty empty statement. By your logic, even if it generates perfect predictions it's still flawed? Yes; see TheGlobalizer's analogy. If the foundation of a model doesn't make sense and can't be reasonably explained, getting lucky does not make it a good model. And every model is flawed, this one moreso than others. It is, IIRC, primarily based on unchecked but time-weighted polling information (hence WV), which is cool but not an especially good model. Title: Re: WI: Other Source: Obama up 15 in WI Post by: Gern Blandsten on October 21, 2008, 02:12:42 AM Well, that's a pretty empty statement. By your logic, even if it generates perfect predictions it's still flawed? Yes; see TheGlobalizer's analogy. If the foundation of a model doesn't make sense and can't be reasonably explained, getting lucky does not make it a good model. And every model is flawed, this one moreso than others. It is, IIRC, primarily based on unchecked but time-weighted polling information (hence WV), which is cool but not an especially good model. So, by your standard, how many times does a model have to be accurate (and how accurate must it be) to pass the "randomizer test"? It seems to me that the logic underlying the model makes perfect sense. Candidates tend to win states in which they hold the lead in September, while controlling for the historical leaning of the state. I'm not sure if they weight the poll data, I don't think so. Not sure what you meanby "unchecked." I saw them present it at a conference a couple years ago. It was a real simple model, and I think they were real close to the actual result in 2004. R-Squares over .9 Title: Re: WI: Other Source: Obama up 15 in WI Post by: Alcon on October 21, 2008, 02:15:09 AM In other words, it's just a polling average with no regard for the passage of time, pollster quality, etc.?
I'm sorry, but those things are essential in a model to me, or you get some zany results (see: WV currently, or almost in '04.) They shouldn't just coast on the availability of polls like that. And what were they within that margin on? |