Daily Kos

View Story | 217 comments

  •  Where do you get 60K? (none / 0)

    Where do you get 60K?  I read Mr. Schaller's post several times and I can't find any number for Dean's hard count.  All I see is that 60-65% of Dean's hard count are first-time caucus voters but that does not mean much if his hard count is 25K (by example).

    Am I missing something?

    by gimmeaD on Sun Jan 18, 2004 at 04:14:10 PM PDT

    [ Parent ]

    •  Math (4.00 / 5)

      Can't speak for Karmakin's math, but just crunch some numbers yourself.

      Polls usually go off of old voting lists. So if Dean got 20% of a (normal year) 130K caucus, that would be 26K experienced voters. But if new voters are 60% of Dean's hard count, then that's an additional 39K for 65K total.

      Highly unscientific, but it makes the significance of the rumors a lot clearer.

      by ploeg on Sun Jan 18, 2004 at 04:33:20 PM PDT

      [ Parent ]

      •  I'm not sure (none / 1)

        I'm not sure that follows.  Even if 60-65% of Howard Dean's hard count is from first-time caucus-goers, a great many of those would still appear on the voter registration lists (no matter how old they are).

        by gimmeaD on Sun Jan 18, 2004 at 04:41:04 PM PDT

        [ Parent ]

        •  ...as democrats (none / 0)

          Most polls go off the SOS voter registration list, and take into account only those who were registered as Democrat. Certainly there would be a certain number of registered Democrats who have never taken part in a caucus before, but I suspect most of the first timers are people who have registered recently or changed affilliation.

          Certainly there is cause for dissatisfaction with the 65K figure, so crunch your own numbers. Still adds up to a lot of people (assuming that the 60% figure is true, and assuming that they actually show up).

          by ploeg on Sun Jan 18, 2004 at 04:53:03 PM PDT

          [ Parent ]

          •  That's fine (none / 0)

            That's fine and I'm not referring to the polls.  The polls would obviously be innacurate if they're based on faulty calling lists.  What I'm referring to is the 'hard count' which has nothing to do with polls.  Where is the 60K number coming from?  The highest I have seen for Dean from any remotely reliable source was 40K, and that was from before he started losing support.

            by gimmeaD on Sun Jan 18, 2004 at 05:09:36 PM PDT

            [ Parent ]

            •  Hard count (none / 0)

              As I said up above in my first post, I don't know where Karmakin got his 60K number. And nobody that I know of ain't talkin' just yet.

              I discerned that you were having trouble seeing the significance of the (rumored) 60-65% first-timer figure, so I decided to put together an estimate of my own, based upon the (systematically inaccurate) poll numbers and the (speculative) first-timer estimate. Take it as you will.

              by ploeg on Sun Jan 18, 2004 at 05:29:31 PM PDT

              [ Parent ]

              •  Oh, I see (none / 0)

                We must've miscommunicated.  Probably my fault.  No, I think the first-timer figure is quite significant.

                by gimmeaD on Sun Jan 18, 2004 at 05:33:45 PM PDT

                [ Parent ]

              •  If Dean's Hard Count is 40,000 . . . (none / 0)

                He may win this in a walk.  Record turnout it 120k, if I'm not mistaken.

                A 33% increase over the previous record is only 160k.  Dean would have 25% of that.

                If there are candidates -- the "other three" who pony up 40k of their own, or more, they deserve the accolades.

                Keep that numerator at 40 and ratchet down the demoninator and things look better and better for Dean.

                I'm ready for anything on Tuesday, but the confidence level -- give the huge slide in the polls -- must mean something.

                Trippi et al have made me nervous with "questionable" promises for expectations but have yet to let the DFA community down.

                An if Dean's HC is 50k or 60k, fagghidaboudit.

                John McCain a/k/a John Sidney "Grampy McSame"

                by MRL on Sun Jan 18, 2004 at 05:40:52 PM PDT

                [ Parent ]

          •  Wrong (none / 0)

            Pollsters do NOT call lists of registered Democrats.  We call random households, asking them demographic questions (in addition to the poll questions) in order to properly weight the respondent.  

            The poll questions that determine "likely voter" status go like this:

            -What is your political party?
            -Are you registered to vote?
            -Did you vote in the last election?
            -Do you intend to vote in the next election?

            Answer "Yes" to the last three questions, and BINGO...  You are a "Likely Voter."

            by Politus on Sun Jan 18, 2004 at 06:55:03 PM PDT

            [ Parent ]

            •  How useful are these questions in Iowa? (none / 0)

              Since voters can register or change affiliation at the caucuses, the first three questions will certainly leave out some likely voters.

              by gxschier on Sun Jan 18, 2004 at 07:57:37 PM PDT

              [ Parent ]

            •  You obviously aren't working for the DM Register (none / 0)

              http://tinyurl.com/ytwbf

              The Iowa Poll, conducted for The Des Moines Register by Selzer & Co. Inc. of Des Moines, is based on interviews with 606 registered Iowa voters who say they definitely or probably will attend the Democratic caucuses. Interviewers contacted voters registered as Democrat, Green or no-party by using randomly selected telephone numbers from the Iowa secretary of state's voter registration list. Percentages based on the full sample may have a maximum margin of error of plus or minus 4 percentage points. Republishing the copyrighted Iowa Poll without credit to The Des Moines Register is prohibited.

              Further, your methodology has its own obvious issues (given that the number of honest-to-God caucus goers is so small compared with the total number of households).

              by ploeg on Sun Jan 18, 2004 at 07:58:11 PM PDT

              [ Parent ]

            •  Politus Got It Wrong (none / 0)

              Politus: you're the one who is wrong.  Most of the candidates are using listed sample, which is typical in primary/caucus races.  RDD is more common in general election races.

              by The Fool on Sun Jan 18, 2004 at 08:08:51 PM PDT

              [ Parent ]

      •  I just want to say... (none / 0)

        That I almost did a terrible thing just now.

        I was going to rate this diary.  But just as I was clicking on the rate-all button, my finger touched the roller & it changed my rate to a 1, rather than the rating I meant.

        This is always my fear when I want to rate posts.  I always worry that If I don't press the rate-all button for each one, that I will accidentally scroll a previous rating to something I don't mean.

        Sorry about being off topic, but it really rattled me.

        by katiebird on Sun Jan 18, 2004 at 04:42:14 PM PDT

        [ Parent ]

        •  Ratings backtrack (none / 0)

          I'm nearly sure that you can change a rating by simply entering a new one and hitting the ratings trigger. I haven't tested it, but the ratings-entry window is still there even after rating a post.

          -- Rick Robinson

          The best fortress is to be found in the love of the people - Niccolo Machiavelli

          by al Fubar on Sun Jan 18, 2004 at 04:49:52 PM PDT

          [ Parent ]

        •  Rate All (none / 0)

          BTW, I've been wanting to ask wtf does "rate all" mean?  I just want to rate one post once in a very great while, since for the most part I consider these ratings a huge pissing contest.  Rate all freaks me out and I've been too lazy to look up the instructions that came with this game.

          by bbuster on Sun Jan 18, 2004 at 05:06:50 PM PDT

          [ Parent ]

          •  me too (none / 0)

            I have this same question, and I did try to look it up in the instructions.  Why does it say "rate all" next to each individual comment rating box?  What happens when you click it?  It seems to only register the rating for that one comment, but I've never been absolutely sure.

            by jsavage on Sun Jan 18, 2004 at 05:19:36 PM PDT

            [ Parent ]

            •  Rate all (4.00 / 2)

              What you do is set the number on all of the posts that you'd like to rate from the pull-down menus.  Then, once you've hit the bottom of the page, you click "Rate All," and the ratings get applied to every one of the posts you've set a number to.  It will not rate every post; just the ones you've put a number on.

              blog | -6.13, -5.95 | Live every week like it's Shark Week.

              by folkbum on Sun Jan 18, 2004 at 05:29:15 PM PDT

              [ Parent ]

              •  Ratings (none / 0)

                OK. So why do some posts have a rating of none, but a count of 1 rather than 0?

                This has bothered me for a while. I have yet to see a count of 2 or higher for a rating of None.

                Democrats stand for Liberty, Security, Support of Families and Opportunity Whiskey Tango Foxtrot - over

                by Rick B on Sun Jan 18, 2004 at 06:58:51 PM PDT

                [ Parent ]

                •  ya need 2 ratings (none / 0)

                  to make an average that gets used.

                  none/1 is when just 1 person has rated the post.

                  i toss out ratings often, just because i believe good posts should be rewarded, and because i feel it's part of our duty to try and rate accurately.

                  a sometimes genius

                  by skaiserbrown on Sun Jan 18, 2004 at 07:15:18 PM PDT

                  [ Parent ]

                •  Duh (none / 0)

                  Those counts are the TOTAL number of ratings given by dKos readers.

                  You're not alone here, don't you know? ;)

                  The "none" (or whatever you set it to) in the dropdown box is the rating that you have given (or will give if you hit "Rate all").

                  by Felix Deutsch on Sun Jan 18, 2004 at 07:25:44 PM PDT

                  [ Parent ]

                  •  Ratings (none / 0)

                    If so, then there should be very few None/0 if there are None/1 's. The reverse seems to be the case. Most ratings are None/0 with a few None/1 's.

                    I am of course ignoring the real ratings that also appear in the list for this discussion. Any rating that is 1 through 4 also seems to have a reasonable number after it, such as 4.00/2 or something. It is merely the None/1 's which strike me as curious.

                    Democrats stand for Liberty, Security, Support of Families and Opportunity Whiskey Tango Foxtrot - over

                    by Rick B on Sun Jan 18, 2004 at 08:40:04 PM PDT

                    [ Parent ]

                    •  Aha! (none / 0)

                      Ok, I was under the impression that you were talking about the "none" in your ratings box.

                      The one next to the totals means overall rating for this post and it takes at least 2 ratings to change a "none" to an overall average. Otherwise "none/1" is show, regardless of the rating.

                      by Felix Deutsch on Sun Jan 18, 2004 at 08:52:55 PM PDT

                      [ Parent ]

                      •  Ratings (none / 0)

                        Oh! The light goes on.

                        If I rate your comment as a 4:Excellant [as, of course, everything which falls on the screen from your keyboard obviously deserves], and I am the first to do so, it will appear as None/1 because there is only a single rating of any numerical value. Only when a second rating is applied would there be an average posted.

                        Have I got it?

                        Though, if that is the case, I see nothing wrong with a 4.00/1 as an average.

                        Democrats stand for Liberty, Security, Support of Families and Opportunity Whiskey Tango Foxtrot - over

                        by Rick B on Sun Jan 18, 2004 at 09:09:40 PM PDT

                        [ Parent ]

                        •  Yes (none / 1)

                          Though, if that is the case, I see nothing wrong with a 4.00/1 as an average.

                          It's because it should be at least two votes for an  overall rating to count, for they contribute to your trusted user status (many recent good ratings give you special powers).

                          BTW you can check out the individual ratings by clicking on the "none/1" (or whatever is displayed).

                          by Felix Deutsch on Sun Jan 18, 2004 at 09:49:03 PM PDT

                          [ Parent ]

                          •  Ratings - reply to Yes (none / 0)

                            OK. I think I've got it. Clicking on the reported rating made it clear what you were telling me. Thank you very much.

                            Now, for the always existing political question. Who established the rule that only two or more ratings count? Kos? Whoever designed the system we are using? Someone else?

                            And is there a user's guide of all the features on this commenting system? If so, is there an ~index~ that we not-so-geeky-ones can use to get quick answers to the system details (not to say quirks - no no, I would NEVER say quirks.)

                            Democrats stand for Liberty, Security, Support of Families and Opportunity Whiskey Tango Foxtrot - over

                            by Rick B on Mon Jan 19, 2004 at 07:37:28 PM PDT

                            [ Parent ]

                  •  Duh (none / 0)

                    Maybe I wasn't clear.

                    I understand that each time some separate individual rates a comment from 1:Troll through 4:Excellant, there will be an average provided with the number of separate ratings included in the average. 3.33/3 is clearly three ratings, one 4 and two 3's made by three people.

                    "None" is not a numerical rating, however. Yet I see ratings of None/0 and None/1. I have not seen a None/2 or higher denominator.

                    My question was what is the difference between None/0 and None/1? None is the default, that is, unrated.

                    It isn't a major problem, just a matter of mild curiosity. Does it have any meaning? Or is it an undocumented feature of the kind that MickySoft seems so happy to provide us with?

                    Democrats stand for Liberty, Security, Support of Families and Opportunity Whiskey Tango Foxtrot - over

                    by Rick B on Sun Jan 18, 2004 at 08:55:11 PM PDT

                    [ Parent ]

          •  Rate All only for some windows (none / 0)

            The Rate All box doesn't show up in the Dynamic Threaded setting but it does in Threaded (I can't get Dynamic Threaded to work on my work computer and still open inner comment threads without crashing the Windows Explorer [??]).

            since for the most part I consider these ratings a huge pissing contest.

            And don't knock the rating system. I rarely have the time to run through 147+ comments but when I know I have 15 minutes to check in it's nice to be able to scan quickly to the couple "hot spots."

            The polls don't tell us how a candidate is doing; they tell us how the media is doing.

            by Thumb on Sun Jan 18, 2004 at 06:41:19 PM PDT

            [ Parent ]

      •  Caveat (none / 0)

        That all being said, 130K plus 39K is a whole lot of caucus goers. So take it as you will.

        But it still gives you an idea of what can happen if the figures are true (and all those people show up).

        by ploeg on Sun Jan 18, 2004 at 04:42:29 PM PDT

        [ Parent ]

      •  Caucus Math (3.50 / 4)

        If Dean got 20% of a normal years' 130K, that would be 26 thousand, yes.  By your math, that 26 thousand voters would be only 40 % of the actual number committed to Dean, so the Dean total would actually be 65 thousand caucusers.  The additional 39K of voters would increase the caucus total to 169K.  Now 65K over 169K equals 38%!!  Incresing numerator and denominator one-for-one rapidly increases the ratio.

        If Dean got that proportion, wildly out of line with the polls, he would be a huge winner.  

        Even allowing for ploeg's comment below, it seems there is a real potential for Dean getting close to 30% of caucus voters at the end, not counting second choices people may make in some precincts.  

        Hmmm, I'm beginning to think there's a reason why the Dean camp is not worried by the polls.  Interesting electoral (caucus) math, indeed.

        by DenfromWisc on Sun Jan 18, 2004 at 05:19:03 PM PDT

        [ Parent ]

        •  re: Caucus Math (none / 0)

          > I'm beginning to think there's a reason why the Dean
          > camp is not worried by the polls.

          And moreover, having the polls indicate that the race is tight ensures that Dean's supporters come through tomorrow night. For example, if the weather is bad, you would have a lot of people who would say that, "Well, Dean's going to win anyway, so I don't need to brave the snow", and then Dean support doesn't translate into numbers tomorrow. But with the polls so close (whether or not that reflects actual support), few supporters will fail to fulfill their duty.

          by sg3000 on Sun Jan 18, 2004 at 05:49:07 PM PDT

          [ Parent ]

          •  Yeah.... (none / 0)

            and what I don't know is how I'm going to get thru the next 40 hours.  I'm practically tuned in full-time to Kos...

            by DenfromWisc on Sun Jan 18, 2004 at 06:01:11 PM PDT

            [ Parent ]

            •  Kos on computer, C-Span onTV (none / 0)

              except now.  I'm not gonna watch Republicans.  I'll come back to it at 3:30....

              by katiebird on Sun Jan 18, 2004 at 06:09:45 PM PDT

              [ Parent ]

              •  I have a nagging negative thought.... (none / 0)

                What if Dean wins the caucuses big, but the media focuses on just the final poll numbers.  They could peddle the Dean victory as a meaningless organizational victory, whereas the typical Iowans actually preferred someone else (edwards/kerry)?

                What ever happens, I'm not going to lose much sleep.  The first real test is still New Hampshire.  Next week is going to be nerve wracking as hell.

                Patriotic, flag waving, radical centrist Howard Dean Democrat. Until we stand on principle and lose our fear of defeat we will never win.

                by rusrivman on Sun Jan 18, 2004 at 06:32:26 PM PDT

                [ Parent ]

                •  Dean will keep going (none / 0)

                  Whatever happens, the Dean campaign will forge ahead. The campaign has outrun the media all along the way and it won't stop now

                  by gxschier on Sun Jan 18, 2004 at 06:44:27 PM PDT

                  [ Parent ]

                •  Doesn't matter (none / 0)

                  The caucuses are about showing off your organization, and finding the issues so that you can fine-tune things before the next rounds. Historically, candidates who win Iowa don't necessarily go on to win the nomination, and candidates who lose aren't killed unless they're hurting badly already.

                  That's why I'm kind of scratching my head about Clark giving Iowa a pass. Certainly he was late to the party and he's never done this before, but that's all the more reason to compete in the early rounds, so that he can get things ironed out. Next week, he'll be discovering the types of problems that the candidates competing this week will have detected and solved already. Well, we'll see how it works out for him.

                  by ploeg on Sun Jan 18, 2004 at 08:12:38 PM PDT

                  [ Parent ]

                •  Maybe not (none / 0)

                  > They could peddle the Dean victory as a meaningless
                  > organizational victory

                  That would be hard to do since all the news programs I've seen today are [gleefully?] showing Dean in 2nd or 3rd according to two specific polls.

                  So the news story, should Governor Dean win, would be how did he break from 2nd or 3rd into first place. Trying to tell your average non-news-junky that Dean won even though he was "behind" just a day before due to "organizational abilities" is pretty nuanced for most news programs to want to bother with. In fact, it's about as easy to describe as explaining why polls based on random number dialing are more accurate than polls taken from outdated Democratic party voter lists.

                  by sg3000 on Sun Jan 18, 2004 at 10:33:23 PM PDT

                  [ Parent ]

          •  Caucus rather than vote (none / 0)

            I think that if people show up at a caucus and there is a large group of Dean supporters, they are going to look around and see who is second. Then the people who came in intending to vote for the persons who turn out to be third or fourth will switch.

            If the Dean voters are mostly young true-believers, then my bet is that most of the switchers will go with number two, or with Gebhardt (if he is not number 2) because of familiarity.  

            A caucus is a lot more dynamic than a straight vote because everyone is there at once, and there is no secret about who is voting which way and who is getting how many votes. That's going to make the polls mostly useless in predicting votes other than Dean's.

            Similarly, a lot of people who have never turned out for a caucus before is going to make it difficult for pollsters to determine who is even a likely voter.

            My bet is that Gebhardt and Dean have the best organizations, and they will wind up splitting the top two groups of voters. The average age will be a great deal lower for Dean. Then they go into New Hampshire against Clark.

            Dean will then win NH based on organization, money and carry-over from his postion of number 1 or 2 in Iowa. Clark and Gebhardt will then split the older group of voters. I'll bet on Clark there. It will be a battle between Clark and Gebhardt for number two until March 2, and that will pretty much determine the Clark vs anti-Clark race.

            We are definitley look at interesting times.

            Democrats stand for Liberty, Security, Support of Families and Opportunity Whiskey Tango Foxtrot - over

            by Rick B on Sun Jan 18, 2004 at 07:31:20 PM PDT

            [ Parent ]

            •  Caucus dynamics (none / 0)

              I think that if people show up at a caucus and there is a large group of Dean supporters, they are going to look around and see who is second. Then the people who came in intending to vote for the persons who turn out to be third or fourth will switch.

              Rick B, have you been to an Iowa caucus? At the one I used to attend, everyone broke up into their groups at once - and people weren't allowed to come late. So it's not a matter of coming in and seeing groups already divided. You hear the call to divide, you go to your corner, and then you look around and see how big the other groups are.

              Respectful of Otters

              by Rivka on Sun Jan 18, 2004 at 08:18:55 PM PDT

              [ Parent ]

      •  I see your logic, but (none / 0)

        You make the assumption it seems that the 20% in the polls is 20% of a "normal turnout" of 130k caucus goers.

        The poll ranking of 20% was of those polled which was a very small number.

        I can certainly see the logic of a great number of new registrants not showing up on the list that is used to poll folks making the list not very relevant, but there were screening questions - especially in the DMRegister poll, that accounts for the distribution.

        I understand the difficulty of polling results for a caucus because of the dynamics of "viability" on a precinct by precinct basis, but that doesn't make the poll itself invalid.

        by Kathleen in CO on Sun Jan 18, 2004 at 06:34:11 PM PDT

        [ Parent ]

        •  Admittedly... (none / 0)

          my methodology is not airtight, and I accept all your criticisms gladly.

          If anything, though, I'd question the 60-65% first-timers figure. Numbers like that are unheard of. If they can deliver that, it would be a huge story.

          by ploeg on Sun Jan 18, 2004 at 08:04:11 PM PDT

          [ Parent ]

      •  Umm.... (none / 0)

        Hate to break everyone's bubble, but polls don't work off of ANY voter list. A long time ago they did, but these days to avoid problems with unlisted numbers they just randomly dial digits.

        So whatever the problems with the voter list the Iowa Dems are hawking, they won't be the source of error for the polls.

        by ColoDem on Sun Jan 18, 2004 at 08:46:50 PM PDT

        [ Parent ]

        •  We settled this earlier (none / 0)

          Polls like this have to work off of voter lists. Otherwise it's impossible to separate the (relatively small) pool of likely caucus goers from the (relatively large) pool of people with phone numbers.

          For example: the Iowa Poll:

          http://tinyurl.com/ytwbf

          The Iowa Poll, conducted for The Des Moines Register by Selzer & Co. Inc. of Des Moines, is based on interviews with 606 registered Iowa voters who say they definitely or probably will attend the Democratic caucuses. Interviewers contacted voters registered as Democrat, Green or no-party by using randomly selected telephone numbers from the Iowa secretary of state's voter registration list. Percentages based on the full sample may have a maximum margin of error of plus or minus 4 percentage points. Republishing the copyrighted Iowa Poll without credit to The Des Moines Register is prohibited.

          by ploeg on Sun Jan 18, 2004 at 09:10:55 PM PDT

          [ Parent ]

        •  LA Times is only IA poll to do RDD (none / 0)

          and nit had Dean with comfortable lead

          do we still have a Republic and a Constitution if our elected officials will not stand up for them on our behalf?

          by teacherken on Sun Jan 18, 2004 at 09:58:21 PM PDT

          [ Parent ]

View Story | 217 comments