Skip to main content

View Diary: Daily Kos Elections Live Digest: 5/13 (381 comments)

Comment Preferences

  •  Sabato: Repubican Wave Inevitable (17+ / 0-)

    (link)

    See, this is why I hate professional pundits. He defines a wave as Dems losing seats, but redefines it as "merely a ripple" if we, say, lose WV and SD.

    So let's go into the murky mind of Larry Sabato!

    In order for us to "only" lose AR, WV, SD, and MT, Obama needs to be at 50% approval and the Republicans need to commit several gaffes/Akin moments.

    In order for us to "only" lose AR, LA, MT, WV, and SD, Obama needs to stabilize and Republicans need to make a couple errors. This is deemed the most likely scenario.

    Why yes, Mr. Sabato, let's ignore all the polling that shows AR is actually our best chance at holding one of the Endangered Seven. Let's ignore the very real possibility of pickups in GA and KY, where our candidates are polling neck and neck with the opposition and both are outraising their likely opponents. In short, let's ignore reality to fit the Dems-in-Disarray theme the media has going for this cycle.

    It's not that I don't think it's possible we'll lose the Senate, it's just that the polling has shown it to be improbable. The most likely outcome at present isn't 50/50 control, but rather 52 or 53 to 48 or 47, where we lose MT, SD, and WV, but possibly gain one of KY or GA. That's what the polls say right now - and I'm not entirely sure when Sabato and Co. decided to become the Unskewed Polls (or, in this case, Ignored Polls) guys.

    TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

    by Le Champignon on Tue May 13, 2014 at 06:31:17 AM PDT

    •  Considering Obama is polling (5+ / 0-)
      Recommended by:
      Avenginggecko, LordMike, TofG, JBraden, askew

      at or near 50 in several polls now including Ras daily trackers, I'd say things aren't looking so bad after all.  Especially if you take Sabato's prognosis at face value.  Which of course I don't being that several polls now show Pryor ahead in AR and a few now have him ahead by double digits.  So that means Dems lose WV, SD, MT which most of us already expect.  However if Dems gain either KY, GA or both then the GOP wave will have resulted in a HUGE gain of +1-2 seats for a solid minority of 46-54 or 47-53.  Considering this was a tough cycle for Dems, I'd say the Dems will gladly take it and move on to 2016 when they show Sabato what a REAL wave will look like.  

      This is your world These are your people You can live for yourself today Or help build tomorrow for everyone -8.75, -8.00

      by DisNoir36 on Tue May 13, 2014 at 06:42:18 AM PDT

      [ Parent ]

      •  What's this? (11+ / 0-)

        There is one poll where Obama is at 50, and it's Rasmussen. In nearly every other poll he's been between 41 and 45 all year.

        I think you guys are whistling past the graveyard a little bit here. We're not behind in any of Alaska, Arkansas, Louisiana, Kentucky, Georgia, or North Carolina, but other than NC those are all very red states, and other than Alaska they're all very southern. This far from an election you can't just look at the polls - campaign dynamics, as well as the natural lean of these states, are more likely than not to assert itself over time.

        If we manage to keep a bare majority I'll be thrilled.

        •  What difference does it make? (3+ / 0-)
          Recommended by:
          DCCyclone, TofG, JBraden

          If they're red states? Polling is polling. I could just as easily say ... yeah, the R's are close in AR, AK, LA, and NC, but those are all incumbents with huge $$ advantages.

          •  This is the difference: (9+ / 0-)
            State partisanship. As Dan Hopkins wrote at FiveThirtyEight last week, races of all kinds have become more and more correlated with presidential results in recent years. So the Partisan Voting Index (PVI), which compares how a state voted in the past two presidential years against the national popular vote, is also a useful tool for congressional races. At this early point in the cycle, there’s reason to be skeptical of races where the polls are out of step with how the state usually votes; states often revert to their partisan mean once more voters engage with the campaign.
            That's from Nate Silver, whose predictions from late March still look basically right to me everywhere but Arkansas (where we've improved a bit) and Michigan (where I think we were always in better shape than it appeared).
            •  lol, Nate Silver (3+ / 0-)
              Recommended by:
              TofG, askew, WisJohn

              The man is the best example of a fall from grace that I've ever seen. Also, he too blew multiple Senate races.

              TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

              by Le Champignon on Tue May 13, 2014 at 08:23:50 AM PDT

              [ Parent ]

              •  What Senate Races Did He "Blow"? (0+ / 0-)

                Nobody I know has a perfect track record of predicting these things.  There was plenty of reason why even the most seasoned analyst would think MT and ND would have elected Republican Senators in 2012.

                •  ND (2+ / 0-)
                  Recommended by:
                  TofG, MichaelNY

                  he was sure the sate's natural political lean would give it to Berg even though many others believed that it was more about the candidates in ND in 12.s

                  Born NY-25, Raised TN-05, Worked NY-11 and NH-01 and NH-02, Live in DC Ward 6

                  by dsh17 on Tue May 13, 2014 at 08:36:23 AM PDT

                  [ Parent ]

                •  MT and ND were bad calls on his part (10+ / 0-)

                  because his algorithm kept calling them very likely Republican all the way up to election day, despite poll after poll showing them to be tossups.  Regular forecasters like Rothenberg had MT and ND as tossups, but Silver's model kept insisting they were GOP-favored (at around 65% and 80%, if I recall correctly).  For all the fanciness of his algorithm and percentages, regular people just sizing up a race outdid him.

                  •  People don't understand what Nate does (2+ / 0-)
                    Recommended by:
                    Mark27, wadingo

                    If his model says 65% Republican, that does not mean that the Republican is guaranteed to win.  It means the Republican, under given conditions, will win 2 out of 3 times.  It also means that the Democrat will win 1 out of 3 times.  As he explicitly states, upsets are bound to happen, and his model accounts for that.

                    •  I understand Nate's analysis just fine (5+ / 0-)

                      I know 80% doesn't mean "guaranteed".  But are we really supposed to believe that Heitkamp winning in North Dakota really was just a 1/5 shot that went our way?  Despite all the reputable polls showing the race to be a tossup?  Despite heavy spending by both sides indicating it was highly competitive?

                      Rothenberg and Cook had ND-Sen as a tossup and they were right.  It could have gone either way.  To have claimed 80% certainty is silly in hindsight, and saying "upsets happen" doesn't excuse poor forecasting and ignoring signs.

              •  He didn't blow anything (1+ / 0-)
                Recommended by:
                Mark27

                He specifically says that his model's percentages are supposed to account for upsets.  He is not in the same category as Sabato, Cook, etc.  He models the odds of things happening rather than predicting them.  

                •  Well that's convenient for him (3+ / 0-)
                  Recommended by:
                  aamail6, Skaje, MichaelNY

                  Because it means he can never be wrong.

                  TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

                  by Le Champignon on Tue May 13, 2014 at 10:35:57 AM PDT

                  [ Parent ]

                  •  Well, that's what a probability forecast is (1+ / 0-)
                    Recommended by:
                    Mark27

                    Let me give an example.

                    Let's say that Silver's final predictions for 2014 have Pryor, Landrieu and Hagan all at 66% for retaining their seats.  One might look at that and say the prediction is that all three will retain their seats, since they are all ahead.  But that's not what it's saying.

                    It means that each has a 2 in 3 chance of retaining their seat.  In other words, the "prediction", if you will, is that one of the three will lose, despite the fact that they're all "ahead" according to the model.

                    •  Hi, you're talking to an engineer.. (1+ / 0-)
                      Recommended by:
                      MichaelNY

                      I know what probability forecasts are. I also know their limitations. Silver's in the electoral prediction business. Two of the states where he claimed Republicans were favored ended up going to the Democrats. None of the Democrat-favored ones went Republican. Therefore, he likely has an error in his model.

                      If you can't evaluate the efficacy of your model based on real-world feedback, then your model is inherently unstable. It'd be like flying a plane without any windows or gauges. You can't say, "Oh, well, there was a small chance that the outcome wasn't what my model predicted, therefore my model still works" - otherwise I could easily assign a 99% probability that some random person would win a given race, and when that person doesn't win, I could simply point to that "small 1% chance".

                      It doesn't work that way. Nate's good with presidential elections, but sucks ass at Senate elections.

                      TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

                      by Le Champignon on Tue May 13, 2014 at 10:53:26 AM PDT

                      [ Parent ]

                      •  So the fact that he got two wrong (1+ / 0-)
                        Recommended by:
                        jncca

                        Makes him "suck ass"?  I'm also not sure what party has to do with anything.  So what if none of the D-favored seats went Republican?  It's not supposed to be even across party lines, it's supposed to be even overall.

                  •  Can't His Model Merely Be A Useful Tool..... (4+ / 0-)
                    Recommended by:
                    Trosk, wwmiv, jncca, wadingo

                    .....to analyze the probability of victory for a given race without making hard and firm predictions?  Is Nate Silver bad at what he does because his model showed Rick Berg would win ND-Sen, even when the overwhelming majority of analysts in the prediction business made the same call?  I think you need to put your own skin in this game and see how your predictions work out before sneering at somebody like Silver for not achieving 100% perfection with prior guesses.

                    •  why have models at all (3+ / 0-)
                      Recommended by:
                      Skaje, andgarden, MichaelNY

                      when prognosticators like Rothenberg and Cook, and some of us here, looking at the same polling, looking at the same historical leans of states, looking at fundraising, and hearing some of what is going on on the ground, are about as accurate?

                      ...to the philosopher it is iron and grain that made men civilized and brought on the downfall of the human race. - Rousseau, Discourse on the Origins of Inequality Among Men

                      by James Allen on Tue May 13, 2014 at 11:04:36 AM PDT

                      [ Parent ]

                      •  Kind Of Like Asking..... (3+ / 0-)
                        Recommended by:
                        Trosk, jncca, wadingo

                        .....why have Fords when there are Chevys?  I find it useful to have both.  If Silver's mathematical formulas really come up short at some point and blow a bunch of races, then they will be discredited and Cook, Rothenberg, and DKE old school analysis will persevere.  In the meantime I will enjoy seeing how well Silver's models up hold up.  

                      •  Because then you can be smug (1+ / 0-)
                        Recommended by:
                        MichaelNY

                        Nate Silver is none but a pundit. His opaque "model" is a hair more interesting that Tom Friedman's taxi driver.

                        Ok, so I read the polls.

                        by andgarden on Tue May 13, 2014 at 05:58:04 PM PDT

                        [ Parent ]

                •  And can his odds be wrong? (1+ / 0-)
                  Recommended by:
                  MichaelNY

                  That's what we're discussing here.  We all understand very well what odds are as opposed to predictions.  There's just a disagreement over whether his odds are particularly accurate.

                  •  I'd say his odds are pretty damn accurate (0+ / 0-)

                    Yes, ND-SEN was a fuck up.  I'd say MT-SEN was reasonable, but even if you think that was a fuck up too, he got every other Senate race right, along with every Presidential state.

                    •  But that's the thing (2+ / 0-)
                      Recommended by:
                      James Allen, MichaelNY

                      I also got every other senate race right because I'm a fan of polls. Why develop a model if it can't call the close races?

                      TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

                      by Le Champignon on Tue May 13, 2014 at 11:15:14 AM PDT

                      [ Parent ]

                      •  He's gotten 4 Senate races wrong since 2008 (0+ / 0-)

                        Two of them were pretty egregious - Nevada in 2010 and ND in 2012.  The other two, in my opinion, were acceptable.  He also got every state right in the 2012 Presidential election, and all but Indiana right in 2008.  Unless you're looking for 100% accuracy, which you'll find nowhere, what are you complaining about?

                    •  It's not hard to get the vast majority correct (6+ / 0-)

                      I predicted every Senate race correctly but Nevada (I thought Berkley would narrowly win).  It's really not hard to look at polling averages and fundamentals, and call pretty much everything correctly, especially when there aren't many tossups.  In the last presidential election, it didn't take a genius to see that Obama was probably going to win every competitive state except North Carolina.  Florida was the only really tough call.

                      Anyway, I dug up his final 2012 numbers, and he actually had ND-Sen at 92.5% chance of a GOP win, being less competitive than WV-Sen which was at 91.8% chance of a Dem win (Manchin won 61-37).  He also had NV-Sen at a 83.4% chance of a GOP win, and AZ-Sen at a 95.9% chance, pretty confident numbers considering how close both races turned out (Heller won by 1.2%, and Flake by 3%).

                      If his model really is so accurate, you'd expect the bad calls to be in both directions.  Instead, all of his bad odds leaned in the GOP's direction.

                      •  I wouldn't expect that given the map (0+ / 0-)

                        I don't see any 2012 Senate seats where a bad call in our direction was reasonable.  Nevada and Arizona were literally the only two possibilities, but those states were much less competitive than ND and MT.

                      •  Same here, in my last prediction (0+ / 0-)
                        I predicted every Senate race correctly but Nevada (I thought Berkley would narrowly win).
                        I ultimately came around on ND-Sen right near the end of the race.

                        Formerly Pan on Swing State Project

                        by MichaelNY on Tue May 13, 2014 at 11:00:01 PM PDT

                        [ Parent ]

            •  It's not just AR and MI (1+ / 0-)
              Recommended by:
              James Allen

              He's got OR and MN at 90%.  Really, a 1 out of 10 chance?  And if you do the math (1 - 0.9*0.9) that's a 19% chance of Republicans winning at least one of those two.

              I went and calculated the percentage chance of Republicans winning at least one of the safe Dem seats based on Nate's percentages (which is found by 1 minus the multiplied probabilities of Democrats winning each race) and got a 41% chance Republicans snag at least one seat out of MN, OR, NJ, VA, HI, MA, IL, NM, DE, or RI.

              If you add in Iowa and New Hampshire (both at just a 75% chance of Dem retention), the odds go up to 67% that Republicans take at least one of those Senate seats.

              •  Thats not how probability /really/ works (1+ / 0-)
                Recommended by:
                Trosk

                24 Burkean Post Modern Gay Democrat; NM-2 (Raised), TX-20 (B.A. & M.A. in Political Science), TX-17 (Home); 08/12 PVIs

                by wwmiv on Tue May 13, 2014 at 11:51:03 AM PDT

                [ Parent ]

                •  Assuming independent variables (0+ / 0-)

                  which is what I thought we were doing here.  If you roll a die three times and are trying to get a 6, the probability of getting it at least once is the 1 minus the total probability of never getting it at all, which is 5/6 for each roll.  For independent rolls (they don't influence each other) you just multiply the probabilities.  So 1 minus [(5/6)^3] which comes out to a 42% chance of rolling at least one 6 if you get three chances.

                  That's pretty much basic prob and what I was doing to Nate's numbers.  Unless you're making the argument that Senate races aren't independent from each other (an argument I am sympathetic to).

                  •  No (1+ / 0-)
                    Recommended by:
                    Skaje

                    I was under the impression (after quickly skimming your comment... I'm at work, after all) that you were adding, rather than multiplying.

                    But, there is also this:

                    Unless you're making the argument that Senate races aren't independent from each other (an argument I am sympathetic to).
                    I think we can agree on that point.

                    24 Burkean Post Modern Gay Democrat; NM-2 (Raised), TX-20 (B.A. & M.A. in Political Science), TX-17 (Home); 08/12 PVIs

                    by wwmiv on Tue May 13, 2014 at 12:46:26 PM PDT

                    [ Parent ]

              •  Um.... (1+ / 0-)
                Recommended by:
                wwmiv

                That's not how probabilities work.  You don't just add them up....

        •  They didn't in 2012 (3+ / 0-)

          "Natural lean" just doesn't apply like y'all think it does, at least not downballot and especially with incumbents. If, for instance, the polls are saying Pryor wins, then Pryor wins. Trust the polls. All the ones showing Heidi Heitkamp competitive despite ND being a "very red state" turned out to be accurate, no?

          TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

          by Le Champignon on Tue May 13, 2014 at 07:33:11 AM PDT

          [ Parent ]

        •  Another just came out today (2+ / 0-)
          Recommended by:
          TofG, MichaelNY

          with Obama at 48 approval in GA.  48 in any poll will likely be his high water mark in GA as he's averaged 48 for 5+ years now nationally (and much lower in red states).  If hes anywhere near that in red states then the Dems in those states have not only a fighting chance but a good shot at winning.  50% is frankly an artificial level and has little bearing on senate races anyway.  Its a brutal cycle for Dems.  Obama will never poll above 40% let alone 50% in many of those states but that doesn't mean the Dems will automatically lose those races because he doesn't poll at 50%.  The quality of the candidates in those states and their opponents will play a far bigger role than Obama does.  With the GOP it's far more likely that an Akin-like moment will have any effect on the races than Obama will as any effect he might have is already factored in.

          This is your world These are your people You can live for yourself today Or help build tomorrow for everyone -8.75, -8.00

          by DisNoir36 on Tue May 13, 2014 at 07:45:02 AM PDT

          [ Parent ]

    •  Sabato was once reality based (8+ / 0-)

      I'm not sure what's happaned to that guy in recent years.  Seems as though he's become more or less just another beltway hack.  He still has Kentucky-Senate rated as Likely R for some reason.  His projections these days seem even more conservative than Stu Rothenberg's.

    •  This is going to be the narrative until Election (10+ / 0-)

      day.

      Only a poor showing for the GOP in the election will disprove the narrative.

      It's not a perfect analogy, but I think this is similar to 1998, in that, a lot of pundits expected big GOP gains, in part because of the Lewinsky scandal, and they only re-evaluated after the GOP ended up doing poorly.

    •  Mark Halperin (4+ / 0-)
      Recommended by:
      James Allen, TofG, askew, MichaelNY

      I usually have Morning Joe on in the morning, mostly as background noise.  I did catch Mark Halperin saying as if he'd been to the future was was reporting back to us that Republicans are likely to take the Senate, still.  In which universe do these people live?

      Don't Blame Me, I Voted for Kodos.

      by MetroGnome on Tue May 13, 2014 at 06:58:52 AM PDT

      [ Parent ]

      •  If Mark halperin is predicting a GOP Senate (18+ / 0-)

        then I'm laying money on the Dems holding.  Halperin is one of the many beltway weenie munchers who has a Reverse Midas Touch.  I bet the stupid fucker STILL thinks McCain has a shot.

        HALPERIN: My hunch is this is going to end up being one of the worst moments in the entire campaign for one of the candidates but it’s Barack Obama.[...] I believe that this opened the door to not just Tony Rezko in that ad, but to bring up Reverend Wright, to bring up his relationship with Bill Ayers.
        That was him referring to McCain's blunder over his houses.  I think the phrase 'is great for McCain' was made in reference to this idiot's uncanny ability to turn everything from Sarah Palin, to the Russia-Georgia conflict into something that would be great for McCain.  This idiot even predicted that Obama would pick Lugar as a VP.  That was before calling PRESIDENT Obama a dick on TV.  Why anyone would ever listen to anything he has to say is beyond me.    He should stick to spouting bullshit about Death Panels.

        This is your world These are your people You can live for yourself today Or help build tomorrow for everyone -8.75, -8.00

        by DisNoir36 on Tue May 13, 2014 at 07:26:18 AM PDT

        [ Parent ]

    •  What's Sabato's track record? (3+ / 0-)
      Recommended by:
      jj32, MrLiberal, MichaelNY

      I'd rather judge him by the past effectiveness of his predictions than what we think his rate will be this year.

      My guess is that he's still mostly relying on fundamentals at this point, since it's really too early for polls to mean that much for Election Day.  Just look at Pryor.  In the minds of just about everyone here on DKE, he went from our most endangered incumbent to one of our safest in a matter of months.  Who's to say that AR won't become vulnerable again by the end of the summer?

      Let's not jump down the pundit's throats just yet.

      •  Pretty good (2+ / 0-)
        Recommended by:
        LordMike, MichaelNY

        "When dealing with terrorism, civil and human rights are not applicable." Egyptian military spokesman.

        by Paleo on Tue May 13, 2014 at 07:12:53 AM PDT

        [ Parent ]

      •  He blew ND and MT (10+ / 0-)

        last cycle.

        In 2010 in May, he had CA as tossup, CT as tossup, DE as Lean-R, NV as Lean-R, and WI as Likely-D. In November, it was CA as Lean-D (lol, she won by 10%), CO as Lean-R, CT as Lean-D (he ended up winning by 12%), DE as Likely-D, NV as Lean-R (Reid won by 7%), and so forth. No, I wouldn't say he has a great track record.

        TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

        by Le Champignon on Tue May 13, 2014 at 07:24:52 AM PDT

        [ Parent ]

        •  2012 is understandable (6+ / 0-)

          Getting two wrong isn't bad, particularly ND;

          But I didn't realize he did so badly in 2010.  I don't care what his May predictions were, but for November, that's pretty unbelievable.

          That said, I still think it's too early to be judging people's predictions unless they're really off the reservation.  Too much can change between now and November.

        •  In other words, he stunk (3+ / 0-)
          Recommended by:
          TofG, askew, MichaelNY

          Most of the regulars on this site would have done better.

        •  Sabato got rid of Tossups before election day (6+ / 0-)

          It's debatable whether that's a good idea or not, but it's not like he was insanely bullish on Colorado and Nevada. If you get rid of tossups and had to pick one party to take each seat, it was completely defensible to guess that the GOP would win.

          And for what it's worth, the Swing State Project also had California and Connecticut as Lean Dem in November. Sabato's final predictions were on the high end for the GOP but they weren't ridiculous.

          Contributing Editor, Daily Kos Elections. 24, male, CA-18 (home and voting there), LA-02 (resident).

          by Jeff Singer on Tue May 13, 2014 at 08:15:13 AM PDT

          [ Parent ]

        •  2010 ratings (2+ / 0-)
          Recommended by:
          lordpet8, MichaelNY

          We had CA and CT at Lean, but CO and NV at Tossup. What were your ratings?

          Get the Daily Kos Elections Digest in your inbox every weekday. Sign up here.

          by David Nir on Tue May 13, 2014 at 11:15:14 AM PDT

          [ Parent ]

          •  Wasn't in the business back then (1+ / 0-)
            Recommended by:
            MichaelNY

            But that might be a fault in your analysis if you seriously believed CA was competitive. After all, you still have AR at a tossup, tilting in the Republican direction. No rational basis exists for such a call, just like CA in 2010.

            TX-17 (Bill Flores-R), TX Sen-14 (Kirk Watson-D), TX HD-50 (Celia Israel-D). Senate ratings map (as of 3/10/14)

            by Le Champignon on Tue May 13, 2014 at 01:51:52 PM PDT

            [ Parent ]

            •  The polls showed CA-Sen competitive (3+ / 0-)
              Recommended by:
              jncca, Darth Jeff, MichaelNY

              Only a single pollster (and a disreputable one at that) actually showed Boxer hitting the 10 point win she eventually got.  PPP's final poll had Boxer up 4.  CNN had her up 5.  SUSA and Rasmussen had released polls earlier showing Boxer losing.  They were all off, but a Lean Dem rating wasn't ridiculous.  Going with Likely Dem could have been justified (I probably would have rated it that) but I know my ratings tend to be optimistic, and I wouldn't knock anyone who put it Lean Dem out of caution.

              As for AR, there is plenty of rational basis to keep it at a tossup.  A few good polls came out for our side, but the fundamentals haven't changed.  We have an incumbent with moderate approval ratings trying to win re-election against a credible opponent in a rapidly reddening state.

              In any case.  Arguing with SSP/DKE's official ratings is somewhat of a tradition for us commenters, but we got to recognize they've done pretty well.

            •  Don't care for your attitude (0+ / 0-)

              It's one thing to disagree. It's another to sneer that people you differ with have "no rational basis" for their opinions. I actually really like when people question our ratings, because it forces us to consider alternate points of view. But this kind of commentary is not helpful in the least.

              Get the Daily Kos Elections Digest in your inbox every weekday. Sign up here.

              by David Nir on Tue May 13, 2014 at 05:28:01 PM PDT

              [ Parent ]

        •  hmm (1+ / 0-)
          Recommended by:
          MichaelNY

          Was Castle still in the running for senate in May? If so lean-R seems more optimistic for Dems than what I would have rated that race.

          "It is difficult to get a man to understand something when his salary depends on him not understanding it" - Upton Sinclair

          by lordpet8 on Tue May 13, 2014 at 01:04:26 PM PDT

          [ Parent ]

    •  They are moving the goalposts as expected (9+ / 0-)

      Does damage to a lighthouse on an outcrop of rock half way out in the ocean constitute a wave on the mainland?

      "What do you mean "conspiracy"? Does that mean it's someone's imaginings and that the actual polls hovered right around the result?" - petral

      by conspiracy on Tue May 13, 2014 at 09:36:09 AM PDT

      [ Parent ]

    •  Early predictions are often too early to predict. (1+ / 0-)
      Recommended by:
      MichaelNY

      And that is what makes them as well as us look ridiculous when we do it.  Though of course we still do because it's fun.  And even if we're correct on something it is often just happenstance rather than genius.  IE I could take credit for saying I thought Feingold would find himself in trouble in 2010 because of his past tendency to almost blow races with his campaign style even though he was running against a bunch of "some dudes."  But I really knew not a hell of a lot about Ron Johnson just like everyone else.  I also predicted Evan Bayh would be in trouble and was generally in a "sky is falling" mindset.  

      Now at the time the polls were terrible.  But there were some contradictory polls saying things might not be so bad.  Feingold's polls weren't that bad and those that showed him in trouble did have things suspect about them.  Sometimes things turn out better than expected.  For example I don't think anyone really saw the Senate gains we got in 2000 coming even if late polling showed it within the realm of possibility.

      What really makes polling difficult particularly this early is coming up with a turnout model.  You get that wrong and you've skewed it hopelessly from the start.  You only begin to get a handle on what it'll be after labor day and even then it's tough.

      But the fact is anyone who makes early predictions is going to look foolish.  Folks like Sabato, Cook, and Rothenberg just come off more silly because they are so much more high profile than of us.  And while I never had much regard for Rothenberg I don't think Sabato or Cook's track record is that bad.  In fact I can think of a few times we've called out Cook and he turned out to be correct.

      The lady was enchanted and said they ought to see. So they charged her with subversion and made her watch TV -Spirogyra

      by Taget on Tue May 13, 2014 at 11:02:30 AM PDT

      [ Parent ]

      •  They aren't really predictions though (1+ / 0-)
        Recommended by:
        MichaelNY

        They aren't claiming anything other than the state of the race today. The reason we mocked Rothenberg for 2010 was because he was so certain Democrats couldn't lose the House until later.

        "What do you mean "conspiracy"? Does that mean it's someone's imaginings and that the actual polls hovered right around the result?" - petral

        by conspiracy on Tue May 13, 2014 at 11:47:00 AM PDT

        [ Parent ]

    •  I'd focus more on the actual number of seats (1+ / 0-)
      Recommended by:
      MichaelNY

      gained or lost more than the specific seat at stake.Also he's writing this for Politico so if anything I'd focus more on what he writes for his actual site instead.

      Don't get me wrong GA & KY look good for team blue but SD and WV look even better for the GOP.

      "It is difficult to get a man to understand something when his salary depends on him not understanding it" - Upton Sinclair

      by lordpet8 on Tue May 13, 2014 at 12:29:27 PM PDT

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site