There’s a very large gulf between my conclusion, explained on Friday, that Obama is toast on Election Day and confident projections like Nate Silver’s poll-reading model still giving the president (at last check) a 77.4% chance of victory. Let me explain why, and what that says about the difference between my approach and Nate’s.
The Limits of Mathematical Models
“A page of history is worth a volume of logic”
– Oliver Wendell Holmes
Mathematical models are all the rage these days, but you need to start with the most basic of facts: a model is only as good as the underlying data, and that data comes in two varieties: (1) actual raw data about the current and recent past, and (2) historical evidence from which the future is projected from the raw data, on the assumption that the future will behave like the past. Consider the models under closest scrutiny right now: weather models such as hurricane models. These are the best kind of model, in the sense that the raw data is derived from intensive real-time observation and the historical data is derived from a huge number of observations and thus not dependent on a tiny and potentially unrepresentative sample.
Yet, as you watch any storm develop, you see its projected path change, sometimes dramatically. Why? Because the models are highly sensitive to changes in raw data, and because storms are dynamic systems: their path follows a certain logic, but does not track a wholly predictable trajectory. The constant adjustments made to weather models ought to give us a little more humility in dealing with models that suffer from greater flaws in raw data observations, smaller sample sizes in their bases of historical data, or that purport to explain even more complex or dynamic systems – models like climate modeling, financial market forecasts, economic and budgetary forecasting, or the behavior of voters. Yet somehow, liberals in particular seem so enamored of such models that they decry any skepticism of their projections as a “War on Objectivity,” in the words of Paul Krugman. Conservatives get labeled “climate deniers” or “poll deniers” (by the likes of Tom Jensen of PPP, Markos Moulitsas, Jonathan Chait and the American Prospect) or, in the case of disagreeing with budgetary forecasts that aren’t really even forecasts, “liars.” But if history teaches us anything, it’s that the more abuse that’s directed towards skeptics, the greater the need for someone to play Socrates.
Consider an argument Michael Lewis makes in his book The Big Short: nearly everybody involved in the mortgage-backed securities market (buy-side, sell-side, ratings agencies, regulators) bought into mathematical models valuing MBS as low-risk based on models whose historical data didn’t go back far enough to capture a collapse in housing prices. And it was precisely such a collapse that destroyed all the assumptions on which the models rested. But the people who saw the collapse coming weren’t people who built better models; they were people who questioned the assumptions in the existing models and figured out how dependent they were on those unquestioned assumptions. Something similar is what I believe is going on today with poll averages and the polling models on which they are based. The 2008 electorate that put Barack Obama in the White House is the 2005 housing market, the Dow 36,000 of politics. And any model that directly or indirectly assumes its continuation in 2012 is – no matter how diligently applied – combining bad raw data with a flawed reading of the historical evidence.
Different sets of polls are, more or less, describing two alternate universes in terms of what the 2012 electorate will look like, one strongly favorable to Obama, one essentially decisive in favor of Romney. The pro-Obama view requires a number of things to happen that are effectively unprecedented in electoral history, but Nate Silver argues that we should trust them because state poll averages have a better track record in other elections than national polls. The pro-Romney view, by contrast, simply assumes that things have gone wrong in a number of the polls’ samples that have gone wrong before. Sean Trende argues that the national pollsters currently in the field are more reliable, and that this (rather than the history of state and national pollsters in the abstract) should be significant:
Among national pollsters, you have a battle-tested group with a long track record performing national polls. Of the 14 pollsters producing national surveys in October, all but three were doing the same in 2004 (although AP used Ipsos as its pollster that year rather than GfK, and I believe a few others may have changed their data-collection companies). Of the 14 pollsters surveying Ohio in October, only four did so in 2004 (five if you count CNN/USAToday/Gallup and CNN/Opinion Research as the same poll).
Pollsters such as ABC/Washington Post, Gallup, Pew, Battleground, and NBC/WSJ are well-funded, well-staffed organizations. It’s not immediately obvious why the Gravises, Purple Strategies and Marists of the world should be trusted as much as them, let alone more. And since virtually none of the present state pollsters were around in 1996 or 2000 (except Rasmussen Reports, which had a terrible year in 2000 and has since overhauled its methodology), it’s even less clear why we should now defer to state poll performance based upon those years.
In my opinion, which view is correct is not one that can be resolved by mathematical models, but rather by an examination of the competing assumptions underlying the two sets of polls and an assessment of their reasonableness in light of history and current political reality.
Where Polls Come From
Polling is “scientific,” in the sense that it attempts to follow well-established mathematical concepts of random sampling, but political polls remain as much art as science, and each polling cycle presents different challenges to pollsters’ ability to accurately capture public sentiment. Quick summary: dating back roughly to George Gallup’s introduction of modern political polling in the 1936 election, a pollster seeks to extrapolate the voting behavior of many millions of people (130 million people voted in the 2008 presidential election) from a poll of several hundred or a few thousand people. In a poll that seeks only the opinion of the public at large, the pollster will seek to use a variety of sampling techniques to ensure that the population called actually matches the population as a whole in terms of age, gender, race, geography and other demographic factors. In some cases, where the raw data doesn’t provide a random sample, the pollster may re-weight the sample to reflect a fair cross-section.
Political polling is a somewhat different animal, however: not all adults are registered voters, and not all registered voters show up to vote every time there’s an election. So, a pollster has to use a variety of different methods – in particular, a “likely voter” screen designed to tease out the poll respondent’s likelihood of voting – to try to figure out whether the pollster’s results have sampled a group of people who correspond to the actual electorate for a given election. This is complicated by the fact that voter turnout isn’t uniform: in some years and some states Republican enthusiasm is higher than others, in some Democratic enthusiasm is higher than others. You can conduct the best poll in the world in terms of accurately ascertaining the views of a population that mirrors your sample – but if your sample doesn’t mirror that season’s electorate, your poll will mislead its readers in the same way that the Literary Digest’s unscientific poll did in 1936, or the RCP averages in the Senate elections in Colorado and Nevada in 2010, or the polls that failed to capture the GOP surge in 2002.
Technology, economics and other factors affect polling. The rise of caller ID in particular has dramatically reduced response rates – that is, pollsters have to call 8 or 9 people for every one who will answer their poll. That raises the level of difficulty in ensuring that the people who actually do answer the questions are a representative sample. Liberals argue that pollsters undersample people who have only cell phones (a disproportionately younger and/or poorer group) and non-English speakers; conservatives counter that Tea Partiers may be less likely to talk to pollsters and that polls in some cases can suffer a “Shy Tory Factor” where voters are less likely to admit to voting Republican. Partisans dispute the relative merits of in-person versus automated polling and the structure of polls that ask a lot of leading questions before asking for voter preferences. And the economics of the polling business itself is under stress, as news organizations have less money to spend on polls and pollsters do public political polling for a variety of business reasons, only some of which have anything to do with a desire to be accurate – some pollsters like PPP make most of their money off serving partisan clients, news organizations do it to drive news, universities do it for name recognition.
2012, even moreso than past elections, is apt to produce another round of reflection and recrimination on all of these issues, as a great many of the individual polls we have seen so far have been largely or wholly irreconcilable, especially in terms of their view of the partisan makeup of the 2012 electorate. If you assume that (1) the various players in national and state polling have essentially random tendencies towards inaccuracy in modeling the electorate in all conceivable environments and (2) each state’s poll average includes a large enough sample of different polls by different pollsters to bear out this assumption – in that case, state polling averages and the models that rest on them should be good predictors of turnout, as they have been in most (but not all) past elections. But when you consider that 2008 was a very unusual environment and that every turnout indicator we have other than the state poll averages is pointing to a different electorate, these become far more questionable assumptions.
Toplines and Internals
Nate Silver’s much-celebrated model is, like other poll averages, based simply on analyzing the toplines of public polls. This, more than any other factor, is where he and I part company.
If you read only the toplines of polls – the single number that says something like “Romney 48, Obama 47” – you would get the impression from a great many polls that this is a very tight race nationally, in which Obama has a steady lead in key swing states. In an ordinary year, the toplines of the polls eventually converge around the final result – but this year, there seems to be some stubborn splits among the poll toplines that reflect the pollsters’ struggles to come to agreement on who is going to vote.
Poll toplines are simply the sum of their internals: that is, different subgroups within the sample. The one poll-watchers track most closely is the partisan breakdowns: how each candidate is doing with Republican voters, Democratic voters and independent voters, two of whom (the Rs & Ds) have relatively predictable voting patterns. Bridging the gap from those internals to the topline is the percentage of each group included in the poll, which of course derives from the likely-voter modeling and other sampling issues described above. And therein lies the controversy.
My thesis, and that of a good many conservative skeptics of the 538 model, is that these internals are telling an entirely different story than some of the toplines: that Obama is getting clobbered with independent voters, traditionally the largest variable in any election and especially in a presidential election, where both sides will usually have sophisticated, well-funded turnout operations in the field. He’s on track to lose independents by double digits nationally, and the last three candidates to do that were Dukakis, Mondale and Carter in 1980. And he’s not balancing that with any particular crossover advantage (i.e., drawing more crossover Republican voters than Romney is drawing crossover Democratic voters). Similar trends are apparent throughout the state-by-state polls, not in every single poll but in enough of them to show a clear trend all over the battleground states.
If you averaged Obama’s standing in all the internals, you’d capture a profile of a candidate that looks an awful lot like a whole lot of people who have gone down to defeat in the past, and nearly nobody who has won. Under such circumstances, Obama can only win if the electorate features a historically decisive turnout advantage for Democrats – an advantage that none of the historically predictive turnout metrics are seeing, with the sole exception of the poll samples used by some (but not all) pollsters. Thus, Obama’s position in the toplines depends entirely on whether those pollsters are correctly sampling the partisan turnout.
That’s where the importance of knowing and understanding electoral history comes in. Because if your model is relying entirely on toplines that don’t make any sense when you look at the internals with a knowledge of the past history of what winning campaigns look like, you need to start playing Socrates.
Moneyball and PECOTA’s World
Let me use an analogy from baseball statistics, which I think is appropriate here because it’s where both I and Nate Silver first learned to read statistics critically and first got an audience on the internet: in terms of their predictive power, poll toplines are like pitcher win-loss records or batter RBI.
At a very general level, the job of a baseball batter is to make runs score, and the job of a baseball pitcher is to win games, so traditionally people looked at W-L records and RBI as evidence of who was good at their jobs. And it’s true that any group of pitchers with really good W-L record will, on average, be better than a group with bad ones; any group of batters with a lot of RBI will, on average, be better than a group with very few RBI. If you built a model around those numbers, you’d be right more often than you’d be wrong.
But wins and RBI are not skills; they are the byproducts of other skills (striking people out, hitting home runs, etc.) combined with opportunities: you can’t drive in runners who aren’t on base, and you can’t win games if your team doesn’t score runs. If you build your team around acquiring guys who get a lot of RBI and wins, you may end up making an awful lot of mistakes. Similarly, you can’t win the votes of people who don’t come to the polls.
Baseball analysis has come a long way in recent decades, because baseball is a closed system: nearly everything is recorded and quantified, so statistical analysis is less likely to founder on hidden, uncounted variables. Yet, even highly sophisticated baseball models can still make mistakes if they rest on mistaken assumptions. Baseball Prospectus.com’s PECOTA player projection system – designed by Nate Silver and his colleagues at BP – is one of the best state-of-the-art systems in the business. But one of PECOTA’s more recent, well-known failures presents an object lesson. In 2009, PECOTA projected rookie Orioles catcher Matt Wieters to hit .311/.395/.546 (batting/on base percentage/slugging). As regular consumers of PECOTA know, this is just a probabilistic projection of his most likely performance, and the actual projection provided a range of possible outcomes. But the projection clearly was wrong, and not just unsuccessful. While Wieters has developed into a good player, nothing in his major league performance since has justfied such optimism: Wieters hit .288/.340/.412 as a rookie, and .260/.328/.421 over his first four major league seasons. What went wrong? Wieters had batted .355/.454/.600 between AA and A ball in 2008, and systems like PECOTA are supposed to adjust those numbers downward for the difference in the level of competition between A ball, AA ball and the major leagues. But as Colin Wyers noted at the time, the problem was that the context adjustments used by PECOTA that season used an unusually generous translation, assuming that the two leagues Wieters had played in – the Eastern League and the Carolina League – were much more competitive in 2008 than they had been in previous years. By getting the baseline of the 2008 environment Wieters played in wrong, PECOTA got the projection wrong, a projection that was out of step with what other models were much more realistically projecting at the time. The sophistication of the PECOTA system was no match for two bad inputs in the historical data.
My point is not to beat up on PECOTA, which as I said is a fantastic system and much better than anything I could design. Let’s consider for a further example one of PECOTA’s most notable successes, one where I questioned Nate Silver at the time and was wrong; I think it also illustrates the differing approaches at work here. In 2008, PECOTA projected the Tampa Bay Rays to win 88-89 games, a projection that Nate Silver touted in a widely-read Sports Illustrated article. It was a daring projection, seeing as the Rays had lost 95 or more games three years running and never won more than 70 games in franchise history. As Silver wrote, “[i]t’s in the field…that the Rays will make their biggest gains…the Rays’ defense projects to be 10 runs above average this year, an 82-run improvement.” I wrote at the time: “this is nuts. Last season, Tampa allowed 944 runs (5.83 per game), the highest in the majors by a margin of more than 50 runs. This season, BP is projecting them to allow 713 runs (4.40 per game), the lowest in the AL, third-lowest in the majors…and a 32% reduction from last season…it’s an incredibly ambitious goal.”
PECOTA was right, and if anything was too conservative. The Rays won 97 games and went to the World Series, without any improvement by their offense, almost entirely on the strength of an improved defense. I later calculated that their one-year defensive improvement was the largest since 1878. Looking at history and common sense, I was right that PECOTA was projecting an event nearly unprecedented in the history of the game, and I would raise the same objection again. But the model was right in seeing it coming.
If you looked closely, you could see why: the frontiers of statistical analysis had shifted. Michael Lewis’ book Moneyball, following the 2002 Oakland A’s, captured the era when statistical analysts stressed hitting and de-emphasized fielding on the theory that it was easier to use sophisticated metrics to find better hitters, but harder to quantify the benefits of defense. By 2008, the metrics were creating more opportunities to study defense, and – as captured in Jonah Keri’s book The Extra 2% (about the building of that Rays team) – the Rays took advantage.
But for the Rays, the 2008 environment was not so easily repeated in subsequent years. While still a successful club with a solid defense in a pitcher’s park (and still far better defensively than in 2007) they have led the league in “Defensive Efficiency Rating” only once in the past four years. It’s what Bill James called the Law of Competitive Balance: unsuccessful teams adapt more quickly to imitate the successes of the successful teams, bringing both sides closer to parity. Trende, in his book The Lost Majority, applies the same essential lesson to political coalitions. Assuming that the 2008 turnout models, which depended heavily on unusually low Republican turnout, still apply to Obama’s current campaign ignores the extent to which multiple factors favor a balance swinging back to the Republicans. And the polls that make up the averages – averages upon which Nate Silver’s model rests – are doing just that. Nate’s model might well work in an election where the relationship between the internals and the toplines was unchanged from 2008. But because that assumption is an unreasonable one, yet almost by definition not subject to question in his model, the model is delivering a conclusion at odds with current, observable political reality.
Painted Into A Corner
Poll analysis by campaign professionals often involves a large dollop of conscious partisan hackery: spinning the polls to suggest a result the campaigns know is not realistic, in the hopes of avoiding the bottom-drops-out loss of voter confidence that sets in when a campaign is visibly doomed. For the record, unlike some of my conservative colleagues, I don’t think Nate is a conscious partisan hack. I have a lot of respect for his intelligence and his thoroughness as a baseball analyst and we have mutual friends in the world of baseball analysis, and I think he undoubtedly recognizes that it will not be good for his credibility to be committed to the last ditch to defending Obama as a prohibitive favorite in an election he ends up losing. (It’s true that the 538 model is just probabilities, but as Prof. Jacobson notes, Nate won his reputation as an electoral forecaster with similar probabilistic projections in 2008; if you project a guy to have a 77% chance to win an election he loses, that will inevitably cause people to put less faith in your odds-laying later on).
I do, however, think that – for whatever reasons – Nate has painted himself into a corner from which there is no easy escape. If I’m right about the electorate and the polls are right about the internals, Romney wins – and if Romney wins, the 538 model will require some serious rethinking. There’s a bunch of reasons why he finds himself in this position. One is that his model has been oversold: he made his poll-reading reputation based on a single election cycle, in which he had access to non-public polls to check his work. Nate is, in fact, not the first poll-reader to get 49 states right: RedState’s own Gerry Daly did the same thing in 2004, missing only Wisconsin (which Bush lost by half a point) in his Election Day forecast, and Gerry did this through careful common-sense reading of the state-by-state polls checked against the national polls, not through a model that purported to do his thinking for him. (As it happened, the RCP averages at the end of the cycle did the same thing, as they did in 2008.) I’m inclined to listen to guys like Gerry who have been doing this for years and have not only recounted the numbers from past elections but lived through the reading of polls while they were happening. In 2010, the 538 model fared well – but no better than the poll averages at RCP. And that was only after Nate was much slower to pick up on the coming GOP wave than Scott Rasmussen, who called it a lot earlier in the cycle.
There are a raft of methodological quibbles with the 538 model (some larger than others), many of which reek of confirmation bias (ie, the tendency to question bad news more closely than good). For example, while Nate’s commentaries have included lengthy broadsides against Rasmussen and Gallup, his model tends to give a lot of weight to partisan pollster PPP. Ted Frank noted one example that perfectly captures the value of knowing your history; the 538 model’s assumptions about how late-deciding undecided voters will break are tilted towards Obama by including the 2000 election, when Gore did far better on Election Day than the late-October polls suggested. But Gore wasn’t an incumbent, and there was a major event (the Bush DUI story) that had a major impact on turnout and undecided voters. If you make different assumptions based on a different reading of history, you get different conclusions. The spirit of open scientific inquiry should welcome this kind of scrutiny, even in the heat of election season.
None of this is a reason to conclude that the 538 concept is broken beyond repair. If you regard poll analysis as something like an objective calling, you can learn from your failures as well as your successes. If Obama wins, my own assumptions (and indeed, nearly everything we know about winning campaigns) will have to be re-examined. If Romney wins, the model of simply aggregating the topline state-by-state poll averages will have to be sent back to the drawing board. But there will be no hiding, in that case, from the fact of its failure.
One of the more widely-discussed efforts to fix the problem of topline poll data varying by turnout models is Dean Chambers’ UnskewedPolls.com, which takes the internals of each poll and re-weights them for a more Romney-friendly turnout model. In concept, what Chambers is doing is on the right track, because it lets us separate how much of the poll toplines is due to the sentiments of different groups and how much is due to assumptions about turnout. But his execution is a methodological hash.
I haven’t pulled apart all the pieces of Chambers’ model, but my main objection to UnskewedPolls is that it re-weights the electorate twice:
The QStarNews poll works with the premise that the partisan makeup of the electorate 34.8 percent Republicans, 35.2 percent Democrats and 30.0 percent independent voters. Additionally, our model is based on the electorate including approximately 41.0 percent conservatives, 20.0 percent moderates and 39.0 percent liberals.
Republicans are 89 percent conservative, 9 percent moderate and 2 percent liberal. Among Democrats, 3 percent are conservative, 23 percent are moderate and 74 percent are liberal. Independents include 33 percent conservatives, 49 percent moderates and 18 percent liberals.
Our polls are doubly-weighted, to doubly insure the results are most accurate and not skewed, by both party identification and self-identified ideology. For instance, no matter how many Republicans answer our survey, they are weighted at 34.8 percent. If conservatives are over-represented among Republicans in the raw sample, they are still weighted at 89 percent of Republicans regardless.
The problem with this method is that neither the raw data (the current polls) nor a lot of the historical data (past years’ exit polls) has crosstabs showing how the votes of each partisan group break out by ideology. That is, for example, we have nearly no separate polling (certainly none on the polls Chambers is “unskewing”) showing how Romney is polling among independents who self-identify as moderates, or how Obama is polling among Democrats who self-identify as conservatives. That’s aside from the question of whether ideological self-ID is nearly as predictive a variable as party ID. Re-weighting the samples twice by these two separate variables, without access to those crosstabs, means you don’t really have any idea whether you are just adding a mutiplier that double-counts your adjustments to the turnout model. It’s more alchemy than science.
We can’t know until Election Day who is right. I stand by my view that Obama is losing independent voters decisively, because the national and state polls both support that thesis. I stand by my view that Republican turnout will be up significantly from recent-historic lows in 2008 in the key swing states (Ohio, Wisconsin, Colorado) and nationally, because the post-2008 elections, the party registration data, the early-voting and absentee-ballot numbers, and the Rasmussen and Gallup national party-ID surveys (both of which have solid track records) all point to this conclusion. I stand by my view that no countervailing evidence outside of poll samples shows a similar surge above 2008 levels in Democratic voter turnout, as would be needed to offset Romney’s advantage with independents and increased GOP voter turnout. And I stand by the view that a mechanical reading of polling averages is an inadequate basis to project an event unprecedented in American history: the re-election of a sitting president without a clear-cut victory in the national popular vote.
Perhaps, despite the paucity of evidence to the contrary, these assumptions are wrong. But if they are correct, no mathematical model can provide a convincing explanation of how Obama is going to win re-election. He remains toast.
97 thoughts on “On Polling Models, Skewed & Unskewed”
Terrific post. If I may overly simplify – it’s a difference between purely quantitative and qualitative political scientists. The former almost slavishly adhere mere to statistical models, while the former utilize them in an order to arrive at more informed opinions. Numbers should be the start of social research – not the start AND end.
If the election plays out as Silver predicts, will that impact your analysis of his methodology?
Jimmy, I’ll take a stab at your question. I agree with Dan substantially enough, and he referenced me, so why not?
I’ll put it in familiar terms, although I am making up the number. What if I was to say, I think that Nate’s model will be accurate (at least in a meaningful fashion) in 78.3% of elections, and as such I think he is not being quite transparent in conveying to his audience the limitations of which he, being of Sabermetric background, undoubtedly is all too well aware?
What if that number is 70%? 50%? 40%? From how big of a sample size of elections, complete elections which are their own system, do we have to estimate that accuracy rate? Not too many.
Models are great- but limitations must be kept in mind always. And each model will have their own limitations. Models for Presidential elections have a big problem in limited data sets on which to construct the models, infrequent opportunities to test and refine the models, and inputs subject to high errors of measurement.
Does that mean, ignore polls? Hell no! They are a great tool for seeing certain things. But recognize their limitations, and understand that models that are based on them have similar limitations.
Thanks for the response Gerry and I suspect that Crank would agree with you. Now what logically follows from that is, given the relative few possible applications of Silver’s model, is that the sample size argument is going to be there for as long as any of us are alive. And if that’s the road we’re going down, then aren’t discussions about the value of the model ultimately pointless? If Silver nails the November results perfectly, you and Crank are going to be able to dismiss him entirely, opening the door to whatever post facto analysis suits your needs. And if he’s wrong, people who believe in Silver will either find solace in the fact that just like a .100 hitter can come through 1/10 of the time, the twenty-something percent chance that Romney will win that Silver allows for is a scenario that played out. Or they will have faith that Silver will tweak his model and get things right in 2012.
So where does that leave us? Well, for the purposes of this blog, it leaves me waiting for the next post or, even better, hoping Crank revisits his previous post and answers the questions still waiting for him in the comments.
“is that the sample size argument is going to be there for as long as any of us are alive. And if that’s the road we’re going down, then aren’t discussions about the value of the model ultimately pointless?”
To me, yes. And the primary reason for my answer of “no” above. I believe, at least in terms of Presidential elections, that none of us will see a large enough data set for which us to know the accuracy rate of the model with any certainty.
That doesn’t mean I don’t think they are of use. I think they are interesting, and can lead to one having a good sense for where to look for other interesting things. Heck, I use a model of my own construction, just for kicks and to point me to things I might be missing.
One more comment.
“If Silver nails the November results perfectly, you and Crank are going to be able to dismiss him entirely”
Well, as I see it, since Nate is couching things in probability, if he gets this election wrong, he can step back and say “I said 73%” or 64% or whatever number is the final one. So his assertions, without that large sample of iterations, is unfalsifiable.
I also am aware that my perspective is also unfalsifiable with regards to 538, at least within our lifetimes.
I guess it depends on if one thinks what I wrote made sense or not. I think it does. Caveat lector.
I’m kind of hitching on to some of Gerry’s points here, so not claiming to be saying something new, but I think this needs to be said over and over until people finally understand. Nate Silver is not predicting that Obama will win. He’s not predicting anything. He only says Obama has a ~75% chance of winning the electoral vote.
That’s the “defending Silver” part of my post. The other part of my post acknowledges the counter-argument. Of course he could just come up with random numbers and never be proven wrong in one election. Which brings me to my understanding of the situation:
– It would take lots and lots of elections to prove or disprove his model.
– Nate really doesn’t have a track record, yet. He did well in 2008 and 2010, although even then you really should only evaluate his success based on those few states or races that were actually in play. But to evaluate a model’s accuracy in saying “one candidate has a 60% chance of winning” is really “correct” 60% of the time, Nate’s track record is very early on in its lifetime.
– Nate has said this himself, but in order for his model to be correct, he needs to be “wrong” sometimes in order to be “right.” That’s if you consider it “wrong” when 60% odds don’t “hit.” People who say Nate’s model is so great because it’s been “right” so many times are making the same mistake I’m trying to clear up here.
To be fully transparent I read Nate’s blog a lot and hope for Obama to win. That said, I’ve spent all of 30 minutes on this here (Crank’s) blog and am impressed by what I see. Crank does a good job and writes well, and while he has an ideology (don’t we all) and it doesn’t jibe with mine, I don’t see it clouding his judgment. Also am refreshed at the thoughtful comments and lack of trolling. Too bad about the spam ratio though 😉
P.S. I’m not so sure you can say Nate has “painted himself into a corner” as the Left’s last rational hope for optimism – there are a few other models out there trying to predict the electoral college outcome, and not only have I yet to see one calling it for Romney, but Nate’s model is assuredly not the most optimistic for Obama. And as an Obama supporter I feel okay about, but not so great about, ~70% odds. I’d rather be Obama than Romney in this race, but it’s going to be a nailbiter no matter what.
P.P.S. – for better or worse, there is decidedly more than just poll toplines at play in Nate’s model. I don’t think it’s fair to say that’s all he uses.
Repeating myself, but just to address one sentence in Crank’s post. I’m sure Nate is well prepared for the possibility that Romney wins. It will not shock him. Someone beating ~25% odds shouldn’t shock anyone. It happens in sports ever day. In fact, if candidates with 75% odds under Nate’s model won close to 100% of the time, that would give him a better reason to revisit, because there definitely is something wrong with the model in that case.
I don’t think a Romney win will cause much consternation in the Silver camp as far as his model goes. But the superficial media has certainly painted it in a way that Nate will be widely considered “wrong” if Romney wins, and in that sense I’ll feel bad for him. Comes with the territory though, I guess.
Do you try to account for voter fraud in your analysis, Crank, or do you think the effects of voter fraud are negligible?
I think Silvers weakness is he is all science and no art … to assume a GREATER Dem turnout in 2012 than 2008 shows an ignorance of historical electoral shifts that is simply not reasonable … add to that 2010 and all the Govenor races and you have evidence of an expected larger GOP turnout in 2012 …
I’m not trying to be a wiseguy, but what is the value of a model that can never be tested in a meaningful way? Yes, I can see how constructing one, and discussing it, can be fun for those of us with an interest in the activity, but if a model is supposed to be a tool, and we can never have confidence that the tool actually works, are we really doing anything more than the equivalent of visiting the Tarot card reader’s booth, for some mild amusement?
Of course, the same might be said for large segments of the “science” of economics.
To counteract your main point, which is that you just can’t lose indy voters by the percentages we’re seeing and win an election, there’s an obvious counterpoint.
Simply that a large number of formerly republican voters now self-identify as independent from the right (i.e. tea party).
Josh Marshall at TPM has been on this for more than a week already.
Simply that a large number of formerly republican voters now self-identify as independent from the right (i.e. tea party).
Except that the turnout margin in 2010 was R+1. Did all those tea partiers leave the party after an electoral rout?
The biggest problem with 538’s model is their massaged, opaque use of the economy as a Plus-factor for the POTUS. Nobody thinks the economy is a + for Obama. Except Nate.
People are still paying attention to Josh Marshall? He used to be good. That stopped a few years ago when HE became the “tire-swinger” and “bamboozler.”
I don’t think a Romney win will cause much consternation in the Silver camp as far as his model goes. But the superficial media has certainly painted it in a way that Nate will be widely considered “wrong” if Romney wins
To put it in polling terms, the topline may not distress Nate overmuch (his fans are another story), but look at the internals. Silver spent a great deal of time expounding on why Gallup and Ras are shoddy and inaccurate and PPP, Marist, etc. are much better. If the reverse turns out to be true, he WILL have been wrong, multiple times over, and it will be perfectly legitimate to criticize him for giving so much weight to inferior polling, because had he not done so, his model would have been more accurate.
Very interesting. Great comments, too.
I don’t think it will be a nail-biter. Romney wins easily (53%ish; wins OH, VA, FL, one of IA/WI, CO, NV). See Crank’s “toast” post below. These polls do not, I don’t believe, account for how different the turn-out is going to be in 2012 than it was in 2008.
Do you have any comments about Josh Marshall’s article where he offers the hypothesis that conservatives who formerly identified themselves as Republicans now call themselves independents and that is why Romney is winning the independent vote?
I don’t know how much Nate’s “Gallup vs. the World” post has been heard third- or fourth-hand or just lightly skimmed. I welcome anyone who hasn’t to read it (or at least skim it better) because he doesn’t exactly throw Gallup under the bus.
In fact, he gives Gallup more weight in his model than any other daily tracking poll. About half as much weight as all other tracking polls combined (there are pie graphs, so you don’t have to read that part if you don’t want to).
He laments the fact that Gallup’s reliability goes down when it marches out of step and that its numbers may screw up his model, because his model was set in stone months ago and he can’t tinker with it now, but it’s disingenuous to state that he gives higher regard to more Democratic-leaning polls in his calculations.
Charles Pierce nailed this nonsense earlier in the week with this quote:
“…we are listening to complaints about Nate Silver from the members of a professional pundit class that, judging by its track record, would have bet against gravity 30 seconds after Galileo dropped the hammer.”
Also, regarding the whole Independents leaning to Romney, the Tea Party, AKA the GOP, calls themselves Independents*.
*after the world witnessed the disaster which was the Republican W Presidency. BTW, they do it for the same reasons the Nazi soldiers shucked their uniforms in favor of civvies as they ran from the battlegrounds at the end of WWII.
A fact which may be decisively relevant, but is widely overlooked, is that if Romney wins the national popular vote he becomes president regardless of how many states he wins.
Why? California, and 7 or 8 other states, now have laws that give ALL of their electoral college votes to the winner of the nationial popular vote. California has 55 ECV’s. Enough said.
But Obama is NOT behind in the national polls. Of the 7 national polls from the last 4 days, three have shown Obama leads, two have been ties, and two have shown Romney leads.
Isn’t it more likely that the popular vote will be nearly even–perhaps with a slight Obama lead–and that Obama will win those swing states as projected? Rather than the popular vote being tied and all the swing state polling having been *consistently* wrong?
Boy, talk about an army of words accompanying a corporal of thought. Actually, a PFC of thought.
Want to be $1,000 on Obama winning the EC?
You emphasize that Obama is not doing well with independents and that historically bodes badly for an incumbent. But isn’t there this in Obama’s favor?: Obama is unprecedented in being able to bring out minority voters. This is a factor that “election history” is unable to illuminate. Perhaps Obama could afford to lose a whole chunk of independents, given his steadily rising status with blacks and Hispanics.
Very interesting. I won’t pretend to know how this will turn out, but I did read a plausible theory by Steve Kornacki as to how Romney could win independents and still lose the election. Kornacki’s theory is that many of those identifying themselves to pollsters as “independents” are actually tea party members who have always voted Republican, and that when you add them to mainstream Republicans you get more historically representative numbers.
It seems to me that it would be relatively simple to cross check the pollster work. E.g., the latest Washington Post/ABC poll not only has its weighting, but also has historical information.
Obama voters: 83% Obama, 14%Romney
McCain voters: 05% Obama, 94%Romney
New voters(7%): 62% Obama, 34% Romney
Given that Obama won 53% of the vote, and McCain won 46% of the vote, it is relatively ease to plug that information in and get
Obama voters: .93(.83*.53+.05*.46)+.07*.62=47%
Romney voters: .93(.14*.53+.94*.46) + .07*.34=49%
In order to get the WP’s results, we have to assume that 4% of the die-hard Republican voters who went to the polls anyway are not
going to show up at all this election cycle. The issue isn’t over-sampling Democrats actually, its under-sampling Republicans, it seems.
Anyway, why don’t more pollster check there results vis a vis historical data they have included to see if the numbers work out the same. Here using the McCain/Obama numbers to re-weight the poll according to known election results turns the ABC/WP results into exactly the same results as Rasmussen.
I wonder if you think this is a reasonable way to cross check weightings?
“A fact which may be decisively relevant, but is widely overlooked, is that if Romney wins the national popular vote he becomes president regardless of how many states he wins.
Why? California, and 7 or 8 other states, now have laws that give ALL of their electoral college votes to the winner of the nationial popular vote. California has 55 ECV’s. Enough said.”
This is an easily falsifiable fact. The National Popular Vote Interstate Compact only takes effect if enough states sign on to constitute 270 electoral votes. That has not happened.
Not quite. It doesn’t go into effect until more states (combined ECV is >269) sign the Interstate Compact.
[quote]A fact which may be decisively relevant, but is widely overlooked, is that if Romney wins the national popular vote he becomes president regardless of how many states he wins.
Why? California, and 7 or 8 other states, now have laws that give ALL of their electoral college votes to the winner of the nationial popular vote. California has 55 ECV’s. Enough said. [/quote]
Not quite true. At least in California’s case, the law doesn’t take effect until a majority of states follow suit. It does not apply in this election.
crank You are an idiot. In the swing states Obama is NOT losing the IND voters. Not at all .
Obama IS losing the IND voters in KS or GA or IND and over the last several days in NC.
GALLUP as of OCT 20… In the Northeast states Obama was leading 53-47…in the MIDWEST 52-48 in the West 51-49 in the SOUTH Romney was leading by 26 points. This sort of REGIONAL split was seen by the last BATTLE GROUND poll OCT 24.
Romney is going to pile up HUGE popular vote margins in the Rockies and the Southern states but will likely lose states like OH IA WI VA by very small margins
Mick Slayed: Those laws don’t take effect until enough states have signed on to total 270 EVs. Since that hasn’t happened yet, they will not benefit Romney if he wins the popular vote.
NEW FLORIDA POLL OUT …Obama moves up by +2% and +5 in VA
10 days ago 7 days ago Romney was clearly in the lead in FL. But over the least several days that lead is gone with all the data showing FL again being tied and Obama moving ahead in the last few polls.
keep in Mind CRANK that FL has 500,000 MORE registered Democrats than the GOP when compared to 2004 and 2000.
Excellent analysis from another sabermatician. I have been seeing something similar in the polling internals for a while. There is obviously a disconnect between some of the polling groups. If Gallup is correct that Romney is +5 nationally, it would be extremely difficult mathematically for Obama to win enough swing states. I guarantee that if Romney wins the popular vote by at least 2%, he will win the electoral college.
My problem with this post is that you make it sound like predicting % of turnout for Rep / Dem / Ind is somehow new. However, that is at the core of what polling companies do – you can’t release a poll without predicting % of turnout amongst those various groups of voters.
What you are suggesting is in fact then already done in each poll.
Are there problems with turnout prediction? Of course. Could the pollsters by and large be very wrong about turnout? Sure. But the point of aggregating multiple polls is an expectation that with enough guesses at the actual turnout will in the average be more accurate than any single guess.
What you seem to be suggesting is that all the polling companies are dramatically underestimating independent turnout. That may be, but it is less likely than the contrary position, that on the aggregate they are close to correct.
Excellent post! What convinces me is the amount of time the Obama team is spending spouting they are ahead in the states, and there is no Romney surge. They are spending more time telling us how confident they are than they are out campaigning and appearing confident, and Obama’s has gone so nasty and negative that he looks desperate. The Romney team on the other hand isn’t saying much, they are just out there with smiles on their faces looking confident. We’ll know soon enough….but based on what you see, hear, and feel it looks like Obama is indeed toast!
In the swing state of Pennsylvania, a poll was recently done by Franklin & Marshall that had Romney at +16 among Independents. They also had Romney at +2 in the Philadelphia suburbs, but somehow still down 4%. Other swing states are showing Romney with a big edge among Independents as well. this calls into question many of the poll results favoring Obama.
You write that in matters such as this, history is a more reliable guide to the future than mathematics, which is a good point. But if we look at the history of polls in recent elections, we see that they’ve gotten to be pretty good. And Nate Silver is respected not because of the elegance of his algorithms (he doesn’t let anyone see them anyway) but because of his history of calling elections right. He may get this one all wrong, as he himself admits. It’s been an unusual contest from the beginning, and Sandy introduces all sorts of unknowns. Still, poll after poll shows Obama with enough Electoral College votes to win. Were these polls turn out to be wrong, there would have to be some factor, still unidentified, at work. It’s quite possible that this will happen. But it’s looking less and less likely.
If Nate is such a left-leaning political hack, how is RedState any more believable. They have a whole post about gay marriage BEGGING readers to call family members in states “WE” as he describes it, as if he lives in every state that has a referendum and is some how on the TEAM of NO, could win. He say “WE” are on defense in WA, Maryland, and Maine, but if he is a true POLL READER and non-partsian hack, then who is “WE”? Certainly not %54 of Americans who favor same-sex marriage. He begs readers to oppose these measures as much as Nate supports an Obama victory. So only one is correct and the other is a lunatic? I’d say you’ve shown your own leaning in this article SIR. Have you predicted anything correctly before? Not to mention, are you a baseball fan or political scientist……if you write about sports, stick with it… I simply ask, if you want to call someone out as a partisan, while quoting another CLEAR partisan, shouldn’t you look in the mirror at the hypocrite?
There is a sense out there that Obama is losing. The desperation his campaign shows paints him into that corner.
Baseball Crank, you have neatly tied up in a bow many of the inherent problems with statistical forecasting. I have a hefty education in statistics but have been wary of disconnecting my quantitative education away from my qualitative education. I have been roundly criticized for that proclivity.
You provide a brilliant illustrative example of why one can statistically be absolutely right and be utterly wrong in the real world. This is the perfect example why we should be wary of any forecasting model(s).
Empiricism has run amok among our universities and I think is part of why most people are wary of science. Put a number to it and its valid–BS. Have a model that takes serious computing time–wow. But does it do the job? Forecasting is a bitch.
Thanks for the good read. I needed it.
Danny Easterling BA, MS, MPA
FYI, Sandy was not a Hurricane according to the governors of New York, New Jersey, and Connecticut. Do you know how much that cost the insurance companies?
Rommey will to loose! Your a bigget and your poles are rasist!
Very well written. Unfortunately, it still boils down to the following:
* all of the polls that don’t make assumptions about the political leanings of the electorate (even the Fox News polls) are wrong.
* they are all making mistakes in their random samples, although you don’t explain how.
* instead, only the polls that assume (without evidence) a specific electorate that is more Republican that in 2008 are correct.’
* therefore, the average of the polls (in the RCP, Pollster, etc. averages as well as the 538 average) must be wrong.
That logic (such as it is) amounts to saying that observations of reality can’t be right because we don’t accept them. Not too different from saying the world is flat.
Great article–this is the best counterargument to Nate Silver’s methodology that I’ve seen, and it was very well presented.
That said, I take issue with criticisms of Mr. Silver because they appear to be very blatant attempts to ‘cook the books.’ By that I mean those who disagree with Mr. Silver are exclusively those who disagree with him, making it hard to take their arguments seriously. They typically make arguments that start with a conclusion (Romney must win) and use various means of arriving at that conclusion.
The main argument against Mr. Silver here is that he is using top line data only and turnout of various groups will be different on election day. The problem with this line of reasoning is that it has been consistently wrong, especially in Presidential cycles. Mr. Silver has done an excellent job on his blog of proving the consistency and remarkable accuracy of aggregate polling. Is it wrong sometimes? Sure, but from a layman’s perspective it’s wrong just as often as the set of polling we’re seeing now is going to be wrong 20% of the time.
To depart from this article, the most concerning thing to me regarding criticism of Mr. Silver is that it doesn’t seem to come from an honest place. If Romney was winning, no doubt Romney backers would be signing the praises of Mr. Silver’s model and possibly liberals would question its accuracy. Perhaps this is an issue we should revisit outside the heat of an election season, I only hope by then each side has not too deeply entrenched themselves in their positions to not come back to reason.
For Dave’s comment above,
Loved your comment(s). As a Romney supporter, I wish more people would view this whole mess in the same vein. For whatever reason, Nate Silver has become a lightning rod in this election (and i suspect that, no matter which way it goes, all of this publicity will be good for him). But I think that most of my conservative friends have a total misunderstanding of what his goals are.
I am no statistician, but from what I can gather with my limited knowledge is that the bottom line is, no matter which way this goes, he and others are ultimately trying to predict the electoral college via the polling across many states and polling agencies. A difficult thing to do, which should only get easier with each election from here on out. As you said, statistically, it will not be known for a long time.
This is a great article, not in the sense that it proves one thing one way or the other, but that it looks at the information we have in a mostly objective way. It is a shame more people cannot see what Nate’s and this blog’s work are really about. Yes, Nate’s personal opinions may be to the left, but that should not affect his data. If it does, I suspect he will come up with a way to fix it. His previous work, especially in baseball, is something to be admired.
Anyway, not sure if any of that makes sense, just wanted to reflect those statements and say that I would not be surprised by any result on Tuesday.
“Do you have any comments about Josh Marshall’s article where he offers the hypothesis that conservatives who formerly identified themselves as Republicans now call themselves independents and that is why Romney is winning the independent vote?”
Yes. He is wrong because his assumption is that the number of people calling themselves Republicans has dropped since 2008. It has not, it has increased. Gallup, Rasmussen, and Pew all find this.
In addition, the WashPo recently had an article out noting that of “shadow partisan” independents, Romney was winning 8% more right-leaning indies than Obama was winning of left-wing indies.
So basically, what you’re saying is the same conservative meme running around everywhere: Nearly every single poll is wrong. I mean that’s just a tired meme.
When you look at states like Ohio, everyone has Obama winning. Romney can’t become president in too many realistic ways without it. But that’s no problem, you know he’ll win it even though no poll shows this.
Got it. Thanks.
Silly blog. Most of the “Independents” that Romney is clobbering Obama with are traditional Republicans who just became independent this election. If Romney was really up 16% among “true” independents, he would be winning the swing state polls on the topline as well. But I think the argument that of the three groups (Dems, Rep, Independents), Romney is killing Obama in two of them, yet almost every poll shows Obama winning Ohio anyway is spurious. Some of these polls just call up voters using random sampling and don’t oversample Democrats. And in those polls… Obama is winning Ohio outside the MOE. Hard to see how that is possible if Romney is way ahead. Main problem Crank made: sticking to labels, instead of looking at what exactly an “independent” is. Ironically, I think that is a far far bigger error than Nate using topline predictions. Plus Nate’s method is at least testable, while Crank’s isn’t.
What if we held an election and noone showed up?
I was interested in an article I read last week that stated that we would possibly see an empty shell of turnout on the Democrat side.
The argument was simply that the method for showing displeasure or lack of excitement with the President’s agenda was to not show up for voting.
If the independents are swinging as far to the right as has been documented, time and again, by every polling agency, what is the logical outcome of Democrats who choose not to vote for Romney?
Like the Republicans in 2008, they will just … not … show up. This is evidenced by the reduced numbers for early voting as has been documented, again, by every major polling agency.
There will not be a wave of popular support for Romney, there is just a demonstrated lack of enthusiasm for the President who was elected with a great deal of enthusiasm four years ago.
I have talked to many friends in the battleground states. I am from Wisconsin. Where there were oceans of Obama signs four years ago, there are none now.
My belief is that the whole support mechanism is honeycombed and that the get out the vote effort will not sustain Obama at the polls. I voted for Obama in the last election. I do not believe he has earned my vote in this one. But this does not mean that I support Romney. Does this make sense to any of you?
I will be interested in the outcome of the election.
Wow, great article.
Sounds like the Baseball Crank is on crank.
It took me a long time to read this post, both because of it’s length and because the concepts presented required reading slowly to comprehend. I read the post to it’s conclusion anyways, and honestly that’s the best compliment I can give you as a citizen of an internet populated by fast food/ bumper sticker sized intellectualism.
There is nothing more dangerous than a person with a lot of intellectual horsepower who is also terminally stupid. Sort of like the rocket scientists who destroyed two space shuttles, or Donald Rumsfeld. Combine that with a person to say whatever lie is necessary to tell their story, and you get baseball cranks article.
How do we know he is lying? Because if he believed what he said then he would tell you how many shares of Romney that he has (intrade.com and others) He has a chance to triple his money (35 cents for a share of Romney stock will get you a dollar when he wins)
But how many shares does he own? Exactly zero. How do I know this? Because if he were an honest writer rather than a well paid blowhard, he would have told how much of his money was in back of his mouth. He did kinda actually, by saying nothing.
The problem with stating that Obama has a 73% chance of victory is that Nate is being rewarded for making the projection, not for being correct in his projection.
In this election cycle, Democrats are less enthusiastic than Republicans. The faithful may not vote if they hear a prediction of Rain of Toads. Donors may not contribute. Down-ballot races may be effected. Thus the assertion that 2012 will be a more Democratic year than 2008 must be made, and thus it will be made, however absurd such an assertion may be on its face.
There are staggering differences between 2012 and 2008. To name a few, Republicans are not demoralized. The 2010 election happened and it happened for reasons which still resonate. The more people hear about the IPAB, the unhappier they become. No one I know is pleased by the Deloitte study that says no employer wants to be the first to drop health care coverage for their employees, but every employer want to be second. The Republicans in 2012 did not just lose the Senate in 2006 because they lost the middle of the country and their base failed to turn out. The Republican candidate in 2012 was never a favorite of the media, as McCain was when he was attacking Bush (the rogue maverick) and up until he became the Republican nominee (and became too old to be a viable candidate, overnight). Lehman didn’t just implode. Romney picked Ryan — not Sara Palin — for his running mate. Romney wasn’t outspent by a staggering amount because he chose to go with federal funding, where Obama reneged on all previous pledges to do the same. Citizens United was the law of the land during this election cycle. Obama wasn’t an incumbent in 2008 and his rating with independents was 15-20 points in the other direction vis where they are now. McCain didn’t destroy Romney in a debate watched by 70 million people, whereupon Chris Matthews had a meltdown so comic it made its way into SNL.
Any of us could go on. 2012 isn’t 2008, whether Romney wins or loses and the D sample in 2012 will not match of exceed the D sample of 2008. Republicans sense this. Democrats have to free-base MSNBC to convince themselves otherwise.
If Nate’s projections are wrong, the faithful will reward him, next cycle, if he makes the same (wrong) projections. If his projections are correct, he is a prophet. This is because his projects are valuable as a propaganda tool, not a predictive tool.
In baseball, his projections are irrelevant. The Padres will bat X irregardless of Nate’s projections.
In politics, a 2012 D+15 sample, where the 2008 turnout was D+10 … it’s a pretty common poll, brought to you by some tool of the party aparatchiks.
There is a fundamental difference between an analyst, who will tell you what you don’t wish to hear, and a propagandist, whose numbers are based on assumptions so deeply flawed as to be intellectually dishonest.
Nate is spot on. Right now he has Obama 80% and Romney 20%. Still great odds for Romney. If you were told you had a 20% chance to win the lotto you would be pretty excited, right?
Larry Sabato is good. Rove has similar numbers on the electoral college. He is of course in the bag with Romney and Fox News. So he put his spin on it and is predicting 279 EV for Romney.
Romney lying to OHIO voters with this whole Jeep thing really killed his chances in OHIO. You can’t lie and then keep running on it. Especially when you have the Chrysler CEO calling him out. OH, NV, and WI are going to Obama. Romney can clean up with FL, NC, NH, CO, IA and VA and still lose.
As stated already Kornacki made a great assessment. After the Tea Party showed up Republicans labeled themselves as independents. So this lead Romney has with Independents is a false narrative because a large portion of those independents are already going Romney. From my understanding the independent advantage for Romney is +7 on average. +7 is weak when you consider the Pro-Romney Tea Party independents.
Fox cherry picks Gallup when it is +7 Romney and then drops Rasmussen when they have Romney +2. Look at Rasmussen in 2008 for OHIO. The NOV 2 poll had it tied. Obama won +4.6. O’Reilly seems to think Rasmussen was spot on in 2008. Hahah. Nice spin, Bill.
The election wil be close but I see a 280+ EV for Obama. I want to see Hannity’s face on the 7th. Hahaha.
“My thesis, and that of a good many conservative skeptics of the 538 model….”
With respect, I’d be interested in a statistician or political scientist offering a result-blind critique of Silver on methodological grounds, but I’m just not interested in “conservative skepticism” driven by an ideological refusal to accept his result.
The problem with the polls is an undersampling of Republicans (even in the polls that Romney leads in).
Look at the 2010 mid terms…which ended up being I believe a turnout of 35R 35D 30 IND (something like that)…
Nate wants us to believe that the Republican voters that showed up in 2010…are going to stay home in 2012? Huh?
You can make any poll say Obama is going to win if you sample 40 D, 32 R, and 28 Ind…but there’s no way the Dems are going to get that turnout.
In response to Dave’s comment that Nate Silver isn’t predicting Obama will win, only stating there is a 75% probability that he will win…that is not entirely accurate.
Silver’s model may only be stating a probability, but Silver himself is being very open about his prediction that Obama will win. For example, his $1000 bet, that he tweeted on Twitter in response to Joe Scarborough, that Obama will win.
It seems to me there is a big difference in discussing “probabilities” versus “predictions”.
Probabilities can’t necessarily be proven wrong-therefore his mathmatical model can’t be proven wrong. How do you prove something wrong, that has built into it an allowance for being wrong?
However, a prediction is unequivocally able to be shown to have been wrong. This to me is the inherent problem with giving too much credence to models. You’re basing something disprovable (a prediction) on something that is not (a probability).
It all just puts my head in a spin, and reminds me that I’m not the sharpest knife in the drawer and these discussions are WAY above my paygrade. However, as my mother always told me, “Common sense will get you a lot further in life than book learnin’, or at the least make your trip a lot easier.” And common sense tells me the state polls assumptions are wrong, and therefore, any probability based on them is wrong. I guess that’s a really wordy way of saying I agree with Crank-apparently-and I’ll shut up now because I’m rambling!
Selling Obama contracts short and buying Romney contracts on Intrade looks like a really good bet to me. I made the same adjustments to the internal party identifications that Crank did, and for the same reasons, which suggests that unless there is a substantial movement in the electorate toward Obama in the next 4 days (currently most polls still show movement toward Romney), Romney will be our next President. Actual data shows Obama underperforming in early voting patterns in Ohio and other venues.
The points made about polls that do not break down responses by party identification which show Obama leads (still usually within the margin of error) have some resonance. However, because they lack such pary identification data, which actually has a high predictive value, not only on how people vote, but also on enthusiasm, such polls mask the ability for independent analysis.
My primary reservation regarding the reliability of the adjustments Crank and I both have made which then evidence a high probability of a Romney victory are the Intrade numbers themselves. Most of the time when I am wrong, it is because I find my analysis to be an outlier, which Cranks’ and my analysis is when referencing Intrade. But then, that is also the time when one can make money going against the conventional wisdom.
Please watch this informative video:
Now is the time to focus on what matters. Together we can make things better!
Isn’t partisan identification in the electorate pretty easy to check, though? E.g. Pew released a survey in early August showing the following numbers: 38% of likely voters self-identify as Democrats, 33% as Independents, and 25% as Republicans. On these numbers, an Obama victory is a foregone conclusion, even with a 10-point edge for Romney among independents.
Of course, different polls find different results in party identification, but it would appear that Rasmussen and Gallup are outliers in finding anything less than a 3-point Democratic advantage. Further, likely voter models correct for the purported ‘enthusiasm gap’ and are therefore more Republican-leaning (by which I do not mean ‘Republican-biased’). But even so, it seems clear that, given much of an edge in voter self-identification, Obama can pretty easily overcome a (supposed) 10-point deficit amongst independents – especially if the independent numbers are even slightly inflated by the above-mentioned Tea Partier flight.
And what will your post be on November 7th when we learn that Nate Silver underestimated th president’s advantage? Anyone who pays serious attention to politics knows that the swing state polls have been pretty darn accurate in the past two presidential elections. Silver’s model reports the undeniable advantage President Obama has in polls in a way that accounts for their historical track record and then dampens this advantage to account for uncertanty in a way that less sophisticated polling averages do not. The results take conservatives out of their comfy fox news alternative universe so they deride him as some kind of partisan hack. This is understandable given the propensity of conservatives tofocus their attention on sources of news, political commentary and pseudo-science that merely reinforce pre-conceiced notions and tell them what they want to hear. Understandably you project this mode of operation on liberals. You are, however, mistaken. Silver’s loyal readers visit his blog because they want to know what the actual result of the election is going to be! If Romney has a 75 percent chance of winning we want to know that. That is the service Silver provides and he has provided it since he was relatively unknown. And the basic premise of the skewed polls\skewed predictions conspiracy theory is way off. If you want every last Obama supporter in Ohio to vote you don’t tell them Obama has an 78 percent chance of winning the state! You don’t tell supporters in Wisconsin that he has a 90 percent chance or Pennsylvnia that he has a 96 percent chance (silver’s projections are somewhere around there). If you want good turnout you say the state is a tossup. This idea that conservatives will somehow be convinced not to vote by the idea that their candidate has only a 25 percent chance of winning is pretty ridiculous. Any brief analysis of recent electoral history shows this is simply not the case. A recent example can be found in this year’s republican primaries. What did the polls say about santorum’s chances in states he won like mississippi? If you try to actually inform yourself you won’t be a helpless pawn spouting whatever hannity and limbaugh tell you.
Well I hope you’re right.
I’m definitely not as much of an optimist as you are about it — I believe Sandy (and to a lesser extent, Chris Christie) has swung the momentum back in Obama’s favor
However, Romney was winning before this happened, and I hope it’s by enough that your prediction proves correct.
Can you more fully explain what this means and what this data says? This seems to be the crux of the issue? What are these poll sample saying and why could they be right or wrong? Thanks.
“with the sole exception of the poll samples used by some (but not all) pollsters”
I found an interesting pattern in the national poll data that backs up a theory I saw on Twitter. The pattern is that the larger the LV sample, the worse Obama polls. Look it up on RCP. The reason for this is that polling $ has decreased while difficulty in reaching LV’s has increased (or response rates have decreased, same thing). In order to actually get the job done, pollsters must loosen their LV screens, shifting the electorate +D as they do so. In other words, LV’s look a lot more like RV’s because of response rates falling off and $ for polling decreasing. The best-funded polls get more contacts, higher samples, lower margins of error, tighter LV screens, and lower #’s for Obama. That’s just what I’ve observed.
I predict this blogger’s credibility will be toast after the efficacy of Nate Silver’s model is proven yet again. The entire argument is based upon a flawed premise: that because Obama is allegedly losing with independents, he will lose the election.
Right-wing poll truthers do not understand party ID in polls. Poll respondents self-identify and “independents” are not really independent: most lean strongly towards one of the parties. Moreover, it has been shown time and yet again that the party that is less enthusiastic is more likely to “lose” voters to the “independent” label. Thus, a poll with a D+9 or R+5 sample does not necessarily mean the other party is underrepresented in the poll: it just means that for whatever reason — typically enthusiasm issues — one party was more likely to identify as “independent”.
Independent voters are independent like Bill O’Reilly claims to be independent. They’re not undecided. Republicans “unskewing” polls is a fool’s errand. Republicans are already in these polls: they just labeled themselves “independent”. Obama losing “independents” who we’re never going to vote for him anyway means jack squat. Romney still has no path to 270. Nate Silver — who predicted 49 of 50 states and all Senate raclas in 2008 and who called the Republican wave in 2010 — is about to make his critics look very foolish in 2010.
Mick Slayed –
ONLY IF ENOUGH STATES WITLH A COMBINED TOTAL OF 270 ELECTORAL VOTES AGREE TO DO SO. Won’t take effect until then. So far, not enough states that total to 270 have agreed to do so.
Your mind and analysis is tied up in knots because you want to believe something that is not supported by all the facts. You dismiss the polls when they disagree with you and then you turn around and use the polls to draw conclusions that you want to believe. Remember that Godel proved that all logic begins with unprovable assumptions. As a result, math books usually start with the axioms. Of course this is not a new concept just consider the opening statement of the Declaration of Independence which states that “we hold these truths to be self-evident…” So of course, you can reasonably challenge the assumptions. You didn’t state your assumptions at the start in order to make a logical argument. You state them at the end as part of the conclusion. Several of your conclusions are not supported by studies that have examined the voting behavior of undecided voters in past elections. So you have an opinion. So what! we all do.
Problem is these days is that pundits and this column have been focusing on the “horse race” aspect of this election and not the implication aspect.
To women – loss of reproductive rights, no equal pay for equal work,
to minorities – disenfranchisement, back to second class citizenship – all are factors.
Actually, societies with a strong pecking order (“Blacks, et al knew their place,” India and its caste system, South Africa and Rhodesia with their apartheid) were MORE productive and efficient than a “mix-it-all-up” conglomeration society that Dems are pushing for. Of course, that wasn’t good for those at the bottom of the food chain.
Very sad, but true.
Remember, the pundits are “free, white and 21” and will not be affected by a GOP takeover, so they can pontificate.
Jews should remember, too, that once the women and minorities go, they’re next and they shouldn’t feel too comfortable with their new white, Christian GOP friends.
One of the best articles I have read this election cycle. Fantastic unbiased analysis. Good job.
One of the problems with Pew’s self-identifier method for determining partisan idenditification in its samplings is that if it were to have more blacks in its sample than in the population than actually votes this time around, it will skew the sample in terms of self-identified Democrats, since such a high percentage of blacks are Democrats. I use this extreme example, since it is something that right now we can all agree upon. Other groups are more fluid to one degree or another. Since the 2008 election was such a variation from other elections, and since this election is probaly going to return closer to norm, taking into account some modest demographic changes such a couple of percentage more Hispanics in the population, but not necessarily in the voting pool, given the issue of enthusiasm, enthusiam among the demographics and more normal voting patterns suggests that Gallup and Rasmussen are much closer to predictiing the 2012 partisan vote than is Pew. However, we shall not know for sure until after the vote is counted. That is why nothing is decided until after the election. Otherwise, we would never need elections, just polls.
Here is Nate’s problem. In politics being majorly wrong once can destroy your credibility. Hes got Obama up now at over an 80% chance to win. That’s pretty big.
Its all about turnout. In 2004 republicans nationally tied with democrats in turn out and were up by 5% in Ohio and in 2008 Democrats beat republicans by 7% nationally and 8% in Ohio. Ive looked at all the Ohio polls that Nate is using to make his predictions 90% of those polls are saying Democrats will have an 8% advantage or MORE in Ohio.
So which will it be 2008 or 2004? Neither. Its probably somewhere near the middle of those 2. But even assuming that democrats have a 4% advantage in Ohio that turns the 1-2% Obama lead in the Ohio average to a 1% Romney lead and switches the entire election prediction to a near 50/50 probability.
Nate is being intellectually dishonest by taking the polls at face value and could cost him all his political credibility if Obama doesnt eek out that win in Ohio.
I think Silver’s models are a shell game. He averages poll results based on a weighted assumption — based on his own preferences. He covers his “prediction” by saying that Romney might win. So, he can’t be wrong. As for the election, I believe that 2008 was a high water mark for Democrats based on the novelty of Obama’s candidacy. Turnout will return to the mean. Given the reliability of party voters, the independents will be the deciding factor. At the moment, based on early voting, that appears to favor Romney.
This should be required reading for anybody who’s about to base any sort of opinion piece on Nate Silver’s work.
For me, the concern is not with the necessary limitations of any predictive model of elections, which have been well-addressed in this piece – the concern is with the impact a failure to appreciate those limitations has on the wider discourse.
When Silver’s model gives Obama a 77% chance of winning overall, or an 80% chance of winning a state that his own data show is well within the margin of error for either candidate (Iowa was in that position not so long ago, iirc), it LOOKS a lot to the lay reader like Romney is a rank outsider. However much evidence that lay reader encounters about the Republican enthusiasm gap, or the remarkable volte-face in independent support over the past four years, or the somewhat freakish nature of the 2008 electorate, Nate Silver’s prohibitive percentages are there staring him in the face – and they’re the product of a robust and respected algorithmic model, and the lay reader tends to give that more credence than he would a mere human opinion (the lay reader tends to forget that mere humans programmed the algorithms).
This is why Nate Silver’s model bothers me less than the Intrade and other betting-house results, which tend to be even more bullish on Obama. There’s a strong reinforcing element of supporting Obama because the model supports Obama because the poll data the model uses supports Obama because people believe the model… it’s a bubble; it’s tulipmania.
And those who know Nate Silver personally may not doubt either his credentials or his motivations, but when Silver bets Joe Scarborough that his model works, AND THAT THIS WILL BE DEMONSTRATED IF OBAMA WINS – despite the fact that Scarborough contends there’s a 50/50, rather than 70/30, chance of that outcome as well – he is actively participating in the promulgation of misrepresentation of his model. He is actively encouraging the lay person to believe that Obama really is a prohibitive favorite in an election where the raw data on party affiliation, voter enthusiasm, measurable early voter behavior, and a host of other metrics shows that Obama’s base is less enthused than 2008 (and has shrunk besides) while Romney’s is vastly more enthused than in 2008 (and has grown besides).
The mass media, breathlessly reporting whatever sounds good for the incumbent President, are certainly far worse signal-distorters than Nate Silver; but I don’t think he gets off scot-free when he editorializes the way he does about his own model.
Maybe it’s inappropriate to state this openly, but don’t the President’s race and consequent ability to turn out black voters justify weighting the polls in his favor?
Black voters, who are fully 10% of the electorate, are registered in huge percentages, will vote, and will vote for the President almost unanimously. This amounts to one candidate being “spotted” the number of black voters who would not otherwise be voting – which may be as much as 4% of the There is not historical precedent for this in American presidential elections. Indeed, doesn’t it render historical analysis of elections useless?
All that to end with the conclusion that no president has won the electoral college without a clear cut popular vote victory….WOW , Um except George Bush just did a few years back.
Way to to whip your crank out for everyone to see ….sadly its lacking both in girth and lenght
After reading some of the other comments, I have to take issue with the statement that Silver did well in 2010. If you read his final “prediction” for 2010, he indicated ~54 seats for the GOP in the House. His model had actually predicted 59, but he reduced it because he didn’t believe his own model. The actual result was 63 seats. So, at the last minute, Silver ignored his own model to insert his own personal preference. If he’s going to ignore his own model, so am I. The Gallup blog is also illuminating. He wants to believe Gallup is an outlier, but it’s such large constituent in his model he can’t walk away from it. Either be a statistician, or a pundit. At the moment, in my view, he has no credibility as either.
“In my opinion, which view is correct is not one that can be resolved by mathematical models, but rather by an examination of the competing assumptions underlying the two sets of polls and an assessment of their reasonableness in light of history and current political reality.”
No, it can only be resolved by looking at who was right on Wednesday morning.
How about the possibility that a share of the 2008 independents, now self-identify as Dems? This would explain Romney’s double digit in indies and the higher Dem turnout that so many polls observe
Well thought out and well written.
Let’s all hope that you are correct.
A couple of points. Rasmussen has published its monthly political trends poll with a sample size of 15,000 voters since 2002. If you look at the changes over the last decade you’ll see that while partisan identification changes over time, it generally does so slowly.
This poll correctly identified the Democrat margin in 2008 and the Republican margin in 2010. That poll indicates that the as of the month of September electorate is 36.8%-R, 34.2%-D, and 29%-I. Note that voters are abandoning Republican identification for independent status, if anything the Democrats are doing so. If this poll is correct the independents that are breaking for Romney are either true independents, or disgruntled Democrats. There is no spin on the that is good for Obama. There are more self identified Republicans who are voting 92%-92% for Romney, who is also winning 8%-11% of self identified Democrats, and independents by an 8%-19% margin.
This supports the previous observation that the turnout models used in Obama leaning polls are under sampling Republicans and independents.
As for Nate’s 75%-80% likelihood for Obama. Statistically he’s covered but what he is really saying is that is the probability that the projected Democrat turnout advantage is correct. If you look at the debate in these comments, that is the underlying topic. How likely is a D+8 or D+10 turnout for this election. All the measurable data on returned mail in ballots, and who is showing up at the early polls reflect Democrat under performance and Republican over performance compared to 2008. Even in the states where Republicans are behind in actual turnout they are well ahead of 2008 in margins that suggest they will ultimately carry the day. In the face of this Nate’s unstated but underlying assumption of a 75% probability of an overwhelming Democrat turnout advantage is why his reputation will deservedly take a hit if incorrect.
Nate Silver has also verged off into lala land by recognizing polls like RAND and Google Consumer in his national weighting. He’s now went above an 80% chance of Obama winning. Hopefully he’s able to handle the loss well. I have my doubts.
I found Crank’s analysis to be more thoughtful than many and I share his passion for Sabremetrics, I also think his strong partisanship colors his work here as it has in the past. Not in an over-the-top way that makes mind-numbing the reading of many partisans posts, but in a way that still falls short of the objectivity he might bring to baseball.
Consider some of the things Crank wrote in 2008:
“The Gallup poll, which admittedly is one of the more volatile polls . . . ”
“JULY 29, 2008
Polls in Perspective
We’ve had wild-swinging polls lately – national polls showing Obama +9 and McCain +4 – but the important thing is not to panic. Polls go up, they go down, and at any rate while the national polls can give you some idea of the direction and momentum in the race, in the end the only polls that matter are on a state-by-state basis, and those really don’t get hugely meaningful until after the conventions.”
A refutation of Josh Marshall’s point that lots of Republicans turned independents.
Catching up on comments here.
feeblemind – Voter fraud has always been with us. In modern times, it’s not going to swing any races unless they are extremely close. Of course, that’s when it matters, but for predictive purposes a race that close is essentially a tie anyway.
Josh Marshall’s post is built around a chart that shows D+4 and 30% Republican on Election Day 2010. Which is rather dramatically at odds with both the exit polls and the election results that very day. Why? “One point to keep in mind here is that we keep this data set of ‘adults’ rather than registered or likely voters. That makes it somewhat different from the voting electorate.”
The exits showed an electorate that was 36% Republican in Ohio, 37% in Wisconsin.
Marshall is right that if the election is held among all adults, Obama would win. It will, however, be held among actual voters, as all elections past have been. I will put my faith instead in the Gallup and Rasmussen party ID surveys cited in my prior post, which have a solid track record of matching the people who actually vote.
On the general point, independent voters didn’t just emerge as a new thing in 2012. They’re traditionally around a third of the electorate. Obama won them in 2008, and House Dems won them in 2006, and they went GOP in 2009-2010.
Tom – The “minority votes” argument founders when you look at the Upper Midwest.
Doug – I agree that a lot of the polls are doing more undersampling of Republicans than oversampling of Democrats.
Sher – First, I’ve always said that Gallup is unusually volatile and a better guide to trends than the actual result. Rasmussen has a better track record of being close to the pin. And Gallup changed its methodology just about a month ago to be more D-friendly. The conventional wisdom is always that state polls are slower to pick things up than the national polls. My view of the state-by-state polls, like Trende’s, is simply that a lot of the pollsters with better track records and more realistic views of the current electorate are doing national polling right now.
Very impressive, and I concur. If President Obama is reelected, I will be shocked. Nearly every sign points to a Romney victory, by my calculations, and I have been correct in predicting the winner of the last eight presidential elections. I expect to soon make that nine for nine.
Squabbling about what constitutes the independent block of voters is silly. There are more people self-identifying as Republican while less self-identify as Democrats. There are those, like a friend of mine, who self-identify as conservative, and view the Republican party as closer to his politics, but not enough to belong to that political party. There are many in the Tea party who behave the same way. They are true independents, not wanting to belong to either political party, but in their case, choosing to vote for the lesser of two evils. When you consider that these voters are out there, and they have been vilified by the Democratic party since 2008, logic would tell you that they are helping Romney expand his independent vote. I, on the other hand, am a registered Democrat who has been behaving like an independent since 2008. Consider me to be a “bitter clinger” who has never forgiven the Democratic party for their obvious anti-Clinton shenanigans during the last election cycle. Many Democrats left the Democrat party after 2008. I would not be surprised at all if these same voters are also propping up Mitt Romney among the indepent voters. I know that I intend on voting Rommey this year, and I know of other formerly solid Democrats who are doing the same thing as I. Obama is toast, and we are happy to help remove him from office.
Want to say upfront, I respect Dan and have read this blog sporadically for a number of years.
On this matter though – I don’t understand why the polls are not trusted to accurately reflect Democratic turnout, but they ARE trusted to accurately reflect independent turnout, not to mention independent political leanings.
We been all over this one on Common Cents…
I think a quick shorthand for this perspective might be garbage in, garbage out. What Nate Silver’s algorithm boils down to is a very intelligent way of averaging the polls. If most non-Gallup polls with a solid track record are suddenly getting the 2012 electorate wrong, particularly at the state level, Silver’s model goes right along with them. (Although I will say strongly question why Google Consumer Surveys are in Silver’s model at all.) Indeed, other projection sites like RCP will be just as affected.
What I would like to see is a post that turns from Silver’s methodology to the state polls that form the basis of his prediction. Digging into the underlying data, CBS, NBC, ABC, and SUSA all show a remarkably similar universe across multiple swing states. (PPP shows similar effects, with a slight young and female skew that readily explains its extra pro-dem numbers). The universe these polls show is one where the D/R split exactly replicates 2008. I dug into each poll and could find *no* exceptions from this split in these polls in the hotly contested states.
I spent a few hours last night painstakingly going through the crosstabs in the hopes of getting a handle on why the partisan numbers look like 2008 and came up empty. The polls often skew ~1% female, but that is far from explaining the difference. The polls all agree that the 18-29 vote will be at 55-60% of 2008, and that the minority share of the vote will be exactly like 2008 (not unreasonable given that 2008 saw little increase in proportional minority vote compared to 2004.)
If we are to believe the leaks, these polls look similar to Obama internals but drastically different from Romney internals. The question is, why are they showing a 2008 D/R split despite the preponderance of evidence to the contrary? And what changed between 2010, when Nate’s model performed well in predicting the Republican wave, and 2012? I do not have a sense of increasing hostility towards pollsters among Republicans this year. The only inklings I have are a possible rise in caller ID among middle to upper class voters, and a possible oversampling of cell phones.
Rasmussen magically drops to an even race today. From looking at the last week’s worth of swing state polls, including today’s, it’s looking like America likes to eat “toast”, know what I’m sayin?
I don’t understand why everyone is making this so difficult.
This election is being contested in 9 states – Florida, North Carolina, Virginia, New Hampshire, Ohio, Wisconsin, Iowa, Colorado, and Nevada. All the other states are effectively conceded to Obama or Romney – 191 electoral votes to Romney and 237 to Obama. So Obama needs 33 to win, and Romney needs 78 (because he wins ties in the House).
Obama can win with as few as 2 of the swing states, and in all combinations of 5 or more. Romney needs a minimum of 5 and can only lock up the election in all cases with 8 state wins.
The nine states lead to 512 possible combinations of victories for either side, with the number of wins varying from 0 to 9.
If we simplify, ignore the polls, and just say every state is a coin toss like Joe Scarborough would like, only 64 combinations of wins give Obama 32 electoral votes or less, and thus produce a win for Romney.
64/512 = 12.5% chance of victory for Romney.
Remember, that comes from ignoring all of the polls and just assuming its all a 50-50 chance.
If you want to say Romney has a better chance, that demands additional evidence, which means looking at polls, and since the polls are pointing to a Romney loss on their face values, reinterpreting or skewing or whatevering the polls to make them support Romney.
As to one final question, the Gallup poll was showing a 5% lead for Romney because they were getting a 22% lead for Romney in the south and a 4% lead for Obama in the Northeast. Romney was going to win most of the south anyway, so a huge lead there hardly matters to producing an electoral college victory, even if he carries Florida, Carolina, and Virginia by huge margins too. The 4% lead in the northeast for Obama is only possible if Romney has the race close in the populous states like New York, Massachusetts, Maryland, and Pennsylvania. The first President Bush won the northeast by 0.84% in 1988 but still lost many of these states. An additional 5% swing would have erased his victories in Pennsylvania, Maryland, Vermont, and Connecticut. Its not hard to also erase the ~12% 1988 Bush margin in Maine and New Jersey while keeping to a 4% overall Obama margin given that Dukakis won West Virginia, and Obama is certain to lose it, and given that no one is expecting a Romney victory in either state. One need only shift 3% of the expected NY and Massachusetts loss margin to Romney to cover losing New Jersey. Gallup’s margins in the West (6%) and Midwest (4%) are entirely consistent with the state polls available, which has Obama leading in Ohio, Wisconsin, Iowa, Colorado, and Nevada. If Obama wins those 5 states, he wins the election, even though Gallup has him losing the popular vote by 5%.
This should be a self evident result. The purpose of the electoral college is to force candidates to have a national appeal instead of a sectional appeal. Rolling up a huge margin in the south, great plains, and Mormon west =/= an election win, because those states don’t have a majority of the electoral vote.
As information, the current state polls in aggregate are pointing to the following margins in each region: South 10% for Romney. West 8% for Obama. Midwest 2% for Obama. East 17% for Obama. This compares to 17%, 6%, 6%, and 1% for Bush in 1988. Romney’s much worse showing in all regions except the south vs. 1988 is why this election is NOT a walk-off blow out for him rightow.
The problem is state polls are not created equal. The big flaw in poll aggregates used via PUBLIC DOMAIN (Silver) is the quality and accuracy of the data is suspect because polls conducted with Newspapers using less than 1K sample sizes with cost cutting corners will not allow you to be able to predict with any confidence of reliability. Otherwise both campaigns that staff well funded (costing millions) private teams would merely read the crap Silver accesses. There is no way Silver can seriously suggest using these on the cheap polls that Obama has an 80% chance to win. He cannot predict turnout on the 7th, which all the data thus far is not showing to be +7. He certainly cannot do so using mixed open source polls with a straight face…
Is it true that Silver way overweights polls that favor Obama?
How can he do this and still retain credibility?
Oh, i forgot, all you have to do is predict an obama win to be credible in the msm.
Hey Crank, you picked the Tigers to win the World Series and we all saw how well that worked out. About as well as you picking Mitt Robme.
Anyone who thinks Teapartiers are self-identifying as Independents is not part of the movement. Every Teapartier I know self-identifies as a Conservative or Conservative Republican.
When I was a true independent, I split my ticket regularly, and voted for Ross Perot and John Anderson (remember him?) for President. Once I became a right-winger, I took out my frustration with establishment Republicans in Primary elections, but still voted for the Republican in general elections.
There are some on the left that Obama has disappointed by his failure to close Gitmo and end the wars sooner, but I doubt seriously that those folks now self-identify as Independents instead of Democrats.
You say undecided voters leaned towards Gore in 2000 because there was a major event, the Bush DUI.
That really wasn’t an event. It was a revelation that Bush, a recovering alcoholic, got a DUI 20 years ago when he was an active alcoholic. I really don’t think that shocked many people. It got a lot of news coverage because it was sensational.
When talking about what a major event that was, you failed to mention there was kind of a major event really late in this election cycle. It was a hurricane that the federal government provided a competent response to. It simultaneously made Obama look presidential while reminding the nation what a dangerously embarrassing debacle the last Republican administration was.
When did I pick the Tigers? My prediction on Twitter was Giants in 5.
Closing comments here due to tsunami of spam, will start an open thread.
Comments are closed.