Today's Numbers - September 30

A good day for Obama, and a few additions:



As you can see from the chart, I've added a couple things. Most significantly, the "Win Percentage" chart shows the probability of each candidate winning according to four sources: three markets (Intrade, Iowa Electronic Markets [IEM], and Rasmussen Markets) and the simulations on FiveThirtyEight. Intrade and IEM are real money markets, while Rasmussen is not--but evidently it has a serious, fantasy-footballesque following. Its numbers are very similar to Intrade and IEM (the former of which runs the software for Rasmussen Markets), so in the absence of evidence to the contrary I'll include it.

These numbers show a much bleaker position for McCain than the national poll numbers would seem to present. The traders on these markets are looking at all the polls, national and state-level, and analyzing the underlying trends, and their belief is that McCain has barely a 1 in 3 chance of coming back. At these prices, I would personally think McCain is undervalued. One interesting thing about these markets, though, is that they reflect the change I spoke about previously--that each day without any change is a good day for Obama. If the polls stay fairly stable between now and November, expect Obama's numbers to improve in the markets. Similarly, the race-tightening assumption in the FiveThirtyEight model will have an equivalent effect in their model, and we're beginning to see it already--though these aren't quite Obama's strongest poll numbers to date (he was better in the period between his World Tour and the DNC), it reflects some of the highest win percentages FiveThirtyEight has seen, because of the relative nearness of November 4.

While I was at it, I also included IEM's popular vote contracts in the popular vote count. As you can see, they are significantly higher than the other poll-based measures. This is because IEM's figures are the percentage of the two-party vote, excluding votes for Nader, Barr, and others. While this explains why in absolute terms the numbers are higher for both candidates, it doesn't help explain why the margin is so great between the candidates; excluding others would add anywhere from 1-4% to the totals, but in terms of margin, there should be little impact because the percentages that would otherwise go to third parties would be allocated proportionally, leading to an increased margin of maybe a tenth of a percentage point or two. It seems as if the traders on IEM expect Obama to open up a much wider margin between now and November 4--which is an interesting expectation, considering the opposite assumption made by FiveThirtyEight's model (which, as it happens, is starting to catch up to the national polls in terms of popular vote). You can see here how one simple assumption in a model (the direction popular vote numbers will take over time) can lead to significant differences in the ultimate predictions.

New state-level polls are beginning to trickle in as well, and so we're seeing some movement in Obama's direction in terms of electoral vote predictions. (Check Pollster or FiveThirtyEight for summaries.) In the next few days I may drop CNN and/or the New York Times from these models, because of the deadweight factor (they don't seem to be updating on any sort of regular schedule), but I'd be interested in hearing other opinions. Is it a problem to have a less-frequently updated tracker included in the averages?

Soon I hope to start posting extended reviews on the individual trackers, over the next week or two. Stay tuned.

Today's Numbers - September 29

Another stable day:



Still waiting for a new batch of state polls. There were a few that came out today, but not enough to create a change in any of the state-poll based trackers. Tomorrow should be a different story, because this close to the election there are a lot of pollsters in the field and we haven't seen anything significant since Friday.

The national polls stayed pretty much the same today as well, with an insignificant move in Obama's favor. But if we're talking in terms of likelihood of winning, any day where nothing happens is a good day for Obama, because that's one less day when McCain has a chance to alter the direction of the race. Unless something dramatic happens that would shift the race in McCain's favor (a new bin Laden video, perhaps?), expect another out-of-nowhere move from him along the lines of the Palin pick and "suspending" his campaign. There are growing calls from the right to dump Palin now before things get any worse, but Nate over at FiveThirtyEight had a great post yesterday discussing why that's unlikely to happen. What else McCain could do, I'm not sure, but I wouldn't be the least surprised if he decides to "go nuclear" and focus mainly on a lot of the personal issues with Obama that we heard in the primaries--Jeremiah Wright, etc--and stop arguing the issues entirely. Many pundits have called this race a referendum on Obama for months now (since voters seem to think they know McCain well enough already), so it wouldn't be surprising for the McCain camp to try and kick his unfavorable ratings up to the point where he'd be unelectable, with McCain as the supposedly-safe default choice.

If you are over at FiveThirtyEight, Nate also has some interesting posts about the politics of today's bailout vote. I'm not sure I agree with his projections about what will happen in the future, but it's an interesting conversation-starter. (Though I have to give Nate a little grief for the erratic scheduling of his updates--sometimes he posts them in the early afternoon, sometime not til late, so that's why today's numbers from him aren't included in these numbers yet.)

Over the next few days I hope to add a few more trackers to here, including NBC Political Director Chuck Todd's electoral map, as well as a fourth chart that includes the various futures markets' estimated win probabilities. And speaking of markets, for what it's worth, Intrade finally fixed the Average/Leaning problem I discussed yesterday, so you can interpret those numbers the way that they're intended.

Today's Numbers - September 28

Not much movement today:



From yesterday, the FiveThirtyEight EV numbers have changed, as have the Intrade (Leaning) numbers and the RCP numbers with tossups allocated. (Note that the FiveThirtyEight numbers actually skipped a day--they hadn't posted yesterday's simulations by the time I updated, but they've already posted today's.)

The FiveThirtyEight movement comes almost entirely from Obama's improved numbers in the national polls; their simulations incorporate national polls in producing their state-by-state results. The Intrade movement is the product of small changes in their market prices.* And the RCP change comes from one state poll, in North Carolina, that moves that state from just barely McCain to just barely Obama. It's still a tossup, however, so that's why their "with tossups omitted" number doesn't change. Otherwise, everything else stays much the same, mostly because there are very few state-level polls released on the weekend, and this weekend is no exception--nothing remarkable at all. Prepare for a good number of new polls tomorrow, though, that should impact the numbers over the next couple days, probably in Obama's direction.

On the popular vote front, all three numbers move in Obama's favor, owing to another good day in the national trackers.** One interesting note from these polls is that today is the first day that post-debate interviews have been included. Though most of the trackers do not break down the day-by-day results, Obama's strong performance (maintaining or increasing his lead in all of them) strongly suggests that voters' preferences were either not affected by the date or else shifted moderately in his favor.

* - For those who don't know, Intrade is an electronic market where users bet (real) money on the outcome of specific states and of the nation as a whole by purchasing shares which pay a fixed amount if a specific candidate wins. The price of those shares, like the price of a stock, changes as the perceived value of the share (which is the likelihood of that outcome occuring) changes. The "Leaning" figure allocates states in full to the candidate who's favored in that state. The "Average" figure awards each state's electoral votes in proportion to the predicted likelihood of each candidate winning that state. So, for example, if Obama were considered 60% likely to win North Carolina and McCain 40%, the Leaning figure would allocate all 15 electoral votes to Obama, while the Average figure would award 9 to Obama and 3 to McCain.

However, and I don't know why nobody's noticed this yet, a quick count of the states on the Intrade homepage seems to suggest that they have the numbers reversed. 311 is the Average figure and 297 the Leaning for Obama EVs, but if you count up all the blue states on their map, it sums to 311 EV. Somebody should get on that, but for our purposes it doesn't matter--they're both included in our average, so the overall numbers don't change.

** - As Nate at FiveThirtyEight noted today, their popular vote numbers are most likely always going to be closer to 0 than the other sites', because their model specifically includes an adjustment that predicts the race is likely to tighten over time. He notes that if the election were held today, their model would predict a 4.2% Obama victory, as opposed to the 3.3% victory predicted for 37 days from now. The 4.2% figure falls nicely in line with both the Pollster and RCP figures shown above.

Today's Numbers - September 27

First, the numbers:



Right now, the race looks very good for Obama. This is mostly due to a handful of factors going his way lately--McCain's convention bounce fading, the public's disillusionment with Sarah Palin, and the focus over the last two weeks on the economy, an issue on which Obama has a double-digit lead over McCain. The results out of today's major tracking polls (Rasmussen, Gallup, DailyKos/Research 2000, & Diageo/Hotline) all showed 5-6 point leads for Obama.

The main thing to look for in the next few days is whether last night's debate changes anything. In polls released immediately after the debate, Obama seemed to come out on top. In particular, public perceptions of his readiness to be commander-in-chief (one of his weakest points) increased dramatically. We'll have to wait and see whether this turns into a lead in voter preferences, particularly one that lasts more than a day or two. By the time we have Wednesday or Thursday's numbers, the picture should be more clear. After that, the VP debate Thursday night will start having its own effects--that should be interesting in its own right.

My guess? Obama may pick up a point or so from the debate, but if the bailout plan passes soon and the economy stabilizes over the next week, it'll be a wash.

Finally, the popular vote numbers are usually a few days ahead of the electoral vote numbers, because national polls are released daily (between the daily trackers and other polls), while individual states might go a few days or even a week or two between polls, depending upon the competitiveness of the state. As such, the electoral vote numbers you see above are probably a little lower than they will be in a few days (with the exception of the FiveThirtyEight.com numbers, which incorporate national polls in the state-by-state simulations). To get a sense of what we might expect a national lead of 3.7 points to translate into, I'll leave you with a look at past elections' popular vote differences and electoral vote differences:

Year: 2004 2000 1996 1992 1988 1984 1980 1976 1972 1968

PV margin (GOP - Dem): 2.4 -0.5 -8.5 -5.3 7.8 18.2 9.7 -2.1 23.2 0.7

EV margin (GOP - Dem): 34 5 -220 -202 315 512 440 -57 503 110

As you can see, a small change in the popular vote can make a huge difference in the electoral college. If Obama can hold on to his 3.7 point lead, he'll probably win by at least 50 EV. If the lead fades by 2 points or more, it's anybody's game. If he can build on it and get it up to 5 or 6, though, we could be in for a blowout. Remember, though: we still have 3 debates and over 5 weeks to go. Anything could happen.

A "poll of polls of polls"? Seriously?

Yes, seriously--sort of. The idea did start out as a joke. More specifically, it started out as a sarcastic voicemail left after hearing Wolf Blitzer promote yet another edition of CNN's "Poll of Polls".

After hanging up, though, I realized it made a twisted sort of sense. This election, there are a multitude of election trackers--sites which combine poll results to predict an overall outcome--using a wide range of methods for combining polls and producing vastly different results. Some of these, like RealClearPolitics and the New York Times Electoral Map, simply average poll results and award states to the candidate who leads in each. Others use more complicated methods, such as regression estimates (Pollster.com) and Monte Carlo simulations (FiveThirtyEight.com). Each tracker also differs in which polls they include in their models and which they leave out. Because each method produces different results, it's worth seeing what happens when we average over the methods to produce a new total result.

Though it all might sound ridiculously meta, there is a logic behind this. Trackers combine polls from a variety of pollsters because each pollster uses different techniques. They vary in their question wordings, their survey methods (live phone vs. automated phone vs. internet vs. ?), their response weightings (to produce representative samples), their likely voter models (if any), and many other factors. Such factors can have a severe impact (called a "house effect") on the pollster's results. By combining myriad pollsters together, trackers hope to compensate for house effects, because on average pollsters will cancel each other out. In the same way, this site seeks to get closer to the "true" state of the race by letting the effects of different tracker methodologies cancel each other out. We know that each tracker puts out a number that is their best guess of where the race stands, and that it is incredibly unlikely that any of them will get it exactly right. More likely, the true will result will be off to some degree in one direction or another. By using all of the trackers, we can make an educated guess as to which direction each tracker will be off by, and therefore (hopefully) come up with a better result than individual trackers can on their own.

Or, for the more practically minded, there is another, simpler justification for this site. I'm a bit obsessive, and spend my time each day going back and forth between a collection of many different sites. I know a lot of you do as well. By giving one place for you to go instead of many, this site will save you some time, time that you could perhaps be spending out campaigning, trying to save the world. Or sleeping--that's important too.

HOW THE RESULTS ARE CALCULATED: Each day (or so), I'll be pulling the most recent results from an assortment of trackers. They'll be broken up into three categories: electoral votes with tossups allocated to the leading candidate, electoral votes without tossups allocated, and popular vote estimates. In the the first category, each tracker's electoral votes will add up to 538 (the number of votes in the electoral college). The second category is necessary because some trackers do not allocate "tossup" states--states that are too close to call--and thus produce numbers that sum to less than 538. These numbers should not be read as a prediction of the outcome of the election, but instead as the minimum number of electoral votes each candidate should be expected to receive. The final category is popular vote estimates, estimated from either national poll numbers (for RealClearPolitics and Pollster.com) or from simulations of overall vote patterns (in the case of FiveThirtyEight.com). For each category, average results will be computed. These averages are unweighted, though this may change at some point (at which point I'll post an entry describing the changes and update this entry).

I also hope to be able to contribute some insight on the state and direction of the race, to give some analysis of events and polls as they happen, and to provide some perspective as to the advantages and disadvantages of the different methods of polling and poll-tracking between now and November 4.

Lastly, I thought I should give a bit of information about myself, for those of you wondering who's behind this and where my "insights" and "analysis" come from. I am a 3rd-year PhD student, studying American and Comparative Politics at a major research university in the northeast. My specific focus includes voter behavior, public opinion, and electoral strategies. What this means, in practical terms, is that I spend my days neck-deep in survey data and statistical analysis methods and programs. My goal with this site, then, is to take that training and use it to make the competing methodologies of professional polling comprehensible to the non-professional.

If you have any suggestions for topics you'd like to see covered, know of any poll trackers I'm missing (there are many, I'm sure), or just want to give feedback, send me an email: andrew [at] poll3.com. Otherwise, enjoy the site!