Indiana

Right now, many sites are surprisingly bullish on Obama's chances in Indiana. This should be surprising to anyone with historical perspective, since IN has been so consistently Republican in presidential elections. During the 20th and 21st century, IN only voted for Democrats in the Roosevelt and Johnson landslides, and in the weird 1912 election when Wilson won IN with a 43% plurality. Throughout the '90s, IN voted against Bill Clinton, even as Clinton won every surrounding state.

Here's a table of how various people view IN:

Source	Estimate of Obama lead	Probability of Obama win
me	-5.9	9.8
538	-1.3	43
Pollster	+1	n/a
most recent poll	+1	n/a
Intrade	n/a	33-37

I'm clearly the outlier here, but I have the courage of my convictions, and maintain that mine is the most reasonable view of IN. Let's look at the available IN data:

Pollster	Dates	Sample size	Obama lead
Survey USA	6/21-23	627	+1.0
IN Leg Thought	5/27-6/1	601	-9
Downs Ctr	4/28-30	1274	+1
Research 2000	4/23-24	600	-8
Selzer-Star-WTHR	4/20-23	384	+8
Downs Ctr	4/14-16	1254	-7
Survey USA	2/26-28	579	-8.9
Survey USA	2/3-4	499	-10.2
Bush vs. Kerry	11/2/2004	2.5MM	-20.7

And here's an aggregation by month. The "margin of error" is a 95% certainty range for the difference between the candidates, assuming a -100% correlation between the two. (This isn't a perfect assumption, and not what I use in my model, but should be good enough.)

month	Aggregate sample size	Aggregate Obama lead	Margin of error
June	627	+1.0	7.8
May	601	-9	8.0
April	3512	-1.9	3.3
February	1078	-9.5	6.0

The story I get from this table is that there is not much information on IN. Just from looking at the IN, one could come up with several reasonable hypotheses, for examples:

• A moderate McCain lead stretching back a few months, with some small fluctuations through the spring and some pro-Obama shift on net since February. In this view, there's been significant error in a few polls, although not so great as to be incommensurate with the polls' published sample errors. This is roughly the view I have, with my estimate that McCain leads today by 5.9 +/- 1.7.

• Obama has had a big surge from May to June, and the race is now essentially tied. All the polls have been valid, but the race is volatile. This roughly seems to be the view that 538 has, and it's implicitly the view that people have who always use the most recent poll, such as many reporters or the map at MyDD.

• The race has been essentially tied for at least a couple months. The polls that were good for Obama are good, and the polls that were good for McCain are bad. I see no evidence to support this idea, but it might not be wrong.
• McCain has led substantially for at least a couple months. The polls that were good for McCain are good, and the polls that were good for Obama are bad. I see no evidence to support this idea, but it might not be wrong.

Now let's look at what 538 and I think are the changes in the D-R spread since 2004 in IN and nearby states. (I don't mean to pick on 538. I'm using it because it's the most sophisticated example of a site that takes the second viewpoint.)

State	2004 spread	Change per me	Change per 538
IN	-20.7	+14.8	+19.4
IL	+10.3	+14.4	+8.8
MI	+4.4	+3.3	+0.2
OH	-2.1	+7.4	+6.5
KY	-19.9	+6.2	+2.8
US	-2.5	+8.3	+6.1

In order to believe 538's results, you must believe that Obama has improved by much more in IN than in the US as a whole, or in any surrounding state. You must even believe that there has been twice as much pro-Obama swing in IN than in Obama's home state of IL. And most importantly, you must believe all this on the basis of a total of scant recent polling in the state -- only 1228 people polled since the IN primary. To me, this state of affair seems neither plausible nor well-supported by evidence.

By contrast, I think my results make common sense. The distribution of shifts across the states makes sense: Obama does best in his home IL and neighboring IN, mediocre in OH and KY (states where he struggled in the primary), and poorly in MI (where he snubbed the primary entirely). Furthermore, none of my current results are inconsistent with available polling.

The reason that my results don't get out-of-whack with each other or with polling data is that my methodology is centered around two fundamental principles:

1. Output should be based on data, i.e., polls.
2. Opinion shifts in different states are highly correlated.

Sometimes these principles conflict, and there are certainly cases where I predict that opinion has shifted differently in different states -- see IN and MI above for example. But in cases like IN/MI, the divergence is strongly rooted in available data. My methodology would rarely (if indeed ever) predict an 11-point divergence in the IL and IN shifts with only 2 stale polls in one state and 8 scattered polls in the other.

Read more (maybe)!

Obama up by 5.8%; 93% chance of victory

Prediction for Election Day

• Probability of victory: 93% electoral, 92% popular. This includes a <0.5% chance of tie.
• Expected value of vote: 340 electoral votes, popular win by 5.8%.
• 95% range of electoral votes: 231 to 458.
• Must-win states (pseudo-Banzhaf):
  • PA .020
  • OH .019
  • MI .015
  • NM .013
• Bellwether states (correlation):
  • 70%+ OH, PA, MI
  • 50%+ NM, NH, CO
  • 40%+ NV, MO, NJ
  • 30%+ WI, VA
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker blue for Obama):

Prediction if the election were today

• Probability of victory: Obama won both electoral and popular votes in all 40,000 simulations.
• Expected value of vote: 339 electoral votes, popular win by 5.8%.
• 95% range of electoral votes: 306 to 375.
• Must-win states (pseudo-Banzhaf):
  • [insufficient sample]
• Bellwether states (correlation):
  • [insufficient sample]
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker blue for Obama):

Popular vote estimate

(Darker red means more votes for McCain, darker blue for Obama.)
Read more (maybe)!

Upcoming Topics

Listed partially as a teaser, and partially so I remember to circle back if I have more time tomorrow.

• Case study: Why I disagree with everyone else on Indiana.
• An insight I had on how public opinion works. (Summary: How I and everyone else are modeling this completely wrong.)

Read more (maybe)!

Obama up by 5.2%; 91% chance of victory

Prediction for Election Day

• Probability of victory: 91% electoral, 90% popular. This includes a <0.5% chance of tie.
• Expected value of vote: 335 electoral votes, popular win by 5.2%.
• 95% range of electoral votes: 225 to 443.
• Must-win states (pseudo-Banzhaf):
  • OH .024
  • PA .021
  • MI .017
  • NM .016
• Bellwether states (correlation):
  • 70%+ OH, PA, MI
  • 50%+ NHNM, CO
  • 40%+ NV, WI, MO, WI, VA, NJ
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Prediction if the election were today

• Probability of victory: >99.5% electoral, >99.5% popular. This includes a <0.5% chance of tie.
• Expected value of vote: 335 electoral votes, popular win by 5.2%.
• 95% range of electoral votes: 298 to 370.
• Must-win states (pseudo-Banzhaf):
  • [none]
• Bellwether states (correlation):
  • [insufficient sample]
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Popular vote estimate

(Darker red means more votes for McCain, darker green for Obama.)
Read more (maybe)!

Methodology

A revised writeup of the site's methodology is now available at this PDF page. It is intended to require little statistical knowledge, but that doesn't mean it's an easy read.

I'll also make a permanent link in the sidebar.

Read more (maybe)!

Obama up by 5.1%; 90% chance of victory

Commentary

I think this is the first update in which Obama has a >50% probability of winning FL. Right now, in light of the Quinnipiac and ARG polls in FL, and polls elsewhere that show a general Obama surge, I have FL at 50.04% Obama, with a margin of error of 1.67% if the election were held today. In fact, it's quite possible that FL will be about as close as it was in 2000, although this time the results wouldn't matter becaule FL is not a swing state anymore.

Prediction for Election Day

• Probability of victory: 90% electoral, 89% popular. This includes a <0.5% chance of tie.
• Expected value of vote: 333 electoral votes, popular win by 5.1%.
• 95% range of electoral votes: ....
• Must-win states (pseudo-Banzhaf):
  • OH .026
  • PA .025
  • MI .022
  • NM .014
• Bellwether states (correlation):
  • 70%+ OH, PA, MI
  • 60%+ NH
  • 50%+ NM, MO
  • 40%+ NV, CO, WI, VA
  • 30%+NJ, FL
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Prediction if the election were today

• Probability of victory: In 10,000 simulations, Obama won both electoral and popular votes every time. [Edited to add: Make that "70,000 simulations."]
• Expected value of vote: 335 electoral votes, popular win by 5.1%.
• 95% range of electoral votes: ....
• Must-win states (pseudo-Banzhaf):
  • [none]
• Bellwether states (correlation):
  • [none]
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Popular vote estimate

(Darker red means more votes for McCain, darker green for Obama.)
Read more (maybe)!

Not a Swing State Watch: Alaska edition

Jonathan Singer at MyDD says that Alaska is a swing state, and so does Kos. I disagree. One poll has come out suggesting that AK may be close. However, that does not make it a swing state. With Obama up by 3-4 points nationally, AK is still several points more Republican than the rest of the country, even if this latest poll is correct. That means that AK is extremely unlikely to make the difference in the campaign (particularly because it has so few electoral votes), and is a poor indicator of who will win the campaign.

When I wrote about how to measure swinginess a couple weeks ago, I didn't comprehend the extent to which many people confuse "swing state" with "close state." Let us be clear -- MN may have been a close state in 1984, but as the safe home state of the Democratic nominee, it was certainly not a swing state.

Read more (maybe)!

Not a Swing State Watch: California edition

I've seen a couple Republicans suggest that CA is in play. I believe that Obama is up by 17 points in CA, and that CA is not in play. Obama has a better chance of winning TX than McCain does in the Democratic big states of CA, NY, or IL.

Suppose, however, that McCain devoted all his resources to CA, Obama did not change his strategy, and McCain was able to take the state. How bad would that be for Obama?

Not so bad. Obama would be able to easily pick up OH, IA, MO, CO, NM, NV uncontested, which is enough for 254 electoral votes. After that, FL is the obvious place to get the rest of the votes, although VA and two NE votes would be enough to squeak by. It would make it difficult for Obama to achieve an overwhelming victory in the Electoral College. Maybe that's what Republicans are playing for at this point -- prevent Obama from a massive victory that would transform American politics. Or more likely, they're just blowing smoke about CA.

Read more (maybe)!

Funny

Here's a funny coincidence. The day after I criticized FiveThirtyEight.com for having stale predictions in OH/PA/CO/etc. even though new polling elsewhere suggested a national surge for Obama, Poblano did something to address that issue. I haven't read closely enough yet to understand what he's doing, but it does bear noting that his expected EV has jumped from 278 to 308, which is now very close to my prediction of 312.

The main practical differences between the two sites is now 1) Poblano is a much better communicator than I am, and 2) I believe that popular opinion is less volatile than he thinks (which leads him to think that the underdog has a better chance than I believe in most cases).

Read more (maybe)!

Electoral College advantage: Obama

I know that I commented on this several weeks ago (and during the 2004 campaign, when conventional wisdom said that the Electoral College favored Bush), but it bears repeating: The Electoral College favors Obama.

The above chart shows the relation between popular vote for Obama (on the X-axis) and probability of victory for Obama in the Electoral College (on the Y-axis). Clearly, the existence of the Electoral College favors Obama. When the popular vote is split evenly, Obama wins about 60% of simulations, and even when he loses the popular vote by a couple tenths of a percent he still wins about half of simulations.

It is well known that the Electoral College overweights small states, which Republicans tend to win. The states with 3-5 votes are: AK, DE, DC, HI, ID, ME, MT, NE, NV, NH, NM, ND, RI, SD, UT, VT, WV, WY. Of these, 39 votes are clearly Republican in an even election (AK, ID, MT, NE, NV, ND, SD, UT, WV, WY), 21 votes are clearly Democratic in an even election (DE, DE, HI, ME, RI, VT), and 9 votes are unclear (NH and NM). This is a structural advantage in the Electoral College for Republicans.

However, the Republicans essentially squander this structural advantage by winning too big in their safe states. Republicans tended to win their safe states by 10-50 points, while Democrats tended to win their safe states by 10-25 points. The one outlier is DC, but in 2004 Republicans wasted far more popular votes per electoral vote in UT than Democrats did in DC.

Democrats were also relatively spare in their winning margins in large safe states too: Republicans wasted 50000 popular votes per electoral vote to win TX, while Democrats wasted only 30000 popular votes per electoral vote to win CA and NY. In fact, I'd say that the key to Democrats' advantage in the Electoral College is their ability to count on winning CA without a very large margin there.

This waste of popular votes is hard to quantify precisely, but I'd say that in a very close election Republicans would tend to waste about 2-3 million popular votes more than Democrats. This is enough to negate the Republicans' structural advantage, plus a little extra.

Read more (maybe)!

Obama up by 3.6%; 81% chance of victory

Prediction for Election Day

• Probability of victory: 81% electoral, 80% popular. This includes a 1% chance of tie.
• Expected value of vote: 312 electoral votes, popular win by 3.6%.
• 95% range of electoral votes: 199 to 421.
• Swing states (pseudo-Banzhaf):
  • PA .051
  • MI .032
  • OH .029
  • NM .016
  • NV .010
• Swing states (correlation):
  • 80%+ PA
  • 70%+ OH, MI
  • 60%+ NH
  • 50%+ NM, NV, MO
  • 40%+ CO, WI, VA, NJ
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Prediction if the election were today

• Probability of victory: >99.5% electoral, >99.5% popular. This includes a <0.5% chance of a tie.
• Expected value of vote: 313 electoral votes, popular win by 3.2%.
• 95% range of electoral votes: 278 to 352.
• Swing states (pseudo-Banzhaf):
  • [none]
• Swing states (correlation):
  • 50%+ PA
  • 40%+ MI
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Popular vote estimate

(Darker red means more votes for McCain, darker green for Obama.)
Read more (maybe)!

Website roundup

A lot of websites make lists that purport to predict the election. I thought I'd look at each of them and make some comments.

This table includes only states that someone thinks are tossups or leaners or between 25-75% chance of each candidate winning.

Code	Me	538	CNN	MSNBC	NYT	Cook	Intrade	US Election Atlas	Average
AR	1.9	0.6	30	10	10	35.7	19.5	30	17.2
CA	100	96.6	90	70	90	92.9	87.9	90	89.7
CO	92	66.2	50	50	50	50	66.75	50	59.4
DE	98.6	92	90	70	90	78.6	90.8	70	85.0
FL	28.6	20.8	50	50	50	35.7	29	30	36.8
GA	2.5	5.4	30	30	10	7.1	17.45	10	14.1
IA	99.2	83.9	50	50	50	50	77	50	63.8
IN	1.7	29.7	30	10	10	7.1	23	10	15.2
LA	7.1	5.2	30	30	10	7.1	13	10	14.1
ME	99.5	91.3	70	70	90	78.6	86.5	70	82.0
ME1	99.9	"	"	"	"	"	"	90	84.5
ME2	97.4	"	"	"	"	"	"	70	81.7
MI	85.7	49.9	50	50	50	64.3	70	50	58.7
MN	99.2	92.7	50	50	50	50	79.1	70	67.625
MO	52.5	34.7	50	50	10	21.4	42.5	50	38.9
MS	0.8	2.7	10	30	10	7.1	13.75	10	10.5
MT	27.4	19	30	30	10	7.1	17.8	10	18.9
NC	17.8	22.4	30	30	10	7.1	25.45	10	19.1
ND	5.5	26	10	10	10	7.1	11	10	11.2
NH	78	45.2	50	50	50	64.3	52	50	54.9
NJ	94.6	86	70	70	90	92.9	86.5	70	82.5
NM	81.4	62.7	30	50	50	50	66	50	55.0
NV	73.6	49.9	50	50	50	50	49	50	52.8
OH	68.7	49.6	50	50	50	35.7	60.4	50	51.8
OR	99.2	91.9	70	50	50	64.3	86	70	72.7
PA	82.9	67.5	50	50	50	64.3	75.5	50	61.3
VA	41.4	35.8	50	50	50	35.7	49.5	50	45.3
WA	99.8	97.6	70	70	50	78.6	90	70	78.2
WI	93.8	76.4	50	50	50	64.3	81	50	64.4
WV	12.7	11.7	30	10	10	21.4	14.5	10	15.0
EC	306.7	278.1	273.8	258.6	257.8	266	291.1	259.4	273.9

Me

I am the most bullish on Obama of the bunch. While I don't profess to understand what most of the other folks are doing (except for 538, which I understand pretty well), I can hazard some guesses about why this is the case.

1. My results are fundamentally based on where the evidence shows the race stands today, without much regard to 2000 and 2004 results. I believe that most major pundits and outlets are still stuck in the framework of the 2000/2004 elections. The best example of this is IA. Poll after poll shows that Obama is clearly ahead in IA, and yet CNN, MSNBC, NY Times, and Cook all have IA as a toss-up. Of course, I can imagine that a pundit like Cook might sniff at this and say that he looks at more than just polls. But what then are they looking at? Seriously, I can't think of a single "soft factor" that would indicate a competitive IA except that Bush happened to win IA in 2004.
2. Pundits tend to think that popular opinion is volatile. No matter how long Obama leads by a substantial amount, no pundit is likely to say that Obama is highly likely to win, because they believe at heart that popular opinion is volatile. Also, the idea of a volatile election makes for exciting TV. CNN in particular very badly obfuscated that fact that Hillary Clinton was clearly losing the Democratic primary, and I expect them to do so in the general, even if one of the candidates is clearly winning. In contrast to the pundits, I don't see much evidence that popular opinion changes quickly once people are familiar with the candidates. I believe that a race that starts out 50/50 is unlikely to be won by much more than 5-6 points.
3. Pundits tend to think that popular opinion is mean-reverting (and that the 2000 and 2004 results represent the mean). This explains why they believe that IA is a swing state but CT is not, even though polling has been similar in the two states, and the two states have the same number of electoral votes. The pundits believe that IA will tend to revert to the 50/50 results of 2000 and 2004, while CT will tend to revert to the double-digit Democratic wins of 2000 and 2004. I actually do believe that people might be slightly mean-reverting, but I don't think that the 2000/2004 elections give enough data to know what the mean is, so my model assumes neither mean-reversion nor mean-aversion.
4. I react quickly to new information. There are new polls released almost every day, and I incorporate each poll into my understanding of where each state is. By contrast, people like Big Media and Cook don't update often (why would they, since they fundamentally base their predictions on 2004?). Also, by contrast, Poblano/538 bases each state's prediction only on that state's polls, so that he's still rather neutral on several key states (OH, PA, NM, NV, CO) that haven't been polled recently, even though Obama has gained a couple points across the nation in the past two weeks. Interestingly, the other site that quickly incorporates new information is InTrade, which is the one site almost as bullish on Obama as I am.

Poblano/538

As mentioned above, the most important difference between me and 538 is that I incorporate information from all polls into my opinion of each state, using some assumptions about the correlations between states in shifts of voter opinion. This means that I don't have to wait around for polling in a particular state (such as OH currently) in order to detect that OH has likely swung to Obama in the past couple weeks; instead, I can observe that the nation has swung overall to Obama, and can infer (with some amount of uncertainty) that OH has as well. Additionally, this means that I can't be fooled as much by an outlier poll in a lightly-polled state. (Poblano sort of gets around this risk by having a constantly weighted input from his regression analysis, but I'm not a big fan of using regression analyses to predict results.)

I currently disagree most strongly with Poblano in that I am more bullish in a number of close states (OH, PA, CO, et.) that haven't had individual-state polls since Clinton dropped out. I'm pretty sure I have better results in each of these cases, since I'm clearly capturing new information (the fact the Obama has improved by a couple points in the polls) that 538 is not.

I also disagree (I'm more bearish on Obama) on IN. Since Indiana is lightly polled, I think Poblano's simulations are getting fooled by a single result from April that showed Obama to be ahead.

CNN

Now we're getting into the poorly done Big Media lists, where I'll just point out the most egregious idiocies.

CNN thinks that AR is only Lean Rep. Evidently they haven't updated their map since Hillary was ahead. They also think that NM is Lean Rep, even though it is much more arguably Lean Dem. How FL and MO are Tossup while NM is Lean Rep is beyond me, since Obama clearly does much better in NM than in FL or MO. And this isn't even a case of extrapolating 2000/2004 forward, since NM was very close in both those years.

MSNBC

I'm going to give more benefit-of-the-doubt to MSNBC, since their categories are "Battlegrounds," "Wildcards," and "Likelies," which I've (maybe wrongly) chosen to translate into tossups, leaners, and safes. They list some states in Wildcards (CA and MS) that should be pretty safe, but that McCain and Obama supporters badly want to find compelling. I suppose that Wildcard might mean "One of the candidates might be daft enough to waste substantial resources in trying to put this state in play, which would make this state interesting to watch," in which case I'd agree that CA and MS are wildcards.

NY Times

This is the best of the simplistic Big Media maps. There aren't any obvious errors, i.e., all the blue states are more pro-Obama than all the tossup states, and all the red states are more pro-McCain than all the tossup states. I'd suggest that blue states like MN/OR/WI are no more battlegrounds than red states like NC/MO, so that NYT's map is too bearish on Obama.

The main bone I have to pick with NYT is that they're so fixated on the 2004 results as a predictor of 2000. Once it becomes clear that IA is not much of a battleground state, I don't understand how they'll represent that it's clearly switched to Democratic. And similarly if it becomes clear that McCain will win NH. My guess is that they'll keep both states as "battlegrounds" even if they're obvious switches.

Cook Report

Charlie Cook either A) has been slow to react to the bounce Obama's received since Clinton dropped out and the Wright controversy fizzled, or B) believes that states will (with a few western and Floridian exceptions) mean-revert to their 2004 results. Those are the only reasons I can think of for him to be generally reasonable but somewhat too bearish on Obama.

Intrade

If you don't already know, Intrade is an Irish website where people can bet on elections. (They pay 0% on their large margin requirement, they charge significant transaction costs, and there's a significant spread in most markets, which are generally illiquid. The world deserves a better election betting site.) Despite their problems, they have some interesting results. In particular, as I've noted above, they are more bullish on Obama than anyone else but me. I think, though, that many predictions there are too close to 50/50, indicating that people believe there's significant volatility. I don't think the odds are close enough to 0/100 in several of the second-tier swing states, such as AR or MN. Best bets: Go long Obama in CO, NV, MN, and short in AR.

Intrade has the closest correlation to the average of any site, and the lowest sum of absolute differences with the average. I take this to mean that InTrade encapsulates common opinion very efficiently. On the other hand it means that it's a decent opportunity to make money if you know more than conventional wisdom.

US Election Atlas

Like Intrade, highly correlated with the average, which indicates that it follows conventional wisdom, at least in how people ordinally rank the states. The average result is very bearish Obama (259 electoral votes), probably for two reasons: 1) The site equally weights predictions that people make at different points in time. 2) When you make a prediction at the site, the default is to start with the 2004 results, which makes people unconsciously more 2004-centric than they might otherwise be.

Read more (maybe)!

Obama up by 3.2%; 77% chance of victory

Commentary

Obama has widened his lead by about 1.5%, and about half of his new votes are from Undecided. This suggests to me that the Clinton supporters who told pollsters they were undecided while the primaries were ongoing have now come home to the Democratic nominee.

Prediction for Election Day

• Probability of victory: 78% electoral, 77% popular. This includes a 1% chance of tie.
• Expected value of vote: 306 electoral votes, popular win by 1.1%.
• 95% range of electoral votes: 192 to 410.
• Swing states (pseudo-Banzhaf):
  • PA .055
  • MI .034
  • OH .033
  • NM .021
  • NV .013
  • CO .010
• Swing states (correlation):
  • 80%+ PA
  • 70%+ OH, MI
  • 60%+ NH
  • 50%+ NM, NV, MO
  • 40%+ CO, WI, VA, NJ
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Prediction if the election were today

• Probability of victory: 99% electoral, 100% popular. This includes a <0.5% chance of a tie.
• Expected value of vote: 308 electoral votes, popular win by 3.2%.
• 95% range of electoral votes: 274 to 349.
• Swing states (pseudo-Banzhaf):
  • PA .010
• Swing states (correlation):
  • 50%+ PA
  • 30%+ MI
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Popular vote estimate

(Darker red means more votes for McCain, darker green for Obama.)

Technical note

I increased my volatility parameter to assume 3.8% annual volatility in each candidate's national popular vote share, up from 2.7%. This is more in line with observed volatility this year. The effect of this is that more recent polls will have more weight, and that the election-day forecasts will be more uncertain (i.e., the expected probability of victory for each candidate will be closer to 50% than previously, all else equal).

Read more (maybe)!

Obama up by 1.1%; 69% chance of victory

Prediction for Election Day

• Probability of victory: 69% electoral, 64% popular. This includes a 2-3% chance of tie.
• Expected value of vote: 282 electoral votes, popular win by 1.1%.
• 95% range of electoral votes: 192 to 357.
• Swing states (correlation):
  • PA 89%
  • OH 71%
  • NH 67%
  • MI 61%
  • NV 58%
  • NM 57%
  • CO 49%
  • MO 47%
  • NJ 44%
  • WI 41%
• Swing states (pseudo-Banzhaf):
  • PA .094
  • NM .029
  • OH .028
  • MI .027
  • CO .016
  • NV .014
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Prediction if the election were today

• Probability of victory: 94% electoral, 88% popular. This includes a 4% chance of a tie.
• Expected value of vote: 288 electoral votes, popular win by 1.1%.
• 95% range of electoral votes: 253 to 309.
• Swing states (correlation):
  • PA 83%
• Swing states (pseudo-Banzhaf):
  • PA .106
  • NM .026
  • MI .029
  • OH .016
  • CO .016
  • NV .012
• Confidence map (states sized by electoral vote, darker red means higher confidence that McCain will win, darker green for Obama):

Popular vote estimate:

(Darker red means more votes for McCain, darker green for Obama.)
Read more (maybe)!