In the latest Curmudgeon’s Corner Sam, Ivan and Bruce talk about:
Recorded on 21 Oct 2012 Length this week – 1:02:24 |
 |
|||||
|
||||||
|
One change today. It is Florida. It is big. It moves from Leaning Romney to Leaning Obama… but… there are a lot of reasons this move doesn’t represent anything that is actually significant. At least not yet. So lets get into into it:
So, according to my five poll average, Florida moves from an 0.2% Romney lead where it was yesterday, to a 1.8% Obama lead today. But… there is plenty of oddness here. First off, I do a “last five poll” average, with “last five” based on the end dates of the polls in question. Well, the most recent end date we have for polls is Thursday October 18th. But we are reaching the time in the race where some states get polled a lot… and so in Florida I now have SEVEN polls in my spreadsheet that ended on the 18th. So how do I pick which ones to include in the five poll average? You could imagine all kinds of tiebreaker criteria, or you could imagine including more than five polls in situations like this. I essentially go with what is simplest for me to process, rather than something that I can justify from a “this is logically the right way to do this because X”, namely, I use the order I found them and added them to my spreadsheet. So the ones I found out about first in my normal daily sweep of sources age out first. I admit that is kind of questionable and other approaches might be more valid from a theoretical point of view. Some are more complicated, like looking at both the start and end dates of the polls, or looking at the sample sizes of the polls, or things like that. I haven’t been keeping the data necessary to do that, so that would be a bit time intensive. So, what can we look at quickly. First of all, if we looked at all seven of the polls that ended on the 18th instead of just five of them… We would have an Obama lead of 0.1%. OK, what if we looked at all of the polls that ended after the second debate (8 polls)… We would have an 0.6% Obama lead. What about the last 10 polls instead of the last five? Now we would have an 0.7% Obama lead. What about if we just took any poll ending in the last week? Now we would have an 0.9% Obama lead. Well, OK, that is interesting, look at that. In all of these cases, we now have Obama leads. So maybe my choice of “last 5” isn’t really that important after all. The trend is indeed moving back toward Obama, and the chart of Florida polls shows that nicely. But… Another thing to look at… the spread… of the 7 polls that ended on the 18th, we have a range from a 5% Romney lead to a 6% Obama lead. That is quite a spread. The average is slightly on Obama’s side… slightly… but there is a large variance. And even if there was no oddness with how I choose which polls to include in the average, what do we have here? My straight five poll average gives a 1.8% Obama lead… which is still very very close. That is a sliver of a lead that could disappear with the very next poll. Bottom line, Florida is too close to call. All of the states in my “Lean” categories could easily go either way, but Florida remains one of the closest. So, where does this put us:
I won’t delve into all the details because I’m slightly behind schedule today, but Romney’s easiest path to victory is pull Florida back into his column and then win Ohio. This has been the primary scenario for a little while now. We’re off a bit from Romney’s highs, but Obama hasn’t yet started pulling states firmly back into his column. So this could still go either way. We’ll start seeing the final view of what this election will look like once the post 3rd debate polls start coming in. By the time we get a full picture on that though, election day will be upon us. And of course early voting has been going on for weeks now. We’re right up against the end now, with Obama remaining a favorite, but by a slim margin. Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
One state changes categories today, and it is a move toward Romney:
Romney had a lead in the five poll average in New Hampshire early in the year… back in January and February… but since then Obama has maintained a consistent lead in New Hampshire, sometimes close, sometimes considerable. Until now. After Obama peaked at a 9.9% lead at the end of September, Romney now takes a 0.9% lead in New Hampshire. And yes, for those asking, that does include one out of five polls with an end date after the second presidential debate. So, new summary:
While still not as well off as his peak on October 12th, Romney has been moving some states in his direction the last couple of days. While Obama is still ahead, the race remains extremely close. Today let’s once again look at the state of the five poll average in all of the close states. In order by the strength of Romney’s position:
With just the states Romney is ahead in by more than 5%, he gets to 191 electoral votes. Add the four states he is ahead in by margins less than 5%, and he gets up to 252 electoral votes. Those four states are all close. They could slip away. But if Romney holds them, that leaves him 17 electoral votes short of a tie, and 18 electoral votes away from an outright win. What is the easiest way to get there? Yup… Ohio. Obama is still ahead in Ohio, but by a razor thin margin. And that margin has actually decreased from 2.8% before the 2nd debate to 1.6% now. If Romney pulls ahead in Ohio, that gets him to 270 electoral votes… the bare minimum needed to win outright without throwing it to the House. I classify all nine of these states as states that could easily go either way. 5% is a margin that can disappear in a day or two with one candidate or another having a bad news cycle. Despite Obama’s razor thin lead at the moment, this really is anybody’s game right now. Obama is one bad day away from actually being behind for the first time. Obama’s performance in the second debate may have stopped his slide and prevented Romney from taking a definitive lead, but so far at least, we don’t have evidence of Obama pulling ahead in a way that would return him to anything close to his late-September numbers. Instead, we have Obama with a small but real lead, and Romney within striking distance. Just over 48 hours until the third debate, and then we’re in the final stretch… Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
Two states move today, one toward Obama, one toward Romney. As usual, in order of Electoral College strength:
In the last three months Nevada has usually been “Leaning Obama”, meaning Obama has the advantage, but not by very much. In September Obama’s lead headed higher for a little bit, then dove again after the debates. Now it heads up again, with Obama’s lead in the five poll average now at 5.2%. Given this, Nevada now is classified as “Weak Obama” meaning it would take a pretty significant move for Romney to take the lead and win the state. So Nevada is no longer included in Romney’s “best case” scenario.
Pennsylvania is a bit odd. Three days ago Obama’s lead topped 5% in the five poll average. But today we add a new poll that arrived out of order. It was still in the five most recent polls by the last day of polling, so it counted as a “recent poll” for us, but it was also the first poll since February showing Romney ahead in Pennsylvania, and showing him ahead by a relatively substantial 4% at that. This is way out of line with other recent polling in the state, so certainly LOOKS like an outlier. Especially since by the time it was released there were already polls taken after it that did not show that kind of movement. But I don’t do any outlier removal here. I just let the five poll average do what the five poll average does. For the moment, this means that Obama’s lead in Pennsylvania is now down 2.8%, with the peak above 5% erased. Because this was still a “current poll” I do not retroactively adjust the overall historical trend however. This potential outlier is now the 5th oldest poll I have in Pennsylvania, so it will age out with the very next Pennsylvania poll. Pennsylvania may or may not pop back to “Weak Obama” at that point depending on the results of that next poll. So this puts Pennsylvania back in play, with the possibility of Romney winning the state now back in his best case scenario. Which brings us to the updated summary:
We still have a close race. Obama is ahead, but Romney only needs to pull 21 more electoral votes over to his side to win. Right now the easiest way to get there is to hold North Carolina (15 ev, 3.8% Romney lead), Virginia (13 ev, 2.6% Romney lead), Florida (29 ev, 0.4% Romney lead) and then to pull ahead and take New Hampshire (4 ev, 0.5% Obama lead) and Ohio (18 ev, 2.8% Obama lead). There are many other combinations that will do it too. That just looks like the easiest to me at the moment. Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
Only one change today, Montana moving toward Romney:
So, although there was some speculation about Obama contesting Montana in 2008… and he did end up making it close, losing the state by only 2.3%… this time around nobody has been talking about Obama trying to win Montana. Although there was one individual poll in April showing Obama only behind by 2%, the five poll average has never shown Obama closer than 6% behind Romney. For most of the year Montana has been in the “Weak Romney” category, meaning Romney is relatively safe, but if something major happened… perhaps a mirror image of the Obama decline after the first debate… then the state might become competitive. Instead, we see Romney’s lead on an increasing trend over the past few months, and with today’s update, Romney’s lead is now at 10.7% over Obama. So Montana moves to “Strong Romney”… essentially meaning that even if Romney completely implodes, he’d probably still win Montana. Since Montana isn’t even close to the competitive zone, this does not affect our models:
The race remains close. Obama is ahead. Obama has always been ahead. Romney still has NEVER taken the lead in my model. But if Romney wins all the states he is ahead in, he is only 21 electoral votes short of a 269-269 tie. He can get those 21 electoral votes in a variety of ways. Assuming Romney holds his current leads in Florida (29 ev, 0.4% lead), North Carolina (15 ev, 2.8% lead) and Virginia (13 ev, 4.0% lead), probably the easiest way to a victory right now is to win Ohio (18 ev) where he is only behind by 2.8% in the five poll average today, and then win any one of New Hamphire (4 ev, behind by 0.6%), Colorado (9 ev, behind by 1.3%) or Iowa (6 ev, behind by 2.5%). Of course, if he is winning Ohio, he will probably be close in Wisconsin (10 ev, 2.8% behind) and Nevada (6 ev, 3.9% behind) as well, and either of them would do the trick too. As of my sweep of polls a few hours ago, we only had TWO state level polls that finished after the second presidential debate. One in North Carolina and one in Ohio. The North Carolina one slightly improved Romney’s position but not enough to recategorize North Carolina, while the Ohio one had no affect on the five poll average at all, as it matched the margin of the poll it replaced in the average. Bottom line, no where near enough data to judge the effects of the second debate. Maybe in the next few days. Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
Three states move between my categories today. One moves toward Romney, two move toward Obama. That sounds like a mixed day, but on balance the benefit goes to Obama. As usual, lets look at the states in order by electoral college weight, but now I’ll zoom in to only show 3 months of polls rather than a year since we are getting close to the end (under three weeks!) and polls are coming in so quickly…
First up, Colorado. According to the five poll average, the state went from an Obama lead of 4.9% on September 23rd, down to a Romney lead of 1.3% on October 8th. With today’s update, Obama takes a very slim 0.5% lead again. This is essentially a tie, but with this the state tips ever so slightly back onto the Obama side of the ledger.
Georgia is Romney’s one piece of good news for the day, but there are a bunch of caveats. Romney’s lead in Georgia hits 10%, so we move it to the “Strong Romney” category. But this is largely due to one poll from September that looks like an outlier, so absent other movement backing up that poll, Georgia is likely to revert back to “Weak Romney” once that poll ages off. If we even have enough more polls in Georgia before the election to do that. It has been pretty sparsely polled. More to the point though, it doesn’t really matter if Romney’s lead in Georgia is just under 10% or just over 10%. Either way, Georgia is a Romney state, and there is no realistic scenario where Obama wins Georgia. Georgia is safe for Romney. So while winning by a bit larger margin may be “nice” for Romney, it doesn’t actually help him at all in the contest with Obama.
Obama’s lead in Michigan topped out at 9.7% on September 20th. It declined to a 4.0% lead on October 8th. It has now popped back up to 6.1%. It seems that the post-debate movement toward Romney was relatively short lived. In the last three months the degree of Obama’s lead in Michigan has bounced around quite a bit. It has however always been a lead. A few isolated polls aside, Romney has never held the lead in the five poll average in Michigan. For awhile there it looked like Michigan was close. Not so much any more. All of this leads us to the new summary:
Since Georgia doesn’t affect any of these three situations, what we see here is basically just Obama strengthening again. Some of the gains Romney had made in the last couple of weeks are just slipping back away from him. Yesterday Obama’s lead in Pennsylvania once again became substantial enough to not call it a close state. Today the same happens to Michigan. And Obama retakes the lead in Colorado. It certainly looks like Romney peaked between August 12th and August 15th and things are now heading back in Obama’s direction, without Romney ever actually taking the lead in the electoral college model. If initial reactions to the second debate are any indication, at the very least we should expect no further downward plummeting by Obama, and perhaps even more additional upside for Obama beyond that. Caution needs to be urged in interpreting this though. Only a small number of states have moved so far, and just barely over the respective category lines. More polls could easily reverse the recent changes. If these initial trends continue over the next few days though, we’ll get a very solid picture. None of the polls included so far include the time period after the second debate. That will of course be the next thing to watch for. Will there be big noticeable effects from the second debate as there were after the first one? Of course to win, Romney needs to do more than just hold on to his previous gains. He actually needs to bring even more states over to his side in order to win. If he couldnt’ do it last week, then what exactly can he do now to make it happen? At the moment Obama’s lead is still very narrow. Romney has many potential ways to win. It is still a close race. But the movement seems to be shifting toward Obama again. Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
After a few days with no changes, one state changes categories today and hey, for the first time in awhile, the movement is toward Obama:
Obama’s lead in Pennsylvania in my five poll average peaked at 9.4% back in mid-September. On the eve of the first debate, this was down to 8.6%. By October 11th, that lead had dropped to 3.8%. Romney actually made Pennsylvania close… but since then the polls have started looking better for Obama again, and the five poll average rises back up to a 5.4% Obama lead. This pulls the state back out of my swing state category and into the “Weak Obama” category, but just barely. Note that even in the dip, not a single poll showed Romney ahead in Pennsylvania. The last time any poll showed Romney ahead in Pennsylvania was way back in February. Although Obama’s lead dipped in Pennsylvania in the post-debate period, a Romney win here seems pretty unlikely if this is indeed the end of the downward post-debate trend. There are too few polls as of yet to know if this is an actual reversal of the recent trend, or if it is just the polls being jittery. For the moment though, the five poll average pulls the state out of Romney’s best case:
Romney’s best case is 20 electoral points weaker, but his overall position is still way better than it has been for most of the year. Obama is still ahead. Romney has still never actually pulled ahead in the electoral college analysis. But it is close. It is still very close. The question remains what it has been for at least the last few days… Has Romney peaked? Or can he pull a little more support in his direction? Right now Romney is behind, but barely. Just a LITTLE more movement in his direction would have him in the lead. But the first debate had already caused a very large movement compared to where the race stood for most of the year. Is there really room for more movement? Or has Romney hit a ceiling? The move toward Obama in Pennsylvania may well be the start of a more general reversion in the polls to something more like the Obama leads we’ve seen most of the year. Or it might not. Too soon to tell. But with the second presidential debate in less than 16 hours, whatever trends may have been underway up through today will get replaced by reactions to the new news cycle. If Obama puts in another sub-par performance, maybe he drops further and Romney takes the lead. If so, Obama is in real trouble. Otherwise, with an “OK”, “Good” or “Great” performance from Obama, we’ll probably start to see some more Obama gains in the polls as those who moved away from him in the last two weeks think to themselves “OK, maybe he’s not that bad after all.” In that case, Obama retains his position as favorite, and maybe even starts making it look inevitable again if he does really well. So… debates almost never make a difference, except when they do… and this next debate is another critical one. So we’ll see what happens… Note: A poll added today in North Carolina that were older than the “last five” changed the date North Carolina last moved from Lean Obama to Lean Romney from October 4th to October 8th. The historical trend chart has been adjusted starting with today’s update. Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
Three states changing categories today. All three moving toward Romney. As usual, lets look at them from smallest to largest in electoral college strength:
First up, New Hampshire. Before the first debate Obama was up to a 9.9% lead here. Now with 4/5 polls after the debate, that lead slips to 4.6%. This is in line with the drops we’ve seen in a bunch of other states. Obama still has a lead here, but it is now small enough that it could easily slip away. New Hampshire has been here before. It had only spiked to a big Obama lead a few weeks ago, when Romney was at his nadir. So this is perhaps New Hampshire just returning to form as a close state leaning slightly toward Obama.
Last time Virginia changed status, I noted that it had been bouncing between Lean Obama and Weak Obama, but perhaps it was ready to break out of that pattern. Indeed it has. This time Obama’s lead continued to drop. After peaking at 5.1% a few days before the debate, Obama’s lead slowly disappeared, and with today’s addition, the five poll average now stands at a 1.0% Romney lead. Romney has been ahead before in Virginia, but it was very short lived. The question of course is if this time he can hold the lead or if Virginia will revert to “normal” which is a narrow Obama lead.
Just yesterday Obama had retaken a very slight lead in the Florida five poll average. Today that slips away, and with some relatively strong polls, Romney ends up with a 1.8% lead at the moment. Florida is still very close. All the states we have talked about today are very close. But the momentum here, as in most other states in the last week and a half, has been toward Romney. So, the revised summary:
We are still seeing what is fairly obviously fallout from the first debate. It has just taken different amounts of time for there to be enough polling in different states to show it. All considered, it looks like we have seen one of the biggest debate effects ever in a presidential election. The level of self-inflicted damage by Obama is amazing. Of course, when this started, Obama had a substantial lead. So even after all of this, Obama is still ahead and is still the favorite. Surely we must NOW be close to a ceiling for Romney, right? Well, maybe not if Obama doesn’t redeem himself in the next debate. But assuming the absence of another catastrophic fail, it seems like sometime soon we’ll need to start seeing some reversion to the mean, which in this case would mean a bit of strengthening for Obama. I don’t have the time today to again do a rundown of the current margins in all the swing states, but suffice it to say that if Romney holds on to all of the states he is currently ahead in, Romney only needs 13 more electoral votes to win. He can get that in a number of ways at this point. The easiest way is still to win Ohio (18 ev). Romney is currently behind in Ohio by only 2.2%. Ohio is within reach. Romney just needs to convince a few more people, or energize greater turnout. Alternately, and a lot more fun, Romney could win Iowa (6 ev) where he is currently behind by 3.2% and Nevada (6 ev) where he is currently behind by 1.6%. That would result in a 269 to 269 electoral vote tie. Assuming no faithless electors, that would throw the election into the House of Representatives, where Romney would almost certainly win. The last time a Presidential election was thrown to the house was 1824. That would be so much fun to watch happen! It is still unlikely though. But a guy can hope! Bottom line though, right now Obama is ahead, but only by the very slimmest of margins. If Romney can hold on to his gains from the last couple of weeks (Obama actually peaked a little before the first debate) and move things just a LITTLE more, then he can win this. I am quite sure Obama would like the Romney “debate bounce” to be over now. We’ll start seeing polls that factor in the VP debate any time now. It is unlikely to have as profound an effect as the first debate and therefore I don’t anticipate Obama erasing Romney’s recent gains. But perhaps the bleeding will finally stop. Note: Chart and map from the Abulsme.com 2012 Electoral College Prediction page. Both assume Obama vs Romney with no strong third party candidate and show polling as it currently exists. Things will change before election day. On the map red is Romney, blue is Obama, gold states are too close to call. Lines on the chart represent how many more electoral votes a candidate would have than is needed to tie under several different scenarios. Up is good for Obama, Down is good for Romney.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
