Vote In Our Poll -- Movie Night at the White House

Don't forget to vote in our "Movie Night at the White House" poll and encourage all your friends and colleagues to vote.  The poll will remain open for approximately one week.  And during the last week of the campaign, we'll have a couple of polls to allow you to contribute the "wisdom of crowds" in forecasting the election results.  The premise of our poll is that election night is going to be a downer for Democrats so the President and Vice President decide to watch a movie at the White House instead of the returns.  Which movie will they pick? Oh, and if you think we omitted one we should have included, please email your selection to us at electionmarkets@gmail.com.  Now to the results.

Today was another day of basically sideways movement.  Republicans remain solidly positioned to take control of the House posting an 85.3% probability of success today.  The median estimate for pickups was 55 in today's run (177 is the base from which the InTrade contacts work).

The Senate side picture isn't quite as bright for Republicans as Democrats have clawed their way back from the brink of minority status over the past week.  Our simulation of all 37 races, which we run 10,000 times for each of four pricing points, suggests that Democrats should be considered a narrow favorite to elect the Senate's Majority Leader.  The data does suggest that if they maintain control they are likely to be electing a new Majority Leader as Senator Harry Reid has moved from being a slight favorite to a slight underdog with just over two weeks remaining in the campaign.   If and when InTrade offers a contract on who the Majority Leader will be we'll attempt to offer an analysis of that race.  The top candidates appear to be Senator Charles (Chuck) Schumer from New York and Senator Richard Durbin (the current Majority Whip).  Should such a contest emerge, it will be interesting to see how Democrats explain a vote for the man often dubbed the "Senator from Wall Street" after pillorying the financial sector during the campaign.

Democrats' prospects of maintaining outright control of the Senate (not counting Independents who caucus with them) has staged a mini rally over the past week.  The rally has come largely on the back of market perceptions that Democrats have improved their position markedly in several key Senate races.  Our model still suggests that the battle for the Senate will come down to between four and six seats.  The broader group of six, represented in the table below, includes California, Colorado, Illinois, Nevada, Washington, and West Virginia.  Another seat that polling suggests might deserve to be in this mix, but not our market analysis, is Alaska.  Polling suggests that thre is increased risk that Senator Murkowski's write-in candidacy has given Democrats a fighting chance in Alaska.

Our micro and macro models are currently projecting a post-election Republican seat count of 48 or 49.  As we mentioned in a previous post, an interesting factor to watch in connection with the Senate contests is the projected Republican gains in the House.  Many commentators have suggested that it is within the realm of posibility for Republicans to gain in excess of 70 seats.  Our model doesn't project that high of a gain at this point, but it could by election day.  Often some of the largest moves in market perceptions around the margins appear in the last week of the campaign.  If the trend is strong enough to produce Republican gains in the House of more than 70 seats, then it is very likely that Democrats will see their position diminish significantly in the Senate.

The chart to the left shows our micro model's trend over the past month.  The question is whether this represents a temporary step back of just a consolidation.  You will notice that prior retreats during the past month have been quickly followed by a resumption of the upward trend. 


The two charts below show a comparison of our micro model's outcome distributions for today compared with a month ago.  While the press coverage of the election has presented a story line of a Democratic surge, compared with a month ago, Democrats are in a more precarious position today than then.

Make sure you vote in our poll for "Movie Night at the White House" and encourage others to do the same.

Daily Update: Basically Sideways; Dems Improve in Key Races

Today's update is limited to just the numbers since tomorrow will bring a far more extensive analysis of the trends in the various races and the various combinations of outcomes.  Suffice it today, House Republicans maintained an 85% probability of control in the next Congress; in the Senate, the movement at the macro and micro model levels was basically sideways, BUT Democrats did improve in key races in West Virginia and Washington.  Republicans showed improvement in Nevada and California.  It will be interesting to watch markets absorb a large amount of new (and in some cases contradictory) polling as well as recent debate performances.

Tomorrow promises at least one distinct positive -- My beloved Buffalo Bills will end Sunday without a loss for the first time since the regular season began.

As always, send me suggestions and questions by emailing electionmarkets@gmail.com or throwing me a tweet @electionmarkets.

GOP AT NEW HIGH IN HOUSE ESTIMATES; SENATE DEMS IMPROVE PROSPECTS

Expected Republican gains in the House rose on our midpoint estimate to a projected 232 seats.  The low point estimate rose to 231 seats and the upper point estimate rose to 232 seats.  The projections represent highs on all three measures. 

Republicans did not fare so well on the Senate side of the ledger.  Republican seat projections declined on both our macro and micro model estimates to midpoint estimates of 49.0 and 47.9 seats respectively.  Our models still suggest that the final Senate seat count will come down to a limited number of seats.  Our models estimate of probabilities for a Republican win in those seats are displayed in the table below.

Democrats Continue Senate Clawback; GOP Solidifies House Prospects

Democrats continued to claw their way back in the race for control of the U.S. Senate.  While there was essentially no change in the median estimate of 49.5 Senate seats in our macro model, our micro model gave Republicans their lowest estimate (47.9 seats) this month.  The reading of 47.9 represents a decline of 0.6 seats over the past two days.  The macro model's estimate of Senate control also retreated by roughly 5% from yesterday's estimate.  Both models utilize data from InTrade markets in generating their estimates. 

The Republican decline from yesterday centers primarily in a 10% decline in Washington state and an eight percent decline in Colorado in our midpoint estimates.  Republicans are above 50% in 27 of the 37 races which would yield 50 seats if Republicans prevailed in all 27.  But, as the table to the left shows, very little change is needed to tilt the field further back towards the Democrats.

As an illustration of just how volatile these markets can be at this point in time consider the table to the left from the October 18, 2006.  The percentages are the win probability percentages for Democratic candidates on an earlier version of our model.  However, the midpoint price estimate methodology varied only slightly so the comparison is relevant if not exact.

The good news for Republicans is that candidate Jim Webb had roughly the same estimated probability of defeating Senator George Allen (Virginia) as the InTrade market gives Republicans of capturing the Washington seat.  The bad news, Democrats only needed to win two of the three to seize control. They did.  Republicans on the other hand would need to run the table in comparable seats.

On the House side, Republicans continue to solidify their position.  While our midpoint seat estimate remained at 231, the probability of House control soared to 86%.  The big question for House Republicans appears to be whether they will see an increase in their seat estimates.

Daily Update: House Republicans Another Step Forward; Senate Republicans Take a Step Back

Post is a little late:  Have to admit I had to take the time to watch the first miner emerge in Chile.  Certainly an uplifting moment.

Republican prospects in the House of Representatives took a step forward in today's data run as the probability of a House Republican takeover pierced the 80% level.  The midpoint range of our seat estimate is 231 which would represent a 54 seat increase from the 177 Republicans held after the 2008 elections.

On the Senate side of the ledger Democrats produced a small, but significant comeback in today's analysis.  (It is worth noting that the data was extremely volatile today and our data run represents a snapshot of data at about 9 P.M. EDT today.)

The midpoint, or median, estimate of our macro model forecasts 49.5 Republican seats; the micro model forecasts 48.1%.  The macro model currently projects the probability of a Republican takeover of the Senate as 26.8%.

As I mentioned there were several changes in individual races from yesterday.  As the table to the left illustrates, if our lower end estimate in all races proved to be the outcome, Republicans would fall four seats short, while at the upper end estimate they would fall one seat short. 

Remember the markets are extremely volatile at this point in time and it is not unusual to see large one day swings in the data.  The question is whether this is a new trend or simply a one or two day correction.  To know that, we'll have to wait for a few more days worth of data.

Today's data does provide the opportunity to give an illustration of conditional probabilities and how they could be affected as events unfold on election day.

Let's take the hypothetical situation where four races (West Virginia, Washington, Illinois, and Nevada) are the races that will determine Senate control.  From a straight statistical standpoint the probability of Republicans winning all four is 6% -- using our median probability estimate in the races AND assuming they are independent events.

As you watch on election night, as each race is decided this probability will change.  The table at the left shows how the probability of Republican control would change based on results becoming known in the individual races.  As I said, this is hypothetical and is unrealistic from a time perspective, but is intended to provide an example.  If Republicans win West Virginia, then the probability of control would increase to 12%.  If the next state to be decided in the Republicans favor were to be Washington, Republicans chance of control would increase to 26%.  With two races remaining, the news arrives that Republicans have won Illinois and only Nevada is outstanding, then Republicans chances of winning the Senate become 51%. 

We hope to be able to blog updated probabilities throughout the evening on election night as races are decided.  This will vary from the example above not only by order (and perhaps states), but by race probabilities as well.  For instance if Republicans were to score an early and sizeable victory in West Virginia, that would likely alter market participant views of other races.  Conversely, a quick Democratic win, would likely see probability declines for Republicans in other races.

Regardless of how successful we are in live blogging election night, we will have much more to say about scenarios and how to track election night probailities as the returns come in.

If you have any questions or suggestions for topics please email us at electionmarkets@gmail.com.  And I encourage you to let others know about this site.  At the end of this week we will be posting our first online poll -- if the President and Vice President choose to watch a movie on election night instead of the returns, what will it be.  The list is about complete, but email us any suggestions for movies to include.

Republicans Strengthen House Position; Will There be Dino-mite in the Senate?

Top Line Numbers

Republicans scored their highest yet probability of taking the House in today's model run -- 78.1%.  The estimated Republican seat count after the election stands at 230 in both our low-end estimate and the median estimate.   The low-end estimate is also the highest yet for Republicans on this measure.

As for the Senate, our macro model is indicating a Republican seat count of 49.5 and a control probability of 28%.   (Remember -- the InTrade control contract on which our model is based would not count a Murkowski victory toward Republican control; presently our model shows the Alaska race favoring Joe Miller with a 64% probability of winning, Senator Murkowski at 29% chance and Democrat McAdams with a 7% chance).

Our micro level model (still operating on a wide screen for analysis purposes) currently projects a Republican seat count of 48.4 seats. The same caveats apply to our micro model as do our micro model, i.e.; a Murkowski win would not count as a Republican seat.  

The following table shows our current probability range estimates for the most competitive races.

The table lists those Senate races where  the Republican and Democratic candidate both have a victory probability of greater than 20% on our midpoint estimate.  None of the six is a seat currently held by Republicans.  At our low end estimate (for Republicans) the Republican candidate has a greater than 50% chance of winning in four of the six; at the upper end estimate the Republican is above 50% in five of the six.  If Republicans were to win those five they would control the upper chamber.  One of the changes in today's rankings is that our model now gives Sharon Angle a higher probability of victory than Matt Kirk in Illinois, although the margin is very narrow.

Sunday Night Special: Angle Poised to be 50th GOP Senator; Rossi Knocking on Door for 51st

A couple of days ago, the situation was reversed -- and we may see several more reversals in the days to come.  Today's analysis of InTrade contract data shows Sharon Angle pulling slightly ahead of Senate Majority Harry Reid and now stands at a 51% probability of victory.  This is a signficant improvement from the 37% probability of victory she had declined to in mid-September.

Less than three weeks ago our model gave Rossi a 31% chance of victory; today our model assigned a probability of victory of 49.9%.  Rossi did briefly surpass the 50% threshold two days ago.  This race remains incredibly tight not only in the polls, but in our estimates as well.  Our model suggests that Matt Kirk would be the 49th GOP Senator and assigns him a 55% chance of victory -- clearly still in the toss up category.

On the House side, Republicans are in a strong position according to our models.  Based on our analysis of InTrade data Republicans now have a nearly 77% chance of winning control of the House of Representatives.

Today our model estimates that Republicans will hold between 228 and 231 seats in the next Congress.  This would represent a pickup of between 51 and 54 seats from the 177 they held following the 2008 election.

As noted yesterday, our analysis of InTrade data suggests that Republicans have a roughly 1/3 chance of gaining 60 or more seats.

If you want to see some more of today's charts, click read more below.