Announcing the Wigand Electoral Average (WEA): Obama 49.07, McCain 50.93

[phrigndumass]

'Swing State Index' aka The Wigand Electoral Average

.



.



In a new cooperative effort, grantcart and phrigndumass are offering a new indicator that focuses on the critical swing states for the 2008 General Election, the Wigand Electoral Average (WEA).

Based on the same concept of the Dow Jones Industrial Average or S+P 500, we will use probabilities in 12 key swing states to serve as a basis for measuring the likely electoral college success of the Obama campaign in the General Election.

It factors in the 'Wisdom of Crowds' by following the movement of the 12 states based on the daily price of Intrade. We expect that it will be a more accurate indicator of electoral college success in the General Election than other indicators like the daily Gallup or Rasmussen polls. Its methodology follows that of the Dow Jones Industrial Average and will serve the same function -- establishing a bench mark and judging the changes in the "Swing State Electoral College Market." Each state will be weighted according to its Electoral College value in the same way that the Dow Jones weights the capital accumulation in each of the companies in its average.

It should be a clearer indicator of the state of the GE campaign than the Gallup or Rasmussen Daily polls or the RCP Average, which don't take into account either the perception of projected outcome (which Intrade does) or electoral college values.

In order to have an appropriate moniker we have decided to call this new indicator the WIGAND ELECTORAL AVERAGE.

It honors the courage of Jeffrey Wigand who used science to speak to power. His courage enabled the courts and eventually Congress, to advance the will of the people in holding the Tobacco industry liable for its assault on public health. His actions were essential in laying the foundation for the The Tobacco Master Settlement Agreement (MSA). The Bush administration employed key Tobacco lobbyist Charles Black, Kirk Blalock and most famously Karl Rove who worked for Phillip Morris from 1991 to 1996.

http://www.stopcorporateabuse.org/cms/page1418.cfm

The Bush Administration wiped out most of the penalties against Big Tobacco and settled reducing the penalty from $130 billion to $10 billion. So in a real sense the Wigand Electoral Average is an index of America returning to a better time, a time before the Presidency of George W. Bush.

His story became well known in the movie "The Insider" where he was played by Russell Crowe.

Jeffrey Wigand then returned to the classroom:

For four years Dr. Wigand taught Japanese and Science (Biology, Chemistry and Physical Sciences) at duPont Manual High School, a national school of academic excellence, in Louisville, Kentucky. He received national recognition for his teaching skills when he was awarded the Sallie Mae FIRST CLASS TEACHER of the YEAR in 1996. He was one of 51 teachers recognized nationwide.

Naming this new index The Wigand Electoral Average also, by extension, honors all teachers as well.

* * * *

More background on Jeffrey Wigand here:

Jeffrey Wigand in Wikipedia: http://en.wikipedia.org/wiki/Jeffrey_Wigand
Jeffrey Wigand's Home Page: http://www.jeffreywigand.com/index.php

Background on the Wisdom of Crowds prediction theory:
http://en.wikipedia.org/wiki/The_wisdom_of_crowds

Intrade link:
http://www.intrade.com

* * * *

How the Wigand Electoral Average Works

The Wigand Electoral Average uses public polling results to determine base states for both parties and a short list of swing states, and combines the actual electoral votes from base states with a weighted Intrade calculation for electoral votes from swing states to derive an average on a scale of 0 to 100. On this scale, greater than 50.00 is considered a win for Barack Obama. An average of 50.01 would yield 270 electoral votes, and an average of 100.00 would yield 538 electoral votes.

The 12 key swing states used in the Wigand Electoral Average are:

Industrial Midwest/East:
Indiana (11)
Michigan (17)
Pennsylvania (20)
Ohio (20)

Rural Midwest:
Iowa (7)
Nebraska (5)
Wisconsin (10)
Missouri (11)

West:
Colorado (9)
New Mexico (5)

South:
North Carolina (15)
South Carolina (8)

These states are considered Obama base states:
California (55)
Connecticut (7)
Delaware (3)
District of Columbia (3)
Hawaii (4)
Illinois (21)
Maine (4)
Maryland (10)
Massachusetts (12)
Minnesota (10)
New Jersey (15)
New York (31)
Oregon (7)
Rhode Island (4)
Vermont (3)
Washington (11)
Total (200)

These states are considered McCain base states:
Alabama (9)
Alaska (3)
Arizona (10)
Arkansas (6)
Florida (27)
Georgia (15)
Idaho (4)
Kansas (6)
Kentucky (8)
Louisiana (9)
Mississippi (6)
Montana (3)
Nevada (5)
New Hampshire (4)
North Dakota (3)
Oklahoma (7)
South Dakota (3)
Tennessee (11)
Texas (34)
Utah (5)
Virginia (13)
West Virginia (5)
Wyoming (3)
Total (199)

[funflower]

1. Barack has barely BEGUN to fight the general election battle.

A lot of people haven't taken a serious look at him yet. When they do, the same thing will happen as has happened on the dem side.

[phrigndumass]

4. It will be fun to watch this index grow :)

He's on his way up!

[bhikkhu]

5. His sparring with McCain so far is inspiring

The more McCain speaks, the more incompetent he sounds. The more Obama speaks, the more presidential..

Perhaps I am biased, but this is one race I am looking forward to!

[phrigndumass]

8. McCain is Dubya on Geritol

:)

[funflower]

20. BWAHWHAHAHAHA!!! I think he needs more of that Geritol!

(Guess you have to be a certain age to even know what Geritol is!)

[Art_from_Ark]

36. The more bu$h spoke, the more incompetent he sounded

The more Kerry spoke...

[bhikkhu]

2. 1st rec - you guys are the best, and a metric for the GE is welcome!

Keep up the good work...

[MadBadger]

3. How did you determine what are the base states and swing states

Cuz I dont agree with a few of them.

[phrigndumass]

6. We found 12 states (6 R and 6 D) that are currently polling in the middle ...

... and have a chance of going either way. It took us some time to agree on this list, based on intrade and current polls. We had to narrow it down to 12 states to find equal base states.

Tell us which ones you think shouldn't be on the list, and which ones you think should! That's part of the fun of this.

:D

[MadBadger]

7. Nevada, South Carolina, and Virginia were the questionable ones

I think though Rasmussen has it as a five point lead for McCain a while back, in my mind its a swing state that has a better chance of going D this year. Secondly Virginia. Both VA and NC have Obama down 3 according to rasmussen. I havent checked intrade, but it seems like if NC is considered a swing state, so would VA. Personally, I think Obama has a better chance of winning VA than NC, but thats just me.

I would switch Virginia and South Carolina because no matter what some poll months ago says, I think SC will definitely go to the Repukes.

[phrigndumass]

10. The good thing about the WEA index is that those two states

... will be calculated into it when they poll as blue. If Nevada and Virginia are polling as Obama winning, that increases the index.

Neat, huh?

[MadBadger]

12. If you missed my edit, I think by the time a new poll comes out, SC will switch with VA

VA has a much better chance of going blue than SC does.

[grantcart]

29. we don't disagree but having all heavily favored Obama states would reduce the liquidity of

movement in the index for Obama. Kind of like buying low and selling high. The movement of any state from one base to another base would also move the index. In the case of Virginia it would move the index 2.42.

The point of the states was not to pick the 12 most likely to move but the 12 that would be representative of movement, give good geographical distribution and still leave the two bases equal on the first day.

If tomorrow a Virginian is picked for VP and Virginia moved from red to blue it would still move the index.

[grantcart]

13. It was a very involved issue but I can tell you beer was involved

by complete accident the numbers also provided perfect electoral division of the other base states for McCain and Obama.

This was not intentional but seem to be a good omen from the math gods.

[CitizenLeft]

15. hey, you beat me to it...

...about SC & VA. I thought the same thing.

[phrigndumass]

22. I thought so as well at first, but both states are polling McCain +3

and have negative Intrade values for Obama, even though Virginia is higher for Obama. In the end, we arrived at even bases and two of the southeast states represented.

We're expecting Virginia to go blue more than we are expecting South Carolina to go blue, and if (when) that happens, it will increase the index with 100% of Virginia's 13 delegates, as opposed to increasing it with 100% of South Carolina's 8 delegates.

:D

[woolldog]

17. I agree Obama has a better chance of winning VA:

more educated white voters; more translplants from more liberal areas who work in DC and commute etc It appears like there's less racial polarization there also.

Plus Obama actually has polled ahead of McCain in VA. He's never polled ahead of McCain in NC or SC.

[grantcart]

25. because SC is trading lower than Virginia 10 versus 43

it actually favors Obama to have SC listed in the 12 because SC is more likely to move more points than Virginia. We had so many states that Obama was already favored to win big by Intrade that we needed more that were rated low to be able to give Obama the chance to move up. For example CO Iowa NM MI PA WI OH are already substantially in Obama's tent in Intrade at 62 70 62 79 67 70 61 they have little room to move up so unlikely to help Obama but SC at 10 will bring the index up if Obama's chances there improve.

Moreover the index has a fail safe in that if a state moves from a base state of one candidate to another - for example Virginia - then it would also drive the index up.

[hokies4ever]

9. Could you list the base states in order

of most to least likely to be base states? For example, Louisiana is a very secure base state for McCain, whereas Virginia is not as secure of a base state. Also, how is North Carolina considered a swing state but not Virginia?

[phrigndumass]

18. I'll be happy to do that :)

We're tracking Intrade for North Carolina, but if Virginia is polling as Obama winning, that also increases the index.

We found 12 states (we had to narrow it down, it was hard!), six of them considered red and six considered blue, that are currently polling in the middle and could go either way. This is the list we finally agreed upon.

We also didn't include Florida and Nevada, and they will be treated like Virginia and could increase the index when polling on Obama's side.

:)

[hokies4ever]

24. Thanks! Also:

Here's an interesting website someone else pointed out to me on these boards. Just thought that I would share. Keep up the great work. :patriot:

http://www.fivethirtyeight.com/2008/05/oregon-projection-obama-by-13.html

[phrigndumass]

27. Love that website!

I blogrolled it in my journal to read everyday.

Thanks!

:D

[phrigndumass]

38. Base states in order of most to least likely to be base states ...

Per your request:

These states are considered Obama base states, in order of most likely:
District of Columbia, Obama +81
Vermont, Obama +34
Hawaii, Obama +30
Illinois, Obama +29
New Jersey, Obama +24
Connecticut, Obama +17
New York, Obama +17
Rhode Island, Obama +15
Maryland, Obama +14
Oregon, Obama +14
Maine, Obama +13
Minnesota, Obama +13
Massachusetts, Obama +12
Washington, Obama +12
Delaware, Obama +9
California, Obama +7

These states are considered McCain base states, in order of most likely:
Arkansas, McCain +29
Tennessee, McCain +27
Kentucky, McCain +25
Oklahoma, McCain +23
Kansas, McCain +21
Alabama, McCain +19
Wyoming, McCain +19
West Virginia, McCain +18
South Dakota, McCain +17
Louisiana, McCain +16
Georgia, McCain +14
Idaho, McCain +13
Mississippi, McCain +13
Texas, McCain +13
Utah, McCain +11
Alaska, McCain +9
Arizona, McCain +9
New Hampshire, McCain +8
North Dakota, McCain +6
Montana, McCain +5
Nevada, McCain +5
Virginia, McCain +3
Florida, McCain +1

[RichardRay]

40. Since you've got the data already

How about that same list (Most to Least Likely) and add the EV's from each state? I'd like to see how many EV's are how close or far away from moving in the Base Support lists. I have a feeling (and it's nothing more than that) that some of the historical assumptions about 'base states' might get changed this cycle.

[phrigndumass]

45. Updated for you :)

Per your request:

These states are considered Obama base states, in order of most likely:
District of Columbia (3), Obama +81
Vermont (3), Obama +34
Hawaii (4), Obama +30
Illinois (21), Obama +29
New Jersey (15), Obama +24
Connecticut (7), Obama +17
New York (31), Obama +17
Rhode Island (4), Obama +15
Maryland (10), Obama +14
Oregon (7), Obama +14
Maine (4), Obama +13
Minnesota (10), Obama +13
Massachusetts (12), Obama +12
Washington (11), Obama +12
Delaware (3), Obama +9
California (55), Obama +7

These states are considered McCain base states, in order of most likely:
Arkansas (6), McCain +29
Tennessee (11), McCain +27
Kentucky (8), McCain +25
Oklahoma (7), McCain +23
Kansas (6), McCain +21
Alabama (9), McCain +19
Wyoming (3), McCain +19
West Virginia (5), McCain +18
South Dakota (3), McCain +17
Louisiana (9), McCain +16
Georgia (15), McCain +14
Idaho (4), McCain +13
Mississippi (6), McCain +13
Texas (34), McCain +13
Utah (5), McCain +11
Alaska (3), McCain +9
Arizona (10), McCain +9
New Hampshire (4), McCain +8
North Dakota (3), McCain +6
Montana (3), McCain +5
Nevada (5), McCain +5
Virginia (13), McCain +3
Florida (27), McCain +1

[CitizenLeft]

11. just my personal opinion, having been brainwashed by pundits...

I would swap Virginia for South Carolina - move SC to McCain's base, and make VA a swing state instead.

Great work, you two! :applause:

BTW, the Insider is one of my favorite movies - "that's the 10th floor, that's where they fuck with my life!" - very cool, honoring Jeffrey Wigand! :thumbsup:

[MadBadger]

14. That's my opinion too.

[CitizenLeft]

16. yep, just saw you up there!

:hi:

[phrigndumass]

19. Good opinion ... it took us some time to decide between VA and SC

... we ultimately went with SC because using VA would have tipped the base states more in Obama's favor and could possibly skew the average.

Adding VA and removing SC would result in:
Obama - 205 EVs from base states
McCain - 194 EVs from base states

... instead of 200-199.

We want the index to be fair and as accurate as possible, in honor of Wigand.

:D

[funflower]

21. I think Virginia will go blue this year.

It's been trending that way for a while, and Barack will energize democrats and reach out to independents there, while McCain will continue to alienate the Virginia Beach fundy crowd.

[phrigndumass]

23. * * * BASE STATES CAN SWITCH OVER TIME * * *

:D

That's the beauty of this index. If a base state begins polling better for the other side, it will automatically increase the average for the other side.

For example, Virginia could begin polling blue and would increase Obama's index. Since it's not considered one of our twelve swing states, 100% of Virginia's electoral value would be given to Obama, instead of a weighted intrade calculation such as would be done for South Carolina.

It benefits both sides to put Virginia in a base instead of in the middle.

:D

[featherman]

44. Jeepers, I hope so. Having NV, NH, FL, and MT as McCain base states is

very pessimistic. These are all Tossups in my model and I'm adding ND for 2008.

PA, WI and MI are DEM base states. No question in my model. It's been quite a few Presidential cycles since these states voted for a REP and I can see no reason why they would trend red under the current negative GOP political climate and with Democratic Governors and state party control.

My "lean Dem" tossups: OH, CO, NM, NH, NV, MT, FL and IA - all have DEM state governments, dispirited GOP, and voted DEM at least once in the last four cycles

My "lean GOP" tossups: IN, MO, NE (split), ND, VA, NC

SC is GOP base state

[phrigndumass]

46. I think our terminology is confusing

When we use the term "base" states, we really mean "basis" states, as in a basis to begin our model. If NV, NH, FL and MT begin polling blue, it will increase Obama's index. Those states are polling red at the moment.

We needed a 50/50 original split of electoral votes for our basis states, and a 50/50 split of "swing" states (6-R and 6-D) for a fair sample. Our original basis state EVs give Obama 200 EVs and McCain 199 EVs.

"Swing" is confusing as well. Maybe "indicator" states would be a better term. For these 12 states only, we don't use public polling at all. We use Intrade and rely on the wisdom of crowds as referenced in the OP.

:D

[featherman]

49. Thanks for the response. I know it's early and with the high feelings still

going in a hotly contested DEM primary, I have a hard time giving much credence to current head-to-head polling which would tend to over-favor McCain at this point in the cycle.

Instead of taking polls at face value (ie. a McCain "lead" in some state that is within the MOE), I tend to favor past cycle voting patterns and historical trending state-by-state.

The head-to-head polls will all shake out more clearly as the Dems coalesce around their own "winner".

I expect Obama to bounce out of the convention with solid polling leads in an unusually broad base of states and it will be McCain playing catch up allied with the MSM who will do whatever they can to tighten the contest.

[phrigndumass]

51. Your expectations are very realistic, imho

I believe you're right about Obama's convention bounce and the MSM's coming attempt to rein it in to increase their profit margin. It'll be interesting!

:toast:

[Capn Sunshine]

26. Pretty cool, but one thing

The concept of "Swing State" is a corralary of the idea that there is a tiny slice of voters (aka "Swing voters")that decides the election for you , where you win 51%, the opponent gets 49%

That's the Karl Rove- Cintonian era model. Worked when there were close lections, in the 90s and extended into 2004.

2006 BROKE that model. Take a look. It is this NEW model , where all former "red " states ARE IN PLAY that our campaign is focused on. We're planning on a huge margin of victory, at leasy 6% and these voters just don't show up on polling methods currently in use.

Take a look at the model that asks voters to self identify R or D.
D's are up by double digits.

Take a look at the new registrations we have brought in; if we get just a 20% pull through, it will be a landslide.

[grantcart]

30. Well we agree 100% and that really is our motivation in having a neutral index to

document that movement. In any case it is in our interests to be as objective as possible in having a simple indicator that tracks how we are doing in states that are in play. As explained in the thread all states are 'in play' because if one of the base states moves from red to blue it would also move the index.

[phrigndumass]

32. Very true! Thanks for pointing that out

We need to focus on all 50 states!

Focusing in every state will hopefully affect both the polls and Intrade, and will increase the index. If a state that has been considered red begins polling blue, the index will jump. This index covers all 50 states, but gives more weight to the "wisdom of crowds" (Intrade) for the swing states we are using.

Hoping for a landslide! I believe this index will show it coming.

:D

[Greyskye]

28. I wasn't familiar with Intrade

So I checked the concept out.

Here's the basic info I went in wanting to find:

Are Intrade Predictions better than Polls?


Whether prediction markets like Intrade can predict more accurately than opinion polls has been studied. Based on past performance prediction markets seem to be more accurate. Does that mean prediction markets like Intrade will be in every instance? No. But we believe that the market mechanism is superior to a polling mechanism. Also, the Intrade prediction market will aggregate information from the polls (and perhaps poll participants will also be influenced by the predictions from the markets).

How can Intrade's Predictions be So Wrong?


The predictions or prices on intrade are predictions, they are not guarantees or certainties. A market that is trading at 90, means that there is a 90% chance of the event happening. 90% means the event it is highly likely to occur, but not guaranteed. Intrade's predictions have been wrong in the past, and will be again in the future. However, the Intrade market mechanism of aggregating opinions to give the an estimate of likelihood of an uncertain future event occurring is the best one we know.

For what it's worth, here is what Intrade currently shows for the 2008 race:

Barack Obama to be Democratic Presidential Nominee in 2008
Last Price: 92.5 0.5


Hillary Clinton to be the Democratic Presidential Nominee in 2008
Last Price: 6.1 0.4


John McCain to win 2008 US Presidential Election
Last Price: 39.0 0.6

Pretty fascinating, I'm looking forward to seeing how this plays out. Thanks for your efforts on this!

:toast:

[grantcart]

31. Also Intrade is trading Obama and McCain almost the same when it

comes to who will win the nomination of their party 93 versus 95


All of the Clinton people who believe that she is actually going to win should get over their and buy now.

[phrigndumass]

35. Fascinating info, Greyskye!

Thanks for posting that in this thread. I'll probably make reference to it a lot, now that I know where to find it easily (in my journal threads, thanks to you!)

Both grantcart and I have tested this index to death, and I think it's a solid one.

For our swing states, we just let the wisdom of the crowds decide it and let it go. "By letting it go it all gets done. The world is won by those who let it go. But when you try and try. The world is beyond the winning." (or so grantcart tells me, and I tend to believe it)

:toast:

[Honeycombe8]

33. Oh, hogwash. These guys aren't qualified to do a prediction any more than I am.

It's amazing all the loonies who are coming out of the woodwork to make a buck predicting winners and losers.

I took some math classes and read a statistics book once. I'll come up with some predictions and post them next week (I'll even enter something about my wonderful background in Wikepedia).

[grantcart]

34. We will look forward to that - be sure to make sure we see it

also interested in your wikipedia article

on edit

you are aware in fact that we are not making any predictions?

[phrigndumass]

37. If you have a crystal ball, it's a prediction ...

If you have a post-graduate degree in this field, it's qualified.

:hi:

(btw, nice profile)

[Egnever]

39. Sweet cant wait

meanwhile cool stuff you two cant wait to watch it pan out.

[grantcart]

41. for the morning crowd

[Egnever]

43. people read this stuff in the morning?

[Upton]

42. I look forward to a steady Obama climb

certainly states like Wi, Iowa, and Pa. should be put in his column in the future.

[olkaz]

47. George Soro's newest book debunks "The Wisdom of Crowds"

It's very, very flawed.

But I have loved your work over the primary season and PLEASE keep it up.

[phrigndumass]

48. Interesting ... I'll check it out!

I would agree with him that the wisdom of crowds is flawed, particularly in the realm of economics, where he specializes. But when we are considering elections, the wisdom of crowds (or the lack thereof, lol) is all we have.

Our index takes the 12 indicator states out of the hands of the polls and allows the crowds (Intrade) to decide. The remaining 38 states are still tracked by the polls within our index.

Thanks for your nice words, and for giving me a heads up on George Soros' new book!

:hi:

[RichardRay]

52. Do you have a title for the new Soros book in which he

addresses the Wisdom of Crowds? I've seen his critiques and I've not understood them to be questioning the underlying concepts so much as the way people have tried to use it.

[olkaz]

53. He was on NPR on my way to the gym a couple weeks ago

I don't remember the title. I'll take a look.

But anyway, in the interview, he gave a pretty impressive critique of the theory. I can only imagine how deep it gets in the book.

[olkaz]

54. Found it! The New Paradigm for Financial Markets: The Credit Crisis of 2008 and What It Means

http://www.georgesoros.com/creditcrisis08

Enjoy!

[phrigndumass]

55. Thanks for linking!

Hope I have room on my credit card to buy it (lol) ...

:7

[olkaz]

57. Here's also a link to the NPR story w/audio

http://www.npr.org/templates/story/story.php?storyId=90328243

It's worth listening to.

[grantcart]

59. I think phrngdumass has found a very clever formula that combines

polls for 36 states and intrade for the 12 in our swing parameter so that there will be a balance.

Also in this case it is not simply a 'wisdom of crowds' but people actually investing money in their opinion

[Scurrilous]

50. K & R

:thumbsup:

[Basement Beat]

56. Sweet!

This will be a fun poll to watch. :)

[grantcart]

58. Wednesday kick