Electability of 2016 Presidential Candidates as Implied by Betting Markets

http://toddwschneider.com/posts/electability-of-2016-presidential-candidates-as-implied-by-betting-markets/

by lil_tee • 11 years ago

54 61 11 years ago

toddwschneider.com

vixen99• 11 years ago

How depressing that a country of 300+ million individuals has candidates from the same families again and again.

jahnu• 11 years ago

Well, with democracy at least it doesn't take a war to change them....

... um....

uptown3• 11 years ago

Quora: "Are all the U.S. Presidents related to each other?"

https://www.quora.com/Are-all-the-U-S-Presidents-related-to-...

jerf• 11 years ago

Followed link, turns out all but one President can be traced to... King John, the one in the Robin Hood story, born December 24, 1166.

Which in practical terms translates to "no", they are not related.

matwood3• 11 years ago

Actually the rise of Trump, Carson, and to a lesser extend Fiorina shows that people are tired of the professional politician. Even if Bush or Clinton were the best candidates (I don't think they are), they look to be facing a huge headwind of people tired of the Bushs and Clintons.

ck24• 11 years ago

The rise of Trump and the others simply means people want entertainment, that's all, the usa is turning into idiocracy.

happyscrappy• 11 years ago

I hope it is Bernie Sanders vs. Donald Trump. Not because I like Trump but because it would make Sanders more likely to win.

JustSomeNobody• 11 years ago

Ah, but the pool isn't really 300+ million. It takes a person with the qualities of a lawyer, used car salesperson, con artist and book maker to even consider running.

I guess this is why many (even intelligent) people think that their favorite candidate is honest.

ck24• 11 years ago

Right because two sons of a president being a president is exactly the same as a non-blood relation of a president becoming president.

The entire Bush family (and it's rather big) is in politics, except the women.

How many Clintons are in politics?

Lawtonfogle• 11 years ago

Pretty close when you compare the odds and consider the total pool of who could be president. As in 'so close without a microscope I would call them identical'.

zxcvcxz• 11 years ago

That's why I think we should elect officials on a local level, then send the winner to state to compete against the other local winners in the state, then on to the national level. After the local level everything is funded by taxes and highly regulated to attempt to give everyone a fair chance.

I think it would highly diversify the elections and give the people more of a sense of power. A local election is a lot easier to run than a national one.

protomyth2• 11 years ago

Its not exactly a new thing. The 2nd and 6th US Presidents were a father-son combo. The 23rd President was the grandson of the 9th. Multiple Presidents were cousins of other Presidents. But for assassination, a tragic accident, and the moral judgement of a hyena[1], there might have been three President Kennedy brothers and one son.

1) leaving a girl to suffocate in a car underwater didn't end his political career but probably cost him the Presidency - that is what's sad in American politics

kittenfluff• 11 years ago

Interestingly, with Joe Biden in the range (10%, 13.5%) to win the nomination, and (7.7%, 8.3%) to win the Presidency, the lowest his electability could be is

    7.7% / 13.5% = 57%

and the highest it could be is

    8.3% / 10.0% = 83%

so the market seems to be pricing a probability in the range of (57%, 83%) for Biden to be elected president if he won the nomination - compared to Hillary Clinton's range of (56.9%, 57.7%)

I can think of a few explanations -

1. Biden really is a lot more electable than Hillary Clinton

2. Both candidates have electability at the low end of their range (around 57%).

3. The market is wrong, i.e. they are systematically underrating Biden's chance of winning the nomination (and overrating Clinton's) or overrating Biden's chance of winning the election (and underrating Clinton's) or both.

There is no arbitrage, but if you believe 3, there might be a good profit to be made in expectation by backing Biden to win the election, but Clinton to win the presidency (you wouldn't hold it through to 2016, but take off the bet as soon as the odds come back to something that looks more plausible).

I haven't done the analysis to see if it's still worth it after trading costs, but maybe someone else wants to.

matwood3• 11 years ago

This makes a lot of sense. I see Biden as much more electable than Clinton mainly because Clinton will draw out the 'hate Clinton' crowd to vote against her. I think the DNC is also realizing this, hence all the talk of Biden jumping in to begin with.

matthewbauer• 11 years ago

A 4th option could be:

If Biden defeats the established nominee, he must have an exceptionally well run campaign making him more electable. That is by winning the nomination Clinton has just met expectations but Biden exceeds most expectations.

ovet• 11 years ago

I think it could also be that Vice Presidents have generally done well in presidential elections when running after their VP term.

http://www.huffingtonpost.com/jonathan-hobratsch/when-vice-p...

saturdaysaint5• 11 years ago

I don't know if this has any special significance, but I think the number of FB/Twitter followers for the current frontrunners are interesting. I'll use Facebook numbers.

Hillary Clinton - 1.3 million

Jeb Bush - .26 million

Donald Trump - 3.7 million

Joe Biden - .85 million

Bernie Sanders - 1.4 million

Marco Rubio - .99 million

John Kasich - .14 million

Scott Walker - .36 million

Ben Carson - 2.88 million

No, I don't think that Donald Trump is the runaway favorite for president. On the other hand, a look at these would make Walker's collapse and Bush's struggles (despite a goliath $100 million warchest) less of a surprise than the political press has portrayed. Likewise, Sanders' challenge to Clinton would be less surprising.

Bush is the real surprise to me. Endorsements and dollars show that he's the establishment's pick, but people just aren't interested in the guy. It would be kind of amazing if someone went from having a D-list celebrity following to sitting in the oval office in a little over a year.

uptown3• 11 years ago

If that's your metric, it's probably worth running them through a few filters to detect authenticity of their followers:

https://www.twitteraudit.com/

vonmoltke• 11 years ago

That site thinks I'm a fake, so I question the accuracy of its metrics.

saturdaysaint5• 11 years ago

(OP) Good idea - FWIW the numbers look proportionately the same if filtered through that.

Interestingly, Hillary actually towers over Sanders in terms of authentic Twitter follows.

arca_vorago• 11 years ago

Be careful with using social media followers/engagement as a metric, because it's an easily manipulated stat. According to some recent stories, a million of H Clinton's followers are fake or fakeish users, actually putting her closer to Jeb's actual number.

Also, with the increasing rise of sock puppetry, I would not be surprised to see these sort of artificially inflated numbers are not just organic, but a planned/hired out part of a public relations campaign.

rattray• 11 years ago

This is pretty interesting. Consider turning into a short blog post?

ck24• 11 years ago

This is like comparing halloween mask sales volumes to who will be elected.

How many Trumps with that hair are going to be running around at the end of October?

saturdaysaint5• 11 years ago

I acknowledge that this could be completely invalid, but a Halloween mask sits in a closet, while a Facebook/Twitter follow can be (assuming the user or the algorithm don't bury it) more powerful than any campaign link to voters, well, ever.

They're already breaking rules of past elections where a "gaffe" would be endlessly regurgitated by news anchors until the candidate agreed to a heartfelt exclusive sit-down apology with some network. Trump (perhaps unwittingly) has figured out that the direct relationship of social media makes that irrelevant, and is taking full advantage by belittling his social media inferiors.

The candidates with good followings are basically getting free focus group testing - I wouldn't be surprised if Trump ran with immigration after floating dozens of wild causes and noticing that immigration had a particular resonance.

djscram• 11 years ago

Pretty sure that a lot of people who hate Donald Trump are following him on social media because he's the most likely person to say something outrageous.

Chathamization• 11 years ago

Worth pointing out that Rand Paul has 2.06 million likes and he's polling at 2.7%.

saturdaysaint5• 11 years ago

Yeah, I noticed that after posting (he wasn't on the original list). My guess is that he's attracted a following of libertarian college students who pollsters probably (rightly) don't give much weight. Maybe he's an example of how flawed the metric is, but maybe he'll be surprisingly resilient as the field winnows.

Killobyte• 11 years ago

"people just aren't interested in the guy" isn't accurate - "people who use social media aren't interested in the guy" is. The candidates with low numbers on social media seem to also be the candidates whose supporters are less likely to be on social media.

saturdaysaint5• 11 years ago

I acknowledge that this is an imperfect, perhaps fatally flawed, metric - I'm sure it skews young, for example. Something I noticed after posting is that Rand Paul isn't on the linked list (and is barely holding on in polls), but has 2 million FB follows. But who knows, I could see him picking up as the fields starts winnowing.

But I think it's fair to characterize Jeb's numbers as indicative of low interest. There are a ton of people on Facebook in demographics that Jeb needs to do well in to win. To stereotype (using known statistics), white males in their 30's tend to vote Republican - there are plenty of those on Facebook and I see them "sharing" conservative links every day (and I can see that many have followed guys like Carson and Paul). In our increasingly "filter bubbled" world, I think we're seeing people aggressively tune out milquetoast establishment types (on the right at least).

wmil• 11 years ago

Jeb is actually really unpopular with a lot of the Republican base. He gets booed when he's at conventions.

His strategy was to use his Wall St connections (he was working for Lehman Brothers when it went bankrupt) to build a war chest capable of permanently damaging any other candidate in the race with negative ads. Essentially forcing Republican voters to make an early choice between accepting candidate Jeb Bush and getting president Hillary Clinton.

That drove most other candidates from the race, but Trump's entry has completely destroyed that strategy.

dragonwriter2• 11 years ago

That drove most other candidates from the race

Given the unusually crowded Republican field, that's hard to defend as consistent with reality. "Resulted in most of the other candidates that were seen as likely and viable by the mainstream establishment staying out of the race" is, perhaps, not as obviously contrary to reality.

PaulHoule• 11 years ago

Prediction markets often break Republican because there are a lot of rich idiots who want to see Republicans win.

(Maybe that is why there are so many Republican candidates. If you can pass all the litmus tests, you can get millions of dollars to pay your friends to do things for you -- between the Koch brothers and the rest of them you can get a lot of free money if you are a Republican. I think Trump thinks he can make money off the publicity, which is a different business model.)

CyberDildonics2• 11 years ago

Not just paying friends, but things like selling books etc.

If you campaign with free money you are basically marketing your book with free money.

impostervt• 11 years ago

I'd be curious to see Electability numbers given scenarious of one candidate v another. For instance, 'Clinton v Bush', 'Sanders v Trump'.

clarkmoody• 11 years ago

A color-coded matrix of candidate pairs would be a nice addition to this discussion.

ck24• 11 years ago

I don't think people understand how close presidential elections are in the US (even after the nightmare in Florida with the supreme court picking Bush as the winner).

The idea of Trump picking the next few supreme court judges is nauseating.

malloryerik• 11 years ago

I'd like to see a history of betting markets in elections. How accurate have they been in the past?

dragonwriter2• 11 years ago

I'd like to see a history of betting markets in elections. How accurate have they been in the past?

IIRC, like polls, they are fairly accurate -- if you look at the state of the market at the last point before the election.

Earlier than that, not so much.

mabbo• 11 years ago

This same math can be applied to the MLB World Series. (I'm in Toronto, which means for the first time since the early 90s, baseball is interesting!)

Instead of primaries, there's Pennants. Instead of general elections, there's the World Series.

http://www.vegasinsider.com/mlb/odds/futures/

Right now, the Jays are the second most likely team to win their Pennant (2/1 vs 9/5 for the Royals). But at the same time, they are the most likely to win the World Series (17/4 vs 9/2 for the Royals). In terms of the original article, the Jays are more 'electable' than the Royals.

I have no idea how this should affect betting strategies, but it sure is interesting from a statistics point of view.

secfirstmd2• 11 years ago

I worked for a summer at Intrade.com during 2008.

It was a very flawed company (and CEO) but it's a pity it's not around during this election. US Presidential elections were one of the few things were it had enough data and people putting money on it's markets that it's predictions were excellent. (Though I think in hindsight I got the job by telling John Delaney the immediate market highs for the Republican's post-Palin entry was a temporary bounce and Obama would win because the economy was the main issue, the Democrats were polling far far higher on it - guess a political degree can sometimes trump market sentiment :)

In fact, I think it's a pity (with the exception of how badly the company was run) in general that something like Intrade is not around. If I am legally allowed to speculate/gamble on oil futures, I really don't understand why I can't be allowed to speculate/gamble/hedge on geopolitical events.

endianswap• 11 years ago

Isn't predictit.com sufficient for what you want?

secfirstmd2• 11 years ago

Hmm, hadn't come across it before.

They seem to be offering to customers in the USA which is interesting. It was the regulatory issues in the US which was one of the biggest problems with Intrade's business.

throwaway1967• 11 years ago

I really don't understand why I can't be allowed to speculate/gamble on geopolitical events.

That would get people interested in geopolitical events. At least, more interested than the establishment would like them to be.

There'd be books on geopolitical spread betting, how to become informed, how reading unpopular economics blogs can actually help your understanding, and so on.

Lots of truth would come out if this were allowed. I'm just hypothesizing, feel free to disagree!

GregBuchholz• 11 years ago

What do you think of the Iowa Electronic market:

http://tippie.uiowa.edu/iem/markets/pres16.html

...and their "U.S. Presidential Election Market"?

csarsam• 11 years ago

I really don't understand why I can't be allowed to speculate/gamble/hedge on geopolitical events.

There was an interesting Planet Money episode lately about the history of betting on US presidential elections - http://www.npr.org/sections/money/2015/08/21/433557935/episo....

It hadn't occurred to me before how much participation in the public sphere gambling could (and previously did) generate.

imaginenore• 11 years ago

Why in the world would they bet on Jeb Bush over Trump or Carson? Look at the actual polls, Bush is doing worse and worse, while Trump and Carson are rising:

http://i.imgur.com/naxgnhi.png

http://apps.washingtonpost.com/g/page/national/rise-of-the-a...

(yesterday's numbers)

WildUtah• 11 years ago

Republican Party insiders like organizers, lobbyists, congressmen, senators, governors, media personalities, and fund raisers despise Trump and adore Bush. Republican insiders make bank working for defense contractors and Trump was against their Iraq profit center. The donor class is focused like a laser on driving down wages with unauthorized immigration and H1-B that Trump wants to limit. The Republican leadership sees Latin Americans as a voting demographic they can capture 40% of but only by opening the border to unlimited immigration to prove they really love Latinos and Trump threatens both of those hopes. (How Republicans will benefit by increasing the number of people who will vote 60% against them is a mystery to me, but I'm not a Republican.)

As the election approaches, the influence of insiders rises and Republicans especially are very obedient to authority figures. Historically, the candidate next in line according to Republican insiders always and without exception wins the nomination, even when he's very unpopular with the base. Therefore, analysts find Trump a very long shot.

Grue3• 11 years ago

Does nobody remember Newt Gingrich anymore? Or Herman Cain? Both of those were leading in primary polls at some point and ended up nowhere.

VT_Drew• 11 years ago

You put in Ben Carson but not Rand Paul?

protomyth2• 11 years ago

Missed Cruz also who has a pretty big warchest at this point.

njharman• 11 years ago

I wonder when if ever it will be cost effective for a candidate to put many bets in order to manipulate the odds. Idea being when you have reported 75% winning lots of people who would vote against you stay home "cause it's pointless".

Electability of 2016 Presidential Candidates as Implied by Betting Markets

Related Stories