latest Betfair odds

[Mr. Morden]

Betfair’s % probability of each of the following being elected president in 2020:

http://sports.betfair.com/?mi=120637747&ex=1

Trump 27.0
Warren 11.4
Pence 10.6
M. Obama 5.7
Julian Castro 4.3
Rubio 4.0
Sanders 4.0
Booker 3.6
Kaine 2.5
Ryan 2.4
H. Clinton 2.3
Cuomo 1.8
Haley 1.8
Cruz 1.5
Klobuchar 1.5

[7,052,770]

I don't think Hillary's very likely, but she's a good value at 2.3%

[Mr. Morden]

Trump 28.2
Warren 11.1
Pence 10.0
M. Obama 5.7
Sanders 4.2
Julian Castro 3.8
Bloomberg 3.7
Booker 3.1
H. Clinton 2.6
Ryan 2.3
Cuomo 2.2
Rubio 2.0
Haley 1.8
Biden 1.7
Harris 1.7
Kaine 1.7
Klobuchar 1.5

[Crumpets]

Wow. Trump at just 28? That's probably where he was... right before election night. Great.

[Phony Moderate]

Wow. Trump at just 28? That's probably where he was... right before election night. Great.

Indeed. The chances of him defeating President Hillary Clinton in 2020 are pretty low.

[Beet]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

[Mr. Morden]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

If Betfair had existed Since WW2, then it (or any other betting markets out there) would have had the guy who ended up winning the presidency favored in their market on election eve in every presidential election from 1952 to 2012.  Now, because the underdog wins one time, they're "discredited"?

[Phony Moderate]

Remember that their purpose isn't to make predictions anyway.

[Beet]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

If Betfair had existed Since WW2, then it (or any other betting markets out there) would have had the guy who ended up winning the presidency favored in their market on election eve in every presidential election from 1952 to 2012.  Now, because the underdog wins one time, they're "discredited"?

Then all they do is reflect conventional wisdom, and magnify it. There's no evidence they add anything of value.

[Mr. Morden]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

If Betfair had existed Since WW2, then it (or any other betting markets out there) would have had the guy who ended up winning the presidency favored in their market on election eve in every presidential election from 1952 to 2012.  Now, because the underdog wins one time, they're "discredited"?

Then all they do is reflect conventional wisdom, and magnify it. There's no evidence they add anything of value.

Yes, they are conventional wisdom aggregators.  I thought that was obvious.

The point is that it's not always obvious to everyone reading about politics how to synthesize the conventional wisdom into a number.  X is the frontrunner to win the primary or the general election or whatever, but how much of a frontrunner?  How much of a chance does the 2nd or 3rd place guy have?  Sure, any of us who follows this stuff can come up with our own numbers, and if we track the races closely enough, we can potentially even beat the market.  But seeing what the collective CW of the market is in numerical form is still something that's of interest to some of us.

[Beet]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

If Betfair had existed Since WW2, then it (or any other betting markets out there) would have had the guy who ended up winning the presidency favored in their market on election eve in every presidential election from 1952 to 2012.  Now, because the underdog wins one time, they're "discredited"?

Then all they do is reflect conventional wisdom, and magnify it. There's no evidence they add anything of value.

Yes, they are conventional wisdom aggregators.  I thought that was obvious.

The point is that it's not always obvious to everyone reading about politics how to synthesize the conventional wisdom into a number.  X is the frontrunner to win the primary or the general election or whatever, but how much of a frontrunner?  How much of a chance does the 2nd or 3rd place guy have?  Sure, any of us who follows this stuff can come up with our own numbers, and if we track the races closely enough, we can potentially even beat the market.  But seeing what the collective CW of the market is in numerical form is still something that's of interest to some of us.

This is philosophical Mumbo jumbo. The point is, if a 90% chance of a result 30 minutes before poll closing on election night can be wrong, what is the hope of any number being predictive this far out? These markets claim to predict events, not capture conventional wisdom. When the average person looks at these numbers, they are thinking about the likelihood of the event happening. No one wakes up from bed in the morning and thinks, "I wonder what the conventional wisdom about the 2020 election is today? I better check Betfair!"

[Mr. Morden]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

If Betfair had existed Since WW2, then it (or any other betting markets out there) would have had the guy who ended up winning the presidency favored in their market on election eve in every presidential election from 1952 to 2012.  Now, because the underdog wins one time, they're "discredited"?

Then all they do is reflect conventional wisdom, and magnify it. There's no evidence they add anything of value.

Yes, they are conventional wisdom aggregators.  I thought that was obvious.

The point is that it's not always obvious to everyone reading about politics how to synthesize the conventional wisdom into a number.  X is the frontrunner to win the primary or the general election or whatever, but how much of a frontrunner?  How much of a chance does the 2nd or 3rd place guy have?  Sure, any of us who follows this stuff can come up with our own numbers, and if we track the races closely enough, we can potentially even beat the market.  But seeing what the collective CW of the market is in numerical form is still something that's of interest to some of us.

This is philosophical Mumbo jumbo. The point is, if a 90% chance of a result 30 minutes before poll closing on election night can be wrong, what is the hope of any number being predictive this far out? These markets claim to predict events, not capture conventional wisdom.

I don’t understand.  Of course a 90% chance of a result 30 minutes before poll closing can be wrong.  It *should* be wrong 10% of the time.  If it was wrong 0% of the time, then it shouldn’t be given a 90% chance.  It should be a 100% chance.  This holds whether the odds are given 30 minutes in advance or 30 years in advance.  The difference being that you have more information when the event in question is occurring tomorrow than if it’s years in advance, and so you should see more 90% or 95% or 99% odds being quoted if the event is happening imminently, because there are fewer unknowns at that point.

With regard to “conventional wisdom” vs. “likelihood of the event happening”, I think you might be making a false distinction here.  There is no such thing as an absolute “probability of an event happening” in a real world social science event like an election.  You can talk about the probability of winning a hand of poker or something like that.  There’s an objectively right answer to the question of “What is the probability of being dealt a straight flush?”  But “What is the probability of a candidate winning this election?”….unless you have access to quantum level data on the brain states of every individual voter, it doesn’t make sense to talk about an *objective* probability on the outcome.

Now you can make models using objective inputs (usually based heavily on polling numbers), like 538 does.  But there is no one “right” model.  They all make simplifying assumptions about which parameters to include and how, and while some will have better track records than others, there’s no way to prove which one is “right”.

But betting markets make no pretense of being objective.  They’re the collective subjective wisdom of whoever is doing the betting, based on whatever criteria they choose to use (and presumably many of the people betting do take models from folks like 538 into account, though obviously there are no such models yet for an event as distant as the 2020 election).  I’m not sure how anyone can think it works any differently from that.  The markets have no “special knowledge” that the rest of us don’t have access to, unless you’ve got insider trading from someone with access to internal polling numbers, or something like that.  But I’m assuming that that’s rare.

You must be logged in to read this quote.

Well, you’re just flat out wrong on this one, I’m afraid.  I frequently check betting markets to see what the conventional wisdom on elections says.  

If you’re saying that many/most people who look at these numbers think that they’re getting some kind of special knowledge out of it, that goes beyond simply aggregated conventional wisdom, then I agree that the average news consumer probably does think that way.  But the average news consumer is wrong on a lot of things.  

[emailking]

This is philosophical Mumbo jumbo. The point is, if a 90% chance of a result 30 minutes before poll closing on election night can be wrong, what is the hope of any number being predictive this far out?

A percentage chance that is not 0% or 100% cannot be shown to be right or wrong based on the eventual result. You would need a much more sophisticated argument for why the Bayesian prior should have been less than 90%. Yes that's "philosophical Mumbo jumbo", but so is talking about odds for a non-repeatable event.

[Mr. Morden]

Many betting sites are offering odds on who will win the general election, but the only one I’ve found offering Democratic nomination odds is Betway:

https://sports.betway.com/#/politics/us/us-presidential-election/985177_us-presidential-election-democratic-nominee-2020

Warren 28.6
Kaine 18.2
Booker 12.5
Cuomo 11.1
Clinton 9.1
M. Obama 9.1
de Blasio 6.7
Castro 5.9
Klobuchar 5.3
Duckworth 4.8
Harris 4.8
Sanders 4.8

[JJC]

The betting markets are totally discredited. Both Bremain and Hillary traded at 85-90% literally minutes before the decisive votes started getting reported. These reflects the biases of their urban, center-left, sources of dumb money.

Shhh! I made a lot of money on election night betting for the Donald.

Mainly in MI. Prices for a Trump win there were going for only .16/share. Unlike everyone else, who only focused on polls, I was looking at real, actual early voting data and comparing turnout statistics by county with the 2012 election. Based on the data, particularly from Wayne and Oakland county, I knew the race would be extremely tight. Definitely NOT a gimme for Clinton.

I even made a thread on here about MI being a red flag for dems;

You must be logged in to read this quote.

I kept warning dems on this site about MI (and the entire rust belt) for the entire election season.

[Mr. Morden]

Trump and Biden gain, while Warren drops:

Trump 37.0
Pence 10.0
Warren 7.7
Biden 6.2
M. Obama 5.7
Booker 3.8
Julian Castro 2.9
H. Clinton 2.9
Rubio 2.3
Sanders 2.3
Haley 2.2
Cruz 2.0
Gillibrand 2.0 (Ladbrokes)
Cuomo 1.8
Klobuchar 1.8
Ryan 1.8
Bloomberg 1.5
Kaine 1.2
Harris 1.1

Gillibrand isn’t actually listed on Betfair, so I listed her price on Ladbrokes.  She seemed like the most obvious person not included on Betfair’s list.

[jimrtex]

Trump and Biden gain, while Warren drops:

Trump 37.0
Pence 10.0
Warren 7.7
Biden 6.2
M. Obama 5.7
Booker 3.8
Julian Castro 2.9
H. Clinton 2.9
Rubio 2.3
Sanders 2.3
Haley 2.2
Cruz 2.0
Gillibrand 2.0 (Ladbrokes)
Cuomo 1.8
Klobuchar 1.8
Ryan 1.8
Bloomberg 1.5
Kaine 1.2
Harris 1.1

Gillibrand isn’t actually listed on Betfair, so I listed her price on Ladbrokes.  She seemed like the most obvious person not included on Betfair’s list.

Is there any information on volume?

[Mr. Morden]

Warren back up to a tie with Pence for second place:

Trump 38.5
Pence 8.0
Warren 8.0
M. Obama 5.7
Ryan 4.2
Booker 3.7
Clinton 3.7
Biden 3.1

[JustinTimeCuber]

well I hope there's a much more sophisticated 538 forecast for the 2020 primaries.

[TomC]

No Tulsi Gabbard? Surprised she's not on there somewhere.

[Mr. Morden]

Gillibrand’s still not listed on “Betfair Exchange” (the Intrade-like peer-to-peer version from which I normally post prices).  So instead I list her latest share price on the regular Betfair site.

Trump 41.3
Pence 8.7
Warren 8.0
M. Obama 5.3
Biden 5.0
Gillibrand 3.8 (regular Betfair)
Booker 3.4
Clinton 2.9
Sanders 2.5
Ryan 2.3
Castro 2.1

[Eraserhead]

Sanders is way too low.

[justfollowingtheelections]

Gillibrand’s still not listed on “Betfair Exchange” (the Intrade-like peer-to-peer version from which I normally post prices).  So instead I list her latest share price on the regular Betfair site.

Trump 41.3
Pence 8.7
Warren 8.0
M. Obama 5.3
Biden 5.0
Gillibrand 3.8 (regular Betfair)
Booker 3.4
Clinton 2.9
Sanders 2.5
Ryan 2.3
Castro 2.1

Gillibrand has no business being listed as a presidential candidate.

[Crumpets]

Gillibrand’s still not listed on “Betfair Exchange” (the Intrade-like peer-to-peer version from which I normally post prices).  So instead I list her latest share price on the regular Betfair site.

Trump 41.3
Pence 8.7
Warren 8.0
M. Obama 5.3
Biden 5.0
Gillibrand 3.8 (regular Betfair)
Booker 3.4
Clinton 2.9
Sanders 2.5
Ryan 2.3
Castro 2.1

Gillibrand has no business being listed as a presidential candidate.

Why?

[justfollowingtheelections]

Gillibrand’s still not listed on “Betfair Exchange” (the Intrade-like peer-to-peer version from which I normally post prices).  So instead I list her latest share price on the regular Betfair site.

Trump 41.3
Pence 8.7
Warren 8.0
M. Obama 5.3
Biden 5.0
Gillibrand 3.8 (regular Betfair)
Booker 3.4
Clinton 2.9
Sanders 2.5
Ryan 2.3
Castro 2.1

Gillibrand has no business being listed as a presidential candidate.

Why?

Convince me otherwise.