"Putting your money where your mouth is": Betting on Politics

Dec 2018
6,976
4,584
the Heart of America
We've all seen polls about attitudes, opinions, beliefs, etc regarding politics. Depending upon the methodology and reputation of the pollster, some are more accurate than others.

One poll that I think is extremely reliable is are gambling websites since people literally put their money where their mouths are. Please note that a gambler is playing the odds; regardless of their opinion of Trump, when putting up money on the probability of his impeachment, they'll bet on the most likely outcome, not what they desire that outcome to be.

The odds quoted on these websites is read as the amount of money spent to earn one dollar. Example; currently, if a person is betting on whether or not Trump will still be President at the end of 2019, the cost is 88 cents per dollar of potential winnings. This means it's very likely he'll still be President. OTOH, if betting on whether or not there will be a US recession by the end of 2019, the cost of 5 cents per dollar of winnings which means the odds are very low. Better to bet there won't be a recession (95 cents) and earn the extra nickel.



The elections page is very interesting to me: PredictIt

Note that Warren is in the lead with a better than 50/50 chance of securing the nomination but that there's a 41% chance Trump will win the election and only a 35% chance Warren will win.

I think it's more accurate to follow the trends in such websites than in traditional polls.
 
May 2013
19,656
19,640
N Oregon Coast
Yeah, the oddsmakers have a tendency to be more accurate than the pollsters. They always have. As I recall they had Trump winning the EC by nearly spot on the margin.

 
  • Like
Reactions: Bad Bob
Dec 2018
6,976
4,584
the Heart of America
Yeah, the oddsmakers have a tendency to be more accurate than the pollsters. They always have. As I recall they had Trump winning the EC by nearly spot on the margin.

One theory on why betting polls are more accurate than regular polls is because people are already voting with their money whereas a poll is just an opinion and not an actual true representation of who will or won't vote.

Example, Trump's favorability among Gen X and Millennials is very low; much lower than the low national average. That doesn't mean squat if they don't actually show up on Election Day. Boomers do show up on Election Day in higher-than-national-average numbers. Trump's favorability among them is higher than the national average.

3/20/2019





1570300949247.png
 
May 2013
19,656
19,640
N Oregon Coast
  • Like
Reactions: Bad Bob
Dec 2018
6,976
4,584
the Heart of America
Overall - whether gambling odds or traditional polls - Things ain't going so well for the Mango Menace.

Note that the gambling poll only had a 5% chance of a recession this year. Of course, there's always 2020...which your link stated.
 
Dec 2018
6,976
4,584
the Heart of America
Overall - whether gambling odds or traditional polls - Things ain't going so well for the Mango Menace.

On further study, there's this: PredictIt

40% chance of a recession during Trump's first term.
 
  • Like
Reactions: BoiseBo
May 2012
70,330
14,236
By the wall
If polling is done accurately it is around 95% accurate. Some types of polling can even get you to 100% accuracy as we've seen Nate Silver do a couple of times.

Odds makers use the same statistical models that many pollsters do, it's all just math, and they change the probability sometimes a lot or just a little to give the house a better advantage so those odds you are seeing are not the actual probability, it's been altered.

Pollsters don't do that.

Experienced betters know the difference between the actual statistic and how much the casino has changed it and that is what they bet on, it's called value betting. The better they can guess how much a casino has manipulated the results the odds there chances go up and making a successful bet.

Therefore it doesn't really matter what odds Vegas is giving for anything because you don't know what the real value is.
 
Dec 2018
6,976
4,584
the Heart of America
If polling is done accurately it is around 95% accurate. Some types of polling can even get you to 100% accuracy as we've seen Nate Silver do a couple of times.

Odds makers use the same statistical models that many pollsters do, it's all just math, and they change the probability sometimes a lot or just a little to give the house a better advantage so those odds you are seeing are not the actual probability, it's been altered.

Pollsters don't do that.

Experienced betters know the difference between the actual statistic and how much the casino has changed it and that is what they bet on, it's called value betting. The better they can guess how much a casino has manipulated the results the odds there chances go up and making a successful bet.

Therefore it doesn't really matter what odds Vegas is giving for anything because you don't know what the real value is.
Agreed. Usually within a margin of 3%...which is within the margin of error for reputable polls during the 2016 election.

Still, I like the idea of people putting their money where their mouth is; it's called "having skin in the game". Talk is cheap.

The problem with traditional polls, as pointed out earlier, is that just because a person supports Candidate A or hates Candidate B doesn't mean they'll show up on election day. This is what happened in 2016: many people hated Hillary but wouldn't vote for Trump so they stayed home. At first guess, I'd say 3% of them. :D
 
May 2012
70,330
14,236
By the wall
Agreed. Usually within a margin of 3%...which is within the margin of error for reputable polls during the 2016 election.

Still, I like the idea of people putting their money where their mouth is; it's called "having skin in the game". Talk is cheap.

The problem with traditional polls, as pointed out earlier, is that just because a person supports Candidate A or hates Candidate B doesn't mean they'll show up on election day. This is what happened in 2016: many people hated Hillary but wouldn't vote for Trump so they stayed home. At first guess, I'd say 3% of them. :D
To say the polling was wrong would not be an accurate statement. I believe polling correctly figured how all except about 3 states or so were going to go and Hillary barely lost those, they got all the others correct. The part of the polling that wasn't accurate were questions about things like "who do you think will win the presidency'. Nate Silver explains it fairly well in his apology letter he wrote after only giving Trump a 2% chance to win the nomination.

It was more the analysis of the reporting that was wrong than the actual polling.
 
  • Like
Reactions: Bad Bob
Dec 2018
6,976
4,584
the Heart of America
To say the polling was wrong would not be an accurate statement. I believe polling correctly figured how all except about 3 states or so were going to go and Hillary barely lost those, they got all the others correct. The part of the polling that wasn't accurate were questions about things like "who do you think will win the presidency'. Nate Silver explains it fairly well in his apology letter he wrote after only giving Trump a 2% chance to win the nomination.

It was more the analysis of the reporting that was wrong than the actual polling.
Agreed. The polling wasn't wrong. Most were within their limits of error. I think some people who don't understand polls ran amok with the results.

A great way to determine the validity of a poll is, as you pointed out, the questions it asks. Although I don't see a problem with a question like "who do you think will win the presidency", I do have a problem with leading or loaded questions like "who do you think would be the best President, Trump or Rudy Giuliani?" Then exclaim "75% of Americans prefer Trump!"