Early post-convention polling isn’t looking good for the guy in the White House: https://www.cnn.com/2020/08/30/politics/trump-biden-post-convention-polling/index.html
Trump might have gotten a 1-2 point bounce, but he will need more than that to scare Biden. There’s still time between now and November of course.
The lack of bounces does make sense given that fewer people watched the conventions this year, which is especially the case for the Republican convention. [emphasis added]
Nice subtle knife-twist there by CNN. The spectacle of Trump whining about his not-at-all-low, actually-way-better-than-Democrats ratings is one of the few bright spots in an epically shitty week.
I hope Silver is right, because I just checked the same site that prompted the OP, and the odds are even. -110 for each candidate.
Both of them could be right. The odds could look worse than average that Trump will win an election, and the betting markets could also be right that he has a 50/50 chance of being in the white house for all of 2021. That’s my exact take on probabilities as well.
The betting markets aren’t showing 50/50, though. RealClear Politics is showing the betting market average at 51.1 Biden - 48.1 Trump. Of the 7 individual markets in that average, one shows a tie - the others show Biden ahead.
This average doesn’t include Predictit for some reason. At the time I posted this, the Predictit odds are 58 Biden 45 Trump. And Biden is trending up and Trump down.
This is no time for complacently, though.
I’m not sure what measures are being used to make the race “58 to 45.”
It’s worth noting that the material difference between saying “Biden is 55 percent likely to win” and “it’s a tossup” is basically nothing at all. 55-45 is close enough to a tossup to say it’s a tossup.
Sorry to reiterate what others have been saying, but regarding discrepancies between 538 and betting markets, 538 stated that their odds are with the assumption of a fair election, so there is no way they would compare apples-to-apples with someone factoring in voter disenfranchisement.
A few screenshots and a link: Trump is upside down among the military.
Exactly. Guarded optimism. Decent numbers, but I’d like to see better .
I am baffled that anyone ascribes any special insight to betting markets. They’re guessing, same as anyone else, and they’re all working off the same poll information we all have.
Speaking of better, I noticed that yesterday’s very bad for Trump ABC/Ipsos poll (an A rated pollster according to 538.com) has not been added to the 538 polls, what’s up with that?
I think the difference is that the polls are measuring who you intend to vote for, and the betting markets are measuring who you think everyone else is going to vote for. It’s different.
Like I said, guessing. It’s second derivative information, at best. At worst, it’s people actively trying to mess around with the narrative. We know that people in the past have attempted to influence political betting markets in this way, and the actual amount of dollars involved in these markets is triflingly small.
Interesting. So not really an efficient market? Do the large books in Vegas, Europe, etc. offer these bets? If so, the creditworthiness of the books must come into play. Because why wouldn’t a private company use these markets to hedge risk? For example, a profitable company domiciled in the US could place large bet on Biden, hedging some potential income tax increases that would only be possible with a Biden victory.
And yet, prediction markets outperform polling, and debiased prediction markets outperform debiased polling.
e.g. Rothschild, Forecasting Elections: Comparing Prediction Markets, Polls, and Their Biases, Public Opinion Quarterly, Vol. 73, No. 5 2009, pp. 895–916
My analysis finds that in the 2008 election cycle FiveThirtyEight’s debiased poll-based forecasts were, on average, slightly more accurate than Intrade’s raw prediction market-based prices. But when prediction markets are properly debiased, they are more accurate and contain more information than debiased polls; this advantage is most significant for forecasts made early in the cycle and in not-certain races (i.e., the races typically of most interest).
Furthermore, I start two prediction threads every month and the most vocal anti-prediction market people never submit an entry. You can compete head to head with my prediction market based model. Show me that prediction markets are poor predictors don’t tell me.
A large US company couldn’t do anything significant. The only place where Americans are allowed to bet on politics is PredictIt, and they have a $850 maximum.
Thanks. Had no idea. Yeah, then I say the market is nowhere close to efficient.
That’s false - Bovada and MyBookie both offer many political bets.