(highlight:: Prediction Markets Tend to Incorporate Information That Cannot Be Found In Other Models
Summary:
Prediction markets respond rapidly to events such as the capture of Osama bin Laden and the performance of political candidates in debates, incorporating information that traditional models cannot.
Market traders quickly factored in the potential impact of these events on the election outcome, unlike traditional models that rely on slower-changing variables. This demonstrates the ability of prediction markets to capture the influence of variables not included in conventional models, like debate performances.
This characteristic extends to various forecasting problems, including COVID and climate forecasting, making prediction markets valuable for incorporating unique and critical information.
Transcript:
Speaker 1
Prediction markets are very rapid response mechanisms. So I'll give you two examples. So one was capture and killing of Osama bin Laden. When that happened, there was a jump up in the price of the Obama contract to win reelection in 2012. So the market traders believed that this would be an asset for him as he was running for reelection. There are no variables such as opinion polls like economic indicators, popularity of the president. None of that actually changed rapidly enough to change the focus of conventional models. So you get a difference in speed of reaction and also there are certain things that can matter that markets can respond to which the models cannot simply because they're missing the Variable. The second example I'll give you is also from the same time period roughly. It was the first debate between Obama and Romney and Obama was to viewers of the debate to be very unprepared for it that he had not really bothered. He just thought he could coast through it. And it turned out in the judgment of people who are watching that debate, they felt that Mitt Romney had gotten better of him in that debate, quite substantially better. And you saw big movements in the markets. Not huge obviously because you're talking about proportional movements. The price has shifted in a noticeable way towards Romney. But then that effect dissipated over time because there were two other debates and Obama's performance was better in subsequent debates. Model is not going to be able to respond that quickly to an event that's not conventionally a part of the model, such as debate performance. So markets will respond to information that is relevant, but usually you would not find in conventional models. And that applies to COVID forecasting, climate forecasting. It applies to all kinds of things that have very important forecasting problems.)
- Time 0:10:14
- information_flow, prediction_markets, predictive_models, 1socialdont-post,

(highlight:: The Paradox of Prediction Markets: The Tendency for Influential Markets to Be Manipulated
Summary:
Prediction markets, despite their forecasting accuracy, are susceptible to manipulation due to the potential impact on public perceptions and beliefs.
The behavior of traders, such as placing large bets to influence market prices and public opinion, can significantly affect the perceived viability of candidates, potentially leading to self-fulfilling prophecies. The paradox lies in the fact that accurate prediction markets become attractive targets for manipulation because of their potential to sway people's beliefs.
Transcript:
Speaker 1
But models have limitations. Markets have their own limitations. So markets are subject to manipulation. They are subject to hurting sometimes. There's something that I call the paradox of prediction markets, which is that if a prediction markets forecasting performance is really outstanding, like if it is believed to be Extremely accurate, then there's a very strong incentive to manipulate it. So I'll go back to that 2012 election between Romney and Obama because I studied it in some detail. I looked at transaction level data. This is in collaboration with David Rothstra who's at Microsoft Research. And we looked at transaction level data on in trade at that time. And it turned out that a single trader, a single account, ended up betting about $7 million on Romney to win. And the pattern of trades was such that it looked like, we can't be sure because I don't know who this person was. This was a randomized data, but it was a single account. And about $7 million in bets were placed to for on Romney to win over the course of two years on that market. And it looked like it was somebody who was trying to prevent the Romney price from falling too far, setting a wall, if you like, to the price of Romney contract in a ceiling to the price Of the Obama contract. That was the pattern of trades that we observed there. Now, why would somebody do that? It could be that they were just very confident that Romney would win and that they felt the market was wrong and they're willing to put a lot of money behind that. But it could also be that they wanted to affect public perceptions of whether Romney was viable as a candidate. Now, why would they want to do that? Well, beliefs about whether somebody is a credible viable candidate affect things like a morale, turnout, volunteer effort, donations, all kinds of things that could actually affect The objective probability of winning. In fact, if you can convince people that somebody is not a viable candidate, that might actually turn into a self-fulfilling prophecy. They might actually lose because people will lose enthusiasm. They'll stop donations. They'll stop working. And so manipulation of prediction markets is something that one has to think about a little bit. And so the paradox of prediction markets is this. If the market is believed to focus accurately, it's worth manipulating because people who your manipulation will actually move people's beliefs.)
- Time 0:11:55
- market_manipulation, prediction_markets, 1socialdont-post,

(highlight:: The Depolarizing Nature of Prediction Markets
Summary:
Prediction markets operate opposite to social media platforms by incentivizing diversity of opinion among participants.
The market's incentive structure causes it to be depolarizing, attracting individuals with different views and preventing filter bubbles. This contrasts with social media platforms which tend to reinforce users' existing views.
The insight highlights the power of prediction markets in challenging and diversifying opinions by attracting a variety of traders, ultimately leading to a more functional market.
Transcript:
Speaker 1
But even online, prediction markets are in fact the opposite of every other platform in the following sense. If a prediction market starts to just get full of people who just absolutely convinced that Biden's going to win and they cause the price of that contract to go out to let's say 60 cents Or 70 cents to a dollar that will look absurd to people who are absolutely convinced that Trump's going to win and they will be attracted to the market. The market just could not exist in that filter bubble because it would lead prices to move in a direction that look absurd to people who are not in the market. And as they enter that will not just change the price, but it will increase the diversity of opinion of market participants. If everybody thought the same way, they really wouldn't be betting against each other and your market would basically end up with hardly any volume. So the incentive structure of markets causes them to be in some sense depolarizing and in the social media landscape or in the online electronic platform landscape, that makes them Very different from things like Twitter, X, Facebook, threads, Instagram and so on. Because if your own activity involves people who just think like you, you're not going to find many profitable bets to make and others outside of that will be attracted to that environment.
Speaker 2
That is a really powerful realization, isn't it? So you're coming up to elections or coming up to anything you're saying, don't look at Twitter, don't look at any platform that self reinforces your own views, go look at essentially Some of the challenges your own views and prediction markets do it with real money.
Speaker 1
No matter what you view, you'll find somebody on that that would challenge it because if that diversity of traders didn't exist, the market wouldn't really be functioning.)
- Time 0:24:56
- depolarization, polarization, prediction_markets, 1socialpost-queue,

(highlight:: Prediction Markets Tend to Incorporate Information That Cannot Be Found In Other Models
Summary:
Prediction markets respond rapidly to events such as the capture of Osama bin Laden and the performance of political candidates in debates, incorporating information that traditional models cannot.
Market traders quickly factored in the potential impact of these events on the election outcome, unlike traditional models that rely on slower-changing variables. This demonstrates the ability of prediction markets to capture the influence of variables not included in conventional models, like debate performances.
This characteristic extends to various forecasting problems, including COVID and climate forecasting, making prediction markets valuable for incorporating unique and critical information.
Transcript:
Speaker 1
Prediction markets are very rapid response mechanisms. So I'll give you two examples. So one was capture and killing of Osama bin Laden. When that happened, there was a jump up in the price of the Obama contract to win reelection in 2012. So the market traders believed that this would be an asset for him as he was running for reelection. There are no variables such as opinion polls like economic indicators, popularity of the president. None of that actually changed rapidly enough to change the focus of conventional models. So you get a difference in speed of reaction and also there are certain things that can matter that markets can respond to which the models cannot simply because they're missing the Variable. The second example I'll give you is also from the same time period roughly. It was the first debate between Obama and Romney and Obama was to viewers of the debate to be very unprepared for it that he had not really bothered. He just thought he could coast through it. And it turned out in the judgment of people who are watching that debate, they felt that Mitt Romney had gotten better of him in that debate, quite substantially better. And you saw big movements in the markets. Not huge obviously because you're talking about proportional movements. The price has shifted in a noticeable way towards Romney. But then that effect dissipated over time because there were two other debates and Obama's performance was better in subsequent debates. Model is not going to be able to respond that quickly to an event that's not conventionally a part of the model, such as debate performance. So markets will respond to information that is relevant, but usually you would not find in conventional models. And that applies to COVID forecasting, climate forecasting. It applies to all kinds of things that have very important forecasting problems.)
- Time 0:10:14
- information_flow, prediction_markets, predictive_models, 1socialdont-post,

(highlight:: The Paradox of Prediction Markets: The Tendency for Influential Markets to Be Manipulated
Summary:
Prediction markets, despite their forecasting accuracy, are susceptible to manipulation due to the potential impact on public perceptions and beliefs.
The behavior of traders, such as placing large bets to influence market prices and public opinion, can significantly affect the perceived viability of candidates, potentially leading to self-fulfilling prophecies. The paradox lies in the fact that accurate prediction markets become attractive targets for manipulation because of their potential to sway people's beliefs.
Transcript:
Speaker 1
But models have limitations. Markets have their own limitations. So markets are subject to manipulation. They are subject to hurting sometimes. There's something that I call the paradox of prediction markets, which is that if a prediction markets forecasting performance is really outstanding, like if it is believed to be Extremely accurate, then there's a very strong incentive to manipulate it. So I'll go back to that 2012 election between Romney and Obama because I studied it in some detail. I looked at transaction level data. This is in collaboration with David Rothstra who's at Microsoft Research. And we looked at transaction level data on in trade at that time. And it turned out that a single trader, a single account, ended up betting about $7 million on Romney to win. And the pattern of trades was such that it looked like, we can't be sure because I don't know who this person was. This was a randomized data, but it was a single account. And about $7 million in bets were placed to for on Romney to win over the course of two years on that market. And it looked like it was somebody who was trying to prevent the Romney price from falling too far, setting a wall, if you like, to the price of Romney contract in a ceiling to the price Of the Obama contract. That was the pattern of trades that we observed there. Now, why would somebody do that? It could be that they were just very confident that Romney would win and that they felt the market was wrong and they're willing to put a lot of money behind that. But it could also be that they wanted to affect public perceptions of whether Romney was viable as a candidate. Now, why would they want to do that? Well, beliefs about whether somebody is a credible viable candidate affect things like a morale, turnout, volunteer effort, donations, all kinds of things that could actually affect The objective probability of winning. In fact, if you can convince people that somebody is not a viable candidate, that might actually turn into a self-fulfilling prophecy. They might actually lose because people will lose enthusiasm. They'll stop donations. They'll stop working. And so manipulation of prediction markets is something that one has to think about a little bit. And so the paradox of prediction markets is this. If the market is believed to focus accurately, it's worth manipulating because people who your manipulation will actually move people's beliefs.)
- Time 0:11:55
- market_manipulation, prediction_markets, 1socialdont-post,

(highlight:: The Depolarizing Nature of Prediction Markets
Summary:
Prediction markets operate opposite to social media platforms by incentivizing diversity of opinion among participants.
The market's incentive structure causes it to be depolarizing, attracting individuals with different views and preventing filter bubbles. This contrasts with social media platforms which tend to reinforce users' existing views.
The insight highlights the power of prediction markets in challenging and diversifying opinions by attracting a variety of traders, ultimately leading to a more functional market.
Transcript:
Speaker 1
But even online, prediction markets are in fact the opposite of every other platform in the following sense. If a prediction market starts to just get full of people who just absolutely convinced that Biden's going to win and they cause the price of that contract to go out to let's say 60 cents Or 70 cents to a dollar that will look absurd to people who are absolutely convinced that Trump's going to win and they will be attracted to the market. The market just could not exist in that filter bubble because it would lead prices to move in a direction that look absurd to people who are not in the market. And as they enter that will not just change the price, but it will increase the diversity of opinion of market participants. If everybody thought the same way, they really wouldn't be betting against each other and your market would basically end up with hardly any volume. So the incentive structure of markets causes them to be in some sense depolarizing and in the social media landscape or in the online electronic platform landscape, that makes them Very different from things like Twitter, X, Facebook, threads, Instagram and so on. Because if your own activity involves people who just think like you, you're not going to find many profitable bets to make and others outside of that will be attracted to that environment.
Speaker 2
That is a really powerful realization, isn't it? So you're coming up to elections or coming up to anything you're saying, don't look at Twitter, don't look at any platform that self reinforces your own views, go look at essentially Some of the challenges your own views and prediction markets do it with real money.
Speaker 1
No matter what you view, you'll find somebody on that that would challenge it because if that diversity of traders didn't exist, the market wouldn't really be functioning.)
- Time 0:24:56
- depolarization, polarization, prediction_markets, 1socialpost-queue,