Announcing Insight Prediction Partnership

The importance of betting on beliefs

I like arguing. I have strong views on aesthetics, morals, and politics that I think are correct. This of course implies that others are wrong, as there is no point in talking about questions on which everyone agrees.

From since at least the birth of philosophy, the wisest among us have recognized that what is believed by the majority of people or those in power isn’t necessarily what is true. Every prophet, political movement, and keyboard warrior operates on the basis of “I’m right and others are wrong.”

A few decades ago, Phil Tetlock came along and showed that expertise is mostly fake. This doesn’t mean that people who know what they’re talking about don’t exist. Rather, our society does a poor job of figuring out who these people are when deciding who to listen to.

The best systems for making decisions so far invented have involved a price signal component. Prices aggregate information, distill it into numbers, and align incentives so that people behave rationally. This means that if you want to know the value of a major corporation, or a haircut or a loaf of bread in your area, that information is available.

While modern markets are extremely developed, other kinds of decision-making remain at their infancy stage. In politics or public policy, making use of price mechanisms is not so straightforward. Democracy aggregates what voters think, and that may be the best system we have developed so far, but its fatal flaw is that it does not align incentives. The individual investor in a stock either makes a killing or goes broke depending on his own choices. The citizen votes and has an impact that rounds to zero. He therefore indulges in whatever ideas or positions make him feel good, regardless of the impact such views have in the aggregate. The worlds of punditry, academia, and expertise more generally have a similar lack of accountability problem.

For those reasons, I’ve always been a fan of event markets, which allow people to bet on what is going to happen in the real world. Last week, there was controversy over the cause of an explosion outside of a hospital in Gaza. To this layman observer, supporters of the Israelis and Palestinians both seemed to make equally plausible arguments. But early on prediction markets overwhelmingly said that it was likely the Palestinian side that was responsible, and they were correct.

For those of us who go through life constantly shaking our heads at how much foolishness is out there, prediction markets are a miracle device. If you’re truly correct about most things you care about, you’ll often find that other people either aren’t willing to put their money where their mouths are, or if they are, you can simply take that money. And if they’re right and you’re wrong, that’s valuable information too!

Given all of that, I’m excited to announce that I’m transitioning from being a guy who writes about prediction markets to one who is also involved in the business side of making them a larger part of our politics and culture. I have just come on board with Insight Prediction.

Working with them, I’ve created a series of markets that I take particular interest in. They are listed below. If you’re going to start betting on the site, please use this link to sign up, as I will get a cut of the fees. Same if you click through to the markets below. In addition, if you sign up, confirm your identity on the “My Account” page, and make at least $250 in trades in the first week, you’ll get a $50 bonus (the confirmation part is only so people don’t pull a scam by signing up through various accounts and getting multiple bonuses. Confirming your identity is not necessary to withdraw money, only to get the bonus).

Since prediction markets are, with a few exceptions, not legally authorized in the US, the website is only available to non-Americans and runs on crypto. This sounded very complex to me at first and too much to bother with, but it turns out that it’s all very easy. Some Americans use VPNs to bet on websites like this, but I can’t encourage you to do that. As for transferring funds, if you have a Coinbase account, or you belong to any crypto exchange, getting money into Insight Prediction takes literally seconds. See here for a short guide under “How do I deposit money?” You can deposit US Dollar Coins (USDC), which are pegged to the US dollar, if you don’t want to worry about fluctuations in the prices of cryptocurrencies while you’re betting. Easiest path to me seems to be to buy USDC from Coinbase, and then transfer the USDC to Insight.

Here are the markets I’m watching particularly closely.

Will RFK get 5% of the popular vote? I was expecting RFK boosters to take YES, so I thought this would be like a 40-60 market. But the market right now says 80% NO! That looks reasonably priced to me. I guess RFK boosters aren’t willing to actually bet on their beliefs, just like the anti-Israeli types. Markets really do discipline people, and those with anti-establishment and populist leanings end up looking bad whenever they can be relied upon.

Trump sweep early primary states? I think there’s a 60% chance Trump sweeps Iowa, New Hampshire, Nevada, and South Carolina. Really, as I’ve been saying for a very long time, this thing isn’t much of a contest. But the market has him at between 60% and 80% to win each of the four early states. It’s worth noting that 80% would be the same probability of him winning the nomination overall at Insight, and one would think the likelihood of a sweep would be lower, unless the market thinks something is going to happen to Trump between the early primary states and the convention, which isn’t crazy. More likely, this is just a new and low liquidity market, so the probability of a sweep will go down as it becomes established. I think Trump at 85% to win the nomination is about right, so I’d still be buying Trump shares.

Speaking of 2024 nominees, Insight has Biden higher than other prediction sites. If you think there is a decent chance of him going down, it is the best place to bet on that theory.

What will be the balance of power in Congress after the 2024 election? I’m more pessimistic about Republican prospects than most people are, because I think the pro-life movement really drags the party down, and there’s just a general sense that these aren’t normal people. That said, the Senate map is really hard for Democrats. Republicans are almost sure to win West Virginia, then they just need either Ohio or Montana to gain a majority. Both of the latter states have incumbents that have won tough elections before. I also think there’s a chance we may get a surprise like Ted Cruz losing to Colin Allred. A surprise in the other direction, say Republicans taking Michigan or Wisconsin, is unlikely because, again, abortion. With all that said, I think the odds are:

D House, R Senate: 40%

D House, D Senate: 30%

R House, R Senate: 28%

R House, D Senate: 2%

The big difference between me and the market is that it’s much more confident in a Republican sweep, so I’d bet NO on that.

Will Black Students Be Less Than 10% of the Ivy League Class of 2028? I thought of this market as a way to settle debates about the degree to which SFFA v Harvard matters. The only direct way to know for sure is to look at how the racial composition of elite universities changes. We found the black numbers for four Ivy League schools for the class of 2027, which was let in under the last admissions cycle when affirmative action was allowed:

Harvard: 14.1%

Columbia: 20%

Yale: 14%

Princeton: 9%

That’s an average of 14.3%. If next year, blacks are less than 10% on average across these schools, this market will settle as YES. I think there’s a 70% chance this will be the case. The market leans in the same direction. Do you believe I’m being naive and liberal bureaucrats will just ignore the law and do whatever they want anyway? If so, come bet.

Note that if one or more of the four schools listed above don’t report numbers, we will go with those that do. If none of them release any kind of data, we will assume it’s because the numbers are too embarrassing and the market will settle as YES, since what we’re trying to get at is whether the state of the law with regards to affirmative action matters that much.

Will Israel Control Gaza City by December 1st? We’re not used to radical movements actually losing anymore, mainly because powerful states don’t fight wars like they used to. But I think people might be surprised at what collective punishment can accomplish. At the same time, I’ve never thought that it was guaranteed that Israel would go in soon. It seems to make sense to just siege and bomb indefinitely, and Hamas should get weaker over time. But who knows, there might be political motivations to invade soon, or they might think that international pressure will get much worse as time goes on, although I have doubts about the latter worry. I think there’s a good chance Israel waits before sending in a substantial number of ground troops, but if it does in the immediate future it may take Gaza City pretty fast. I’d give this market overall a 30% chance of settling as YES.

I’ll probably be announcing more new initiatives related to prediction markets in the coming weeks and months. For now, happy betting.