Nate Silver is perhaps the most famous oddsmaker in the world. He first rose to fame by correctly predicting nearly every election in 2008, then repeating the feat in 2012. Then, when Hillary Clinton lost the 2016 election, Silver was the subject of a lot of (misguided) scorn, even though he predicted she had a 71% chance of winning.
Silver has now published a new book, “On the Edge: The Art of Risking Everything,” which focuses on people who, like Silver, make a living by calculating odds and placing bets. This includes not only professional gamblers, but also entrepreneurs, venture capitalists, and cryptocurrency speculators. (By the way, Silver is also a big gambler. He is an avid poker player and bet $1.8 million on NBA games in one year, earning a huge profit of $5,242.)
In this book, Silver wants to educate readers about risk-taking and how it applies to people like Sam Bankman-Freed and Sam Altman, and why people in government, academia, the media and other fields are so risk-averse.
But what the world wants to know is who Silver thinks will win the 2024 election. Silver is filling that appetite by running Substack, a social networking site that provides detailed models to subscribers and more general updates to freeloaders.
In a recent conversation, we talked about all that, but I also play into the possibilities here with edited excerpts that focus on his thoughts on the election, plus a look back at his career and a look ahead to the future.
I know you don’t want to be known as a “political forecaster,” but I’m curious how you view the election. Your latest prediction has Trump and Harris neck and neck, but Kamala Harris seems to be gaining momentum. You also wrote that normally we would expect her recent surge to stall, but given how close the election is, that may not be the case. What are your thoughts on that now?
If it was November [Silver’s election model] They’ll be very aggressive. They’ll say that Harris is on the rise and there’s not enough time to reverse.
As it is still only August, we are already seeing significant fluctuations in momentum that call for greater caution.
I think it’s pretty safe to expect the polls to continue to favor her over the next few weeks, and since she has her caucus and Harris and Tim Walz are relative unknowns, there may be a bit more excitement than usual.
Polls usually heat up after party conventions, so she might have a three, four, five point lead in the national polls, and then the September debate, or debates, will be the reality check.
But yeah, I thought she was an underdog too, just not as underdog as Biden, but I think right now you probably want to be in her position.
It’s a very close race. This is no longer an election for Trump to lose. Someone else has to take a chance and win.
Long before the Biden-Trump debate, going back at least a year, you have consistently argued that Biden was incapable of running a real campaign, that his age made him unlikely to win, and that Democrats should have replaced him.
Were you worried about what would happen if you got it all wrong? That people would be as angry with you as they were when you said Hillary Clinton had a overwhelming chance of beating Trump in 2016?
In 2016, I thought people were denying the fact that Trump actually had a pretty good chance. Trump had lost these key states by just a few points, and Clinton had all kinds of signs of weakness. So I was trying to make the point that Trump had a chance. [the fact that people didn’t understand that] I was frustrated.
This time, we were more skeptical of Biden than the polls suggested. Our model had Biden’s chances of winning down to around 26%, but that assumes he runs a normal campaign and doesn’t try to avoid media appearances or let his fundraising dry up. So we thought Biden’s actual chances of winning might be 10% or 15%.
As a poker player, it’s like going all in with an ace against a king. A really good bet. Biden is probably going to lose.
And, certainly, I was taking some risk by being pretty vocal about it – a calculated risk.
But that’s also how I felt as a citizen. Usually these things don’t have one motivation, right? I was very frustrated that I felt there wasn’t a responsible alternative to Trump. And I wasn’t going to vote for Biden. I was going to vote for the Libertarian Party or the Working Families Party or something.
Now, why wouldn’t you want to be known as an expert in predicting elections? Why wouldn’t you want to be known as someone who’s really good at predicting elections?
If I’m to flatter myself, it’s like if you’re an indie band that has a big hit, there’s always pressure to play that hit, and that’s part of it.
One reason is that we live in a world where people may not understand probability as well as I do, so if on Election Day it’s 80/20 that Harris will win and Trump will win, or 80/20 that Trump will win and Harris will win, that’s a career risk.
I first heard about you when you were in an indie band and running your own blog as an otaku idol. Then you worked for two big media companies, The New York Times and Disney, and now you’re independent again. Looking back, do you think it made sense to work for a big company, or do you think a solo career would have been better?
I think I’m naturally suited to working solo, and Substack’s business model is also very attractive from a cash flow standpoint: all revenue is paid up front and we have an 80% gross margin.
But I would like to do a hiring drive soon, and I almost certainly will, but instead of 35 people, it’ll probably be around six.
I have friends who have had this experience in companies of all sizes, from small to medium to large. They were the ones who got lucky and had an idea and became CEO. I don’t think they’re very happy in their professional lives. They spend all their time dealing with internal politics, external politics, compliance, litigation, investor relations. These are not my forte, and I don’t want to spend the rest of my career that way.
So it’s about embracing what you can do. Let’s say I can only hire six people. That’s a constraint. And let’s say I have to write half the signatures. Keep those two constraints in mind and build the best business you can from there. I think constraints are sometimes a good thing.
Can you tell us how many paid subscribers you have on Substack?
That’s great. That’s a very healthy number.
