"I’m pretty sure I’ve gotten more emails/texts/DMs from friends in the past six months about things I wrote at Silver Bulletin than in the nine years I spent at 538@Disney."
I've gotten more texts from friends about you in the past six months than in the nine years you spent @Disney. I think you're onto something about how pushing emails to people builds more of a connection.
Funny how that is! So many apps were marketed as getting writers closer with readers just for simple old email to circle back and be there.
Thanks for sharing your metrics. While I don't always agree with your opinions, I appreciate how thorough you are in articulating your analysis. Looking forward to The Edge - The Signal and the Noise was a classic.
I kinda hope he picks up his ball and leaves TBH. Nate made a career off TDS then proceeded to be wrong about everything else since, with all the data to back it up
And what’s so sadz is just as you were getting over your Hillary Derangement Syndrome you came down with an even worse case of BDS. A symptom of HDS is a nasty sneeze that sounds like “Benghazi”…gesundheit!
I'm glad that 2023 was very happy for you, and I hope that the plans that you've sketched out for 2024 bring you even more happiness. That's what should be prioritized over anything else. Looking forward to the book and to more work over here.
I do hope that the sports models find a good home, even if you're not the one providing or stewarding its home. I definitely missed them as the NFL playoff race this season was kicking into high gear. I could calculate the scenarios myself, but the probability of each scenario based on opponent quality is tougher.
A question on the election models: does Substack have the functionality to do what you want to do with them? That's been a long lingering question on my mind ever since you started this blog. You mentioned making a "low-fi" version here, but how low-fi do you envision that, and will that be satisfactory to what you want to do here in that regard?
Nate, have you considered getting some sort of assistant to help with things like comment moderating? I like this newsletter, but posts older than a few days have nearly unreadable comment sections - e.g. https://www.natesilver.net/p/why-liberalism-and-leftism-are-increasingly has something like 100 different spam comments from the same schizophrenic Nazi. I realize that hitting the optimal balance of stopping low-value posts from gumming up the works while still encouraging real and and times heated debate is challenging, but at a minimum banning pure spam is a no-regrets move.
Being totally honest, I would pay for Nate's takes on electoral politics and polling models but have little interest in the gambling content or other things you mentioned. So this is not an appealing list to me personally but glad you're able to follow your passions and I'm sure you'll find an audience whatever you do
>about two-thirds of readers on a typical post are coming from places other than the email list
Be aware that "from places other than the email list" probably includes Substack app notifications. I think this comment I'm making counts in your statistics as "from the email list" because I happened to be looking at my inbox on my laptop when your message came through, but most of my reading probably counts as "from places other than the email list" because I typically click on a Substack app notification on my phone.
As sad as I would be to see your election model put behind a paywall or completely privatized, I understand that it is your original work and you have full control over what happens to it.
The references to economics are making me nervous, because one of the rare times I've seen you totally strike out with an analysis was the other day when you were posting the economics graphs.
If this is related to the game theory work, that gives you a good foundation for getting into industrial organization (business competition issues). But it won't get you far with other parts of econ like macroeconomics. It's not really possible to correctly do that kind of econ commentary without spending a lot of time specializing in it.
Another potential econ landmine is that economists have very rigorous statistics training -- more than basically any other scientists -- so you have a lot of really serious competition, and the opportunities for an outsider statistician to make contributions are not that common. The point is that it would be a major challenge if that's what you had in mind.
The good news is macroeconomists are wrong about a lot of things and so if people want to weigh in they just need to back up their arguments. Another field that is all about statistics is public health and Nate weighs in on that. I actually disagreed with him in part and I back up all of my arguments. Oh, and I’m the only person in America that spotted the Tennessee anomaly with Covid deaths other than obviously the Tennessee official that corrected it.
Perfect example of the confused Dunning-Kruger type of attitude that gets laymen into trouble when they try to do economics. Public health is not in the same league as economics in terms of statistics. You'll get your ass handed to you by economists if you enter the conversation not realizing that.
I don't know what the Tennessee anomaly is and obviously don't care. Come back when you have a years-long track record of predicting inflation and unemployment.
I did a crash course on inflation a year ago after I finished crunching the Covid data in the spring of 2022. So I’m smart enough to know that I shouldn’t be expressing my opinions on Covid during a pandemic…but I have no problem looking at easily accessible data and coming to conclusions once the pandemic is over. Nobody’s life was at stake with inflation and I doubt the Fed was reading my analysis so I was free to express my opinions in real time with respect to inflation—it was mostly transitory from bullwhip effect.
I don't get it. Do you not understand that every economics grad student can also crunch covid data? What do you think researchers are doing all day? Is the argument that you know how to make a plot, therefore you should be doing macroeconomic analysis?
Do you know what I was doing during the pandemic? Also analyzing covid data. I put together a Bayesian model in Stan to make some forecasts, then stopped when (finally) an actual research group put something similar online. I also posted an analysis suggesting that we'd be locked down long term, back in April 2020 when everyone was saying we'd temporarily "flatten the curve" then get back to normal. https://skeptonomics.com/posts/outbreak-economics
So am I out here now posting about macroeconomics all the time? No! Because doing amateur covid analysis doesn't make you a macroeconomist. I have enough self-awareness not to post a bunch of half-baked nonsense on a subject I haven't taken the time to work through.
And I even have an undergrad degree in economics. Meanwhile you're BS-ing with the made-up term "bullwhip effect", which isn't even a thing in economics.
I mean, sure, you're qualified to BS with your friends at the pub about this. Why not. If your standard is, "no one cares about my analysis so I can write what I want", then go for it. But Nate Silver obviously has better things he could be doing.
I am an expert in politics having worked in GOP politics at a fairly high level during Tea Party movement. So the reason the Covid data is so easy for me to analyze has to do with my knowledge of demographics by state and county. And my expertise with respect to inflation of the last several years is based on the fact I was part of the right wing echo chamber and so I know how it works.
That is the opposite of expertise in macroeconomics. It's been obvious from the first comment you made that you're having a Dunning-Kruger moment about this. This entire conversation gives me second-hand embarrassment
I trust your election modeling more than any other forecaster out there so I hope we see a lot of it discussed here at the newsletter for paying subscribers like me.
I love the newsletter. I read most of them (I’m not a poker player or big gambler). I recommend the newsletter to friends, especially those interested in politics. I’m looking forward to your data-driven election content, and would also be interested in reading your thoughts about college football if you have time and interest. I think NIL and the transfer portal will drive significant change, as will the 12-team playoff and would be interested in your thoughts. I very much appreciate having access to your insights.
Thanks for letting us know your plans. I would be much happier reading one or two thoughtful pieces a week as opposed to 4 pieces of product that you felt compelled to get out. Looking forward to you new book as well as your future insights. Go Blue?
538 was transparent enough in its methodology you can capture most of it on a spreadsheet. the most labor intensive part is the pollster ratings, and those only add significant value when the polling is thin
Tbh I'll probably mostly wait and see what happens in November. No real desire to visit new 538/ABC or RCP, though I'll probably get weak and check one of those a few times.
Wow I love Maria Konnikova -- looking forward to listening to the podcast and readying the book!
"I’m pretty sure I’ve gotten more emails/texts/DMs from friends in the past six months about things I wrote at Silver Bulletin than in the nine years I spent at 538@Disney."
I've gotten more texts from friends about you in the past six months than in the nine years you spent @Disney. I think you're onto something about how pushing emails to people builds more of a connection.
Funny how that is! So many apps were marketed as getting writers closer with readers just for simple old email to circle back and be there.
Thanks for sharing your metrics. While I don't always agree with your opinions, I appreciate how thorough you are in articulating your analysis. Looking forward to The Edge - The Signal and the Noise was a classic.
Paywalling is a necessary evil, but it's nice to leave comments open on the free articles. Will do our best to be civil. Here's a primer on anon culture: https://yuribezmenov.substack.com/p/how-to-stay-anonymous-top-anons-creativity-trust
Make sure you bring an apple for teacher. Kiss ass! ;)
I kinda hope he picks up his ball and leaves TBH. Nate made a career off TDS then proceeded to be wrong about everything else since, with all the data to back it up
This is bizarrely, fractally wrong.
You have BDS and Substack is just a bunch of halfwits making money off people suffering from BDS that have $5/month to flush down the toilet.
Holy shit you Dem fellating clowns will steal everything- there’s only one BDS: Boycott Divestment and Sanctions - against Israel.
And what’s so sadz is just as you were getting over your Hillary Derangement Syndrome you came down with an even worse case of BDS. A symptom of HDS is a nasty sneeze that sounds like “Benghazi”…gesundheit!
I'm glad that 2023 was very happy for you, and I hope that the plans that you've sketched out for 2024 bring you even more happiness. That's what should be prioritized over anything else. Looking forward to the book and to more work over here.
I do hope that the sports models find a good home, even if you're not the one providing or stewarding its home. I definitely missed them as the NFL playoff race this season was kicking into high gear. I could calculate the scenarios myself, but the probability of each scenario based on opponent quality is tougher.
A question on the election models: does Substack have the functionality to do what you want to do with them? That's been a long lingering question on my mind ever since you started this blog. You mentioned making a "low-fi" version here, but how low-fi do you envision that, and will that be satisfactory to what you want to do here in that regard?
Nate: 12% conversation to paid IS fantastic! I think 4 or 5% is more usual.
Nate, have you considered getting some sort of assistant to help with things like comment moderating? I like this newsletter, but posts older than a few days have nearly unreadable comment sections - e.g. https://www.natesilver.net/p/why-liberalism-and-leftism-are-increasingly has something like 100 different spam comments from the same schizophrenic Nazi. I realize that hitting the optimal balance of stopping low-value posts from gumming up the works while still encouraging real and and times heated debate is challenging, but at a minimum banning pure spam is a no-regrets move.
I really miss the Sports statistics and percentages that were on 538. Is there anyone doing something similar now?
Lamar Jackson is good…you’re welcome!
Being totally honest, I would pay for Nate's takes on electoral politics and polling models but have little interest in the gambling content or other things you mentioned. So this is not an appealing list to me personally but glad you're able to follow your passions and I'm sure you'll find an audience whatever you do
Mama needs a new pair of shoes!! Lucky 7, no snake eyes!!
>about two-thirds of readers on a typical post are coming from places other than the email list
Be aware that "from places other than the email list" probably includes Substack app notifications. I think this comment I'm making counts in your statistics as "from the email list" because I happened to be looking at my inbox on my laptop when your message came through, but most of my reading probably counts as "from places other than the email list" because I typically click on a Substack app notification on my phone.
Usually the email notification makes me open up the app and read it there.
As sad as I would be to see your election model put behind a paywall or completely privatized, I understand that it is your original work and you have full control over what happens to it.
The references to economics are making me nervous, because one of the rare times I've seen you totally strike out with an analysis was the other day when you were posting the economics graphs.
If this is related to the game theory work, that gives you a good foundation for getting into industrial organization (business competition issues). But it won't get you far with other parts of econ like macroeconomics. It's not really possible to correctly do that kind of econ commentary without spending a lot of time specializing in it.
Another potential econ landmine is that economists have very rigorous statistics training -- more than basically any other scientists -- so you have a lot of really serious competition, and the opportunities for an outsider statistician to make contributions are not that common. The point is that it would be a major challenge if that's what you had in mind.
The good news is macroeconomists are wrong about a lot of things and so if people want to weigh in they just need to back up their arguments. Another field that is all about statistics is public health and Nate weighs in on that. I actually disagreed with him in part and I back up all of my arguments. Oh, and I’m the only person in America that spotted the Tennessee anomaly with Covid deaths other than obviously the Tennessee official that corrected it.
Perfect example of the confused Dunning-Kruger type of attitude that gets laymen into trouble when they try to do economics. Public health is not in the same league as economics in terms of statistics. You'll get your ass handed to you by economists if you enter the conversation not realizing that.
I don't know what the Tennessee anomaly is and obviously don't care. Come back when you have a years-long track record of predicting inflation and unemployment.
I did a crash course on inflation a year ago after I finished crunching the Covid data in the spring of 2022. So I’m smart enough to know that I shouldn’t be expressing my opinions on Covid during a pandemic…but I have no problem looking at easily accessible data and coming to conclusions once the pandemic is over. Nobody’s life was at stake with inflation and I doubt the Fed was reading my analysis so I was free to express my opinions in real time with respect to inflation—it was mostly transitory from bullwhip effect.
I don't get it. Do you not understand that every economics grad student can also crunch covid data? What do you think researchers are doing all day? Is the argument that you know how to make a plot, therefore you should be doing macroeconomic analysis?
Do you know what I was doing during the pandemic? Also analyzing covid data. I put together a Bayesian model in Stan to make some forecasts, then stopped when (finally) an actual research group put something similar online. I also posted an analysis suggesting that we'd be locked down long term, back in April 2020 when everyone was saying we'd temporarily "flatten the curve" then get back to normal. https://skeptonomics.com/posts/outbreak-economics
So am I out here now posting about macroeconomics all the time? No! Because doing amateur covid analysis doesn't make you a macroeconomist. I have enough self-awareness not to post a bunch of half-baked nonsense on a subject I haven't taken the time to work through.
And I even have an undergrad degree in economics. Meanwhile you're BS-ing with the made-up term "bullwhip effect", which isn't even a thing in economics.
I mean, sure, you're qualified to BS with your friends at the pub about this. Why not. If your standard is, "no one cares about my analysis so I can write what I want", then go for it. But Nate Silver obviously has better things he could be doing.
I am an expert in politics having worked in GOP politics at a fairly high level during Tea Party movement. So the reason the Covid data is so easy for me to analyze has to do with my knowledge of demographics by state and county. And my expertise with respect to inflation of the last several years is based on the fact I was part of the right wing echo chamber and so I know how it works.
That is the opposite of expertise in macroeconomics. It's been obvious from the first comment you made that you're having a Dunning-Kruger moment about this. This entire conversation gives me second-hand embarrassment
I trust your election modeling more than any other forecaster out there so I hope we see a lot of it discussed here at the newsletter for paying subscribers like me.
I love the newsletter. I read most of them (I’m not a poker player or big gambler). I recommend the newsletter to friends, especially those interested in politics. I’m looking forward to your data-driven election content, and would also be interested in reading your thoughts about college football if you have time and interest. I think NIL and the transfer portal will drive significant change, as will the 12-team playoff and would be interested in your thoughts. I very much appreciate having access to your insights.
Thanks for letting us know your plans. I would be much happier reading one or two thoughtful pieces a week as opposed to 4 pieces of product that you felt compelled to get out. Looking forward to you new book as well as your future insights. Go Blue?
For other readers, what's your likely go-to election model if Nate doesn't make one available?
538 was transparent enough in its methodology you can capture most of it on a spreadsheet. the most labor intensive part is the pollster ratings, and those only add significant value when the polling is thin
Tbh I'll probably mostly wait and see what happens in November. No real desire to visit new 538/ABC or RCP, though I'll probably get weak and check one of those a few times.
You don’t want to read endless speculation for 10 months?? Strange. Btw, Super Bowl coverage begins next week!! I can’t wait for the speculation!!!
Love it! Eager to see what you produce.