Mar 16, 2022·edited Mar 16, 2022Liked by Robert de Neufville
While there's a large practical difference between 10% and 4% (my current prediction is 4%) the costs of being wrong are so bad that getting a single sub 7% prediction wrong basically throws you out of the running in the good judgement open.
Revolutions have much more wind up before they start happening, if a revolution that would remove putin from power would happen it likely would have started about 1 week ago. (It takes a really long time to get the momentum for these sorts of things).
Putin is closer to 3-4% to "die" of natural causes based on the tables I could find. https://apps.who.int/gho/data/view.main.61360?lang=en Maybe Putin has good medical care, but from what I can read medical care doesn't extend life all that much. Even if we us US death rates instead, he's still in the high 2%s
Though in spite of believing he has a higher chance of natural death than you put it, I think there's a lower chance of some sort of overthrow than you have it. As you rightfully point out dicatator's rarely if ever get overthrown. Putin is less likely than the base rate to be overthrown, as you yourself point out he has been better than average at stamping out opposition.
Thus I give putin a 95% chance to remain leader of russia next year. (while our forecasts don't differ much in magnitude they do differ in composition)
Thanks, Edmund. I saw the same actuarial tables you did. I don't think medical care will make much difference if Putin gets an aggressive cancer, but my understanding is that wealthier people do live longer. Recent US presidents have lived a strikingly long time. The main thing though is that I'm basically forecasting 3/4 of a year, which already brings it down to around 3%.
I know what you mean about the high penalty you pay if you're wrong about a single low percent forecast. If there's a 7% chance of something, 1 out 14 times it will really hurt your score. Even the best forecasters can do badly for a while just because of luck. But if you make a large enough number of forecasts, it should pay off if your forecasts are well calibrated.
While there's a large practical difference between 10% and 4% (my current prediction is 4%) the costs of being wrong are so bad that getting a single sub 7% prediction wrong basically throws you out of the running in the good judgement open.
Revolutions have much more wind up before they start happening, if a revolution that would remove putin from power would happen it likely would have started about 1 week ago. (It takes a really long time to get the momentum for these sorts of things).
Putin is closer to 3-4% to "die" of natural causes based on the tables I could find. https://apps.who.int/gho/data/view.main.61360?lang=en Maybe Putin has good medical care, but from what I can read medical care doesn't extend life all that much. Even if we us US death rates instead, he's still in the high 2%s
Though in spite of believing he has a higher chance of natural death than you put it, I think there's a lower chance of some sort of overthrow than you have it. As you rightfully point out dicatator's rarely if ever get overthrown. Putin is less likely than the base rate to be overthrown, as you yourself point out he has been better than average at stamping out opposition.
Thus I give putin a 95% chance to remain leader of russia next year. (while our forecasts don't differ much in magnitude they do differ in composition)
Thanks, Edmund. I saw the same actuarial tables you did. I don't think medical care will make much difference if Putin gets an aggressive cancer, but my understanding is that wealthier people do live longer. Recent US presidents have lived a strikingly long time. The main thing though is that I'm basically forecasting 3/4 of a year, which already brings it down to around 3%.
I know what you mean about the high penalty you pay if you're wrong about a single low percent forecast. If there's a 7% chance of something, 1 out 14 times it will really hurt your score. Even the best forecasters can do badly for a while just because of luck. But if you make a large enough number of forecasts, it should pay off if your forecasts are well calibrated.
"An obviously correct but impossible proposal to impose a punitive tax on Russian oil and gas".
Please unconfuse me. When Trump imposed a punitive tax on Chinese goods, that was supposed to be a bad idea. Or not?
I just came across the line you quoted on a different Substack that links here. I think you probably meant to comment on that article, not mine.
What are you quoting? This article doesn’t talk about a proposal for a tax on a Russian oil and gas.
Yes, you're absolutely right. Sorry about that. Must have been a wrong tab.