> A system which stops 50% of incoming missiles means that only half as many people will die in a nuclear war, That doesn't make sense. The population isn't ev

> A system which stops 50% of incoming missiles means that only half as many people will die in a nuclear war, That doesn't make sense. The population isn't evenly distributed, and a target can have multiple missiles directed at it. New START set the number of warheads to 1,550. The 500 largest US cities have a population (2016) of 103,020,808. https://www.cdc.gov/places/about/500-cities-2016-2019/pdfs/500-cities-by-state.pdf https://www.cdc.gov/places/about/500-cities-2016-2019/pdfs/5... . (Note LA, Long Beach, and Anaheim are different cities, when I should be using metropolitan areas, so my estimate is low-balling things.) With 6x coverage of the largest cities, and 2x coverage of the smallest, that's a good chance of ~100 million deaths, even with a 50% missile intercept rate. While a 0% intercept rate will still leave many people alive, in the sparsest parts of the US - not a factor of 2. > [footnote] but the number of targets is cut by 50% either way Note how targets are places or cities, not people. Nuking Gann Valley, South Dakota kills far fewer people, and causes far less disruption, than nuking Seattle. > Obviously, these are one-off tests in ideal conditions Last I knew, we had no good numbers for realistic conditions. I last looked at this years ago. At the time they were doing things like including a transmitter with the missile, and excluding chaff and other counter-measures. Clearly if real-world numbers were a 5% then the entire calculation is wrong. > The last half-century has seen very little in the way of better decoy technology, and massive improvements in discrimination. Citation needed in how those improvements affect the overall intercept rate, and the cost/benefit of running an expensive defense system vs. building a few more nukes. > In practice, no leader this side of Stalin is likely to court those kind of losses unless they really need to, no matter how much they stand to gain from it Really now? No one like Curtis LeMay - willing to nuke Vietnam if that were needed to end the war - will ever become president? And ... won't that encourage North Korea to build more bombs? Because the argument seems to be that with 99% effectiveness, the 30 to 40 nuclear weapons they currently have gives a ~33% of having a strike in the US. So the US could nuke North Korea and risk it. Which means North Korea should build more nukes. Ditto for every other smaller nuclear power. > whose opposition often seemed to flow more from political opposition to Reagan than from a technical evaluation of the merits of the system We've spent how many billions of dollars since the 1980s, to build a missile defense system, and we still don't have something that could handle a 1980s-era Soviet attack? Yeah, I'm going to say the critics were right.