3 ms·

"the advice that wearing a mask doesn't protect you very well from getting Covid has held up pretty good; they're a lot better at protecting everyone else from

"the advice that wearing a mask doesn't protect you very well from getting Covid has held up pretty good; they're a lot better at protecting everyone else from the mask wearer." They're actually no good at anything. You can compare places that were right next to each other and which differed only in mask mandates, or look at the times mask mandates were introduced and removed. Cases don't shift. See the graphs here, where the numbers from different nearby regions are overlayed on the same graph with the times of mask mandate rule changes: https://ianmsc.substack.com/p/every-comparison-shows-masks-are https://ianmsc.substack.com/p/every-comparison-shows-masks-a... That's why when you dig into the citations of a paper claiming masks work, you invariably find some invalid evidence, like they'll be citing model outputs as evidence of real-world effects, or they'll cite a study whose CI includes zero, etc.

chimprich 5y ago
This is nonsense. You can't do comparisons like that. For example, mask mandates tend to get imposed when cases are rising, and removed when they fall. To say they have no effect on cases is to confuse cause and effect. The conspiracy site you link to has no interest in looking at the data dispassionately. There's a reason why we have data science and peer review to investigate these things.
- native_samples 5y ago
  The site I link to isn't a "conspiracy site", it's a blog written by a guy who does lots of data analysis. Please, have you not learned yet that this word is meaningless anyway? There are lots of cases where mask mandates were not added and removed at the times you claim, which were geographically next to each other, and in which there was no observed difference. There are lots of cases where mask mandates were added and the trend of the case graph didn't change at all. These are evidence against the hypothesis that masks work which was, note, the standard belief prior to mid-2020. That's why public health officials all said mass masking doesn't work. It's because it doesn't.
  - chimprich 5y ago
    > The site I link to isn't a "conspiracy site", it's a blog written by a guy who does lots of data analysis. Those aren't mutually exclusive. Anyone can throw some graphs together on a blog. Why should we trust him over papers that have been peer reviewed by experts in their fields? It's a series of cherry-picked comparisons. Why is he comparing Germany to Sweden? They don't even share a border. Why not UK to France or Peru to Ecuador? Did he pick those countries because it happens to support his theory? We don't have to wonder, because he admits it: "Perhaps the clearest example of the delusional political obsession with pretending masks matter is found in Southern California." Oh, great. So why are you showing us the "clearest examples"? Why not a comparison of the other places which aren't as "clear"? How can we determine if it's statistically significant if it's not a systematic study? There are lots of other problems with that page. He's a crank.
    - native_samples 5y ago
      Because the experts aren't really experts and peer review is useless? They routinely publish totally unscientific stuff and peer reviewers don't care, I can list 5 examples off the top of my head. Why compare Germany and Sweden - because their case curves are virtually the same despite wildly different levels of masking. Because providing counter-examples is how you disprove a hypothesis. Remember that he doesn't actually need to do lots of statistical analysis of this because mask theory is a total one. It doesn't come with caveats like "sometimes in some countries it works and others it doesn't for <reasons>", it doesn't even come with effectiveness numbers. It's just masks=less COVID. To disprove a theory this simple you only need one counter example. But this guy shows lots. He makes a big pile of graphs not because it's actually necessary to prove the point, but to try and ram it home to people who are in denial about it. There isn't just one counter example, there are a huge number of counter examples, and any one of them is sufficient to invalidate "masks = universal and effective COVID countermeasure" as a belief. That actually is science! If mask mandate proponents were proposing a very subtle and complex effect where it was expected that sometimes there'd be no impact etc, then yes, the counter-argument would need to be correspondingly complex in return. That isn't the case here though.

chimprich 5y ago

This is nonsense. You can't do comparisons like that. For example, mask mandates tend to get imposed when cases are rising, and removed when they fall. To say they have no effect on cases is to confuse cause and effect. The conspiracy site you link to has no interest in looking at the data dispassionately. There's a reason why we have data science and peer review to investigate these things.

native_samples 5y ago
The site I link to isn't a "conspiracy site", it's a blog written by a guy who does lots of data analysis. Please, have you not learned yet that this word is meaningless anyway? There are lots of cases where mask mandates were not added and removed at the times you claim, which were geographically next to each other, and in which there was no observed difference. There are lots of cases where mask mandates were added and the trend of the case graph didn't change at all. These are evidence against the hypothesis that masks work which was, note, the standard belief prior to mid-2020. That's why public health officials all said mass masking doesn't work. It's because it doesn't.
- chimprich 5y ago
  > The site I link to isn't a "conspiracy site", it's a blog written by a guy who does lots of data analysis. Those aren't mutually exclusive. Anyone can throw some graphs together on a blog. Why should we trust him over papers that have been peer reviewed by experts in their fields? It's a series of cherry-picked comparisons. Why is he comparing Germany to Sweden? They don't even share a border. Why not UK to France or Peru to Ecuador? Did he pick those countries because it happens to support his theory? We don't have to wonder, because he admits it: "Perhaps the clearest example of the delusional political obsession with pretending masks matter is found in Southern California." Oh, great. So why are you showing us the "clearest examples"? Why not a comparison of the other places which aren't as "clear"? How can we determine if it's statistically significant if it's not a systematic study? There are lots of other problems with that page. He's a crank.
  - native_samples 5y ago
    Because the experts aren't really experts and peer review is useless? They routinely publish totally unscientific stuff and peer reviewers don't care, I can list 5 examples off the top of my head. Why compare Germany and Sweden - because their case curves are virtually the same despite wildly different levels of masking. Because providing counter-examples is how you disprove a hypothesis. Remember that he doesn't actually need to do lots of statistical analysis of this because mask theory is a total one. It doesn't come with caveats like "sometimes in some countries it works and others it doesn't for <reasons>", it doesn't even come with effectiveness numbers. It's just masks=less COVID. To disprove a theory this simple you only need one counter example. But this guy shows lots. He makes a big pile of graphs not because it's actually necessary to prove the point, but to try and ram it home to people who are in denial about it. There isn't just one counter example, there are a huge number of counter examples, and any one of them is sufficient to invalidate "masks = universal and effective COVID countermeasure" as a belief. That actually is science! If mask mandate proponents were proposing a very subtle and complex effect where it was expected that sometimes there'd be no impact etc, then yes, the counter-argument would need to be correspondingly complex in return. That isn't the case here though.