3 ms·

> There's no biological theory underpinning cherry picking of a handful of Sweden's neighbours. Uh... the whole premise was to cherry pick Sweden and try to ma

> There's no biological theory underpinning cherry picking of a handful of Sweden's neighbours. Uh... the whole premise was to cherry pick Sweden and try to make a broad determination based on that one data point! You don't get to demand broad statistics if you didn't do it originally.

mike_hearn 6y ago
Sweden isn't cherry picked because there are only a handful of regions that chose not to lock down and the relevant argument is about the impact of locking down or not. Therefore looking at it specifically follows naturally from the question being investigated. The converse, attempting to argue that Sweden suffered harshly by looking only at physical neighbours, doesn't follow naturally from any scientific question anyone is actually raising. For that comparison to be useful you'd need a theory of the form, "Nordic peoples are genetically superior to other peoples in ways that makes them naturally resistant to COVID, and thus whilst original models neglected that factor, Nordic Exceptionalism means Sweden should have been identical to Norway and they weren't therefore their decision was a bad one". But obviously such an argument would be nonsense and nobody is claiming that - just implying it. It's irrelevant for another reason: Sweden only looks like it did "badly" compared to Norway when using percentages. If you look at it in terms of absolute change in excess death, it's small, and depending on how you calculate, you can actually argue COVID had no effect on excess death in Sweden ([1] has an example of such calculations). The impact on Sweden of not locking down, even if you ascribe all differences between these arbitrarily chosen countries to that, is so tiny that it cannot justify the costs of the policy. At any rate, if you like there are other examples and they all support the same argument. For example, in the USA Florida is a reasonable choice. They released their restrictions relatively early without disaster. Or South Dakota (=identical curves to North Dakota despite very different levels of restrictions), or the recent example of Texas unlocking, or Belarus. Sweden is a popular choice because it's a very clear counter-example that's been widely discussed in Europe, and because - as I already noted - no model predictions or epidemiological theories make any exceptions for Sweden or Nordic countries. In fact modellers predicted specifically that Sweden would have had 90,000 deaths from COVID by May, but in fact they had about 98,000 deaths in total from all causes by the end of the year. It's therefore sufficient by itself to disprove the models. [1] https://softwaredevelopmentperestroika.wordpress.com/2021/01/15/final-report-on-swedish-mortality-2020-anno-covid/ https://softwaredevelopmentperestroika.wordpress.com/2021/01...
- newacct583 6y ago
  This is an outstanding Gish gallop, and I salute you for it. All those gaggling facts are just being used to obscure. Really all you're saying here is that the data is noisy (duh). So we can't refute your argument with local data about Scandinavia because that's "cherry picking". But when your argument is about a SINGLE DATA POINT IN ISOLATION (Sweden's results) then... that's totally valid analysis that needs to be accepted without question or worry about confounding noise? No, that's bad science. It's quite clear that nations with both good mitigation strategies and good compliance (Finland, Germany, Japan) had low case counts, and those that didn't (everywhere else) had bad outbreaks. It's true you can find outliers, but so what?
  - mike_hearn 6y ago
    There seems to be a basic logic issue here. Maybe an example will help. If you argue "all swans are white" then to disprove your theory all I have to do is identify a single black swan. One is enough, and presenting such a black swan is not "cherry picking" because the original theory was total and had no exceptions: it posited the absence of "cherries". If your response is, "well obviously all swans are white in this part of the lake" then you've lost the argument, because your original theory is the one being debated, not one with a random ad-hoc revision that makes no sense. Arguments about Sweden and lockdown are like this. The debate in question is whether lockdowns were required. Their justification is entirely based on models that claim to accurately simulate the counter-factual: that not locking down will yield enormous death numbers. To disprove this theory and thus eliminate it as the justification for lockdowns only requires one counter example, because the original theory is total and has no exceptions for proximity to Stockholm/Nordic culture/whatever today's excuse is. Now as I've pointed out, Sweden is not the only counter-example, there are quite a few in different parts of the world. But that doesn't matter because by the nature of the original claim, a single counter-example is sufficient. To re-establish a justification for lockdowns would require a revised theory that is actually coherent and explains all the exceptions, not just a single one: this is basic logic.

mike_hearn 6y ago

Sweden isn't cherry picked because there are only a handful of regions that chose not to lock down and the relevant argument is about the impact of locking down or not. Therefore looking at it specifically follows naturally from the question being investigated. The converse, attempting to argue that Sweden suffered harshly by looking only at physical neighbours, doesn't follow naturally from any scientific question anyone is actually raising. For that comparison to be useful you'd need a theory of the form, "Nordic peoples are genetically superior to other peoples in ways that makes them naturally resistant to COVID, and thus whilst original models neglected that factor, Nordic Exceptionalism means Sweden should have been identical to Norway and they weren't therefore their decision was a bad one". But obviously such an argument would be nonsense and nobody is claiming that - just implying it. It's irrelevant for another reason: Sweden only looks like it did "badly" compared to Norway when using percentages. If you look at it in terms of absolute change in excess death, it's small, and depending on how you calculate, you can actually argue COVID had no effect on excess death in Sweden ([1] has an example of such calculations). The impact on Sweden of not locking down, even if you ascribe all differences between these arbitrarily chosen countries to that, is so tiny that it cannot justify the costs of the policy. At any rate, if you like there are other examples and they all support the same argument. For example, in the USA Florida is a reasonable choice. They released their restrictions relatively early without disaster. Or South Dakota (=identical curves to North Dakota despite very different levels of restrictions), or the recent example of Texas unlocking, or Belarus. Sweden is a popular choice because it's a very clear counter-example that's been widely discussed in Europe, and because - as I already noted - no model predictions or epidemiological theories make any exceptions for Sweden or Nordic countries. In fact modellers predicted specifically that Sweden would have had 90,000 deaths from COVID by May, but in fact they had about 98,000 deaths in total from all causes by the end of the year. It's therefore sufficient by itself to disprove the models. [1] https://softwaredevelopmentperestroika.wordpress.com/2021/01/15/final-report-on-swedish-mortality-2020-anno-covid/ https://softwaredevelopmentperestroika.wordpress.com/2021/01...

newacct583 6y ago
This is an outstanding Gish gallop, and I salute you for it. All those gaggling facts are just being used to obscure. Really all you're saying here is that the data is noisy (duh). So we can't refute your argument with local data about Scandinavia because that's "cherry picking". But when your argument is about a SINGLE DATA POINT IN ISOLATION (Sweden's results) then... that's totally valid analysis that needs to be accepted without question or worry about confounding noise? No, that's bad science. It's quite clear that nations with both good mitigation strategies and good compliance (Finland, Germany, Japan) had low case counts, and those that didn't (everywhere else) had bad outbreaks. It's true you can find outliers, but so what?
- mike_hearn 6y ago
  There seems to be a basic logic issue here. Maybe an example will help. If you argue "all swans are white" then to disprove your theory all I have to do is identify a single black swan. One is enough, and presenting such a black swan is not "cherry picking" because the original theory was total and had no exceptions: it posited the absence of "cherries". If your response is, "well obviously all swans are white in this part of the lake" then you've lost the argument, because your original theory is the one being debated, not one with a random ad-hoc revision that makes no sense. Arguments about Sweden and lockdown are like this. The debate in question is whether lockdowns were required. Their justification is entirely based on models that claim to accurately simulate the counter-factual: that not locking down will yield enormous death numbers. To disprove this theory and thus eliminate it as the justification for lockdowns only requires one counter example, because the original theory is total and has no exceptions for proximity to Stockholm/Nordic culture/whatever today's excuse is. Now as I've pointed out, Sweden is not the only counter-example, there are quite a few in different parts of the world. But that doesn't matter because by the nature of the original claim, a single counter-example is sufficient. To re-establish a justification for lockdowns would require a revised theory that is actually coherent and explains all the exceptions, not just a single one: this is basic logic.