3 ms·

>How is it that this one exception differs from majority of sample which can be explained by hypothesis X, that clearly disproves hypothesis X. This is a simil

>How is it that this one exception differs from majority of sample which can be explained by hypothesis X, that clearly disproves hypothesis X. This is a similar argument to "this year is cooler than the last year so climate change is a hoax." Taking a single exception as a stand-in for the mean IS an invalid argument. At best, it points out that the world is more complicated than explanation X, but it certainly doesn't invalidate X for the sample it is predictive for.

SilasX 9y ago
Well, it depends on what the claim is. If you're claiming that overall temperate (or energy content etc) of the atmosphere is on an upward trend because of specific drivers, then one anomalous year doesn't refute that (so long as the rest of the years and samples are in line with the claim/theory). If you're claiming that a certain ubiquitous dynamic exists that forces all providers in one market to do X to stay profitable, then it would refute that claim to point to one provider that is able to consistently, year after year, stay profitable without doing X.
- noobermin 9y ago
  No one is making the argument with ∀. It would be preposterous to do that in the real world because "for all" quantifiers only work in mathematics. Almost always when someone says "all" in common conversation they either mean some form of "almost always" or "a good majority"; Otherwise, they are thinking too simplistically if they really think it applies to a true "all." Taking exception to "no true Scotsman" arguments of course.
  - SilasX 9y ago
    I know they didn't say "for all", but it's functionally equivalent when the claim is that the very nature of the market forces providers into these patterns. That explanation would indeed be refuted by a persistent counter example.
    - noobermin 9y ago
      Yes, I grant that a persistent counter example would point to other factors that could exist. Counter examples usually point to "the world is more complicated" rather than disprove X if X applies to a majority the set it talks about, as I said. I'd imagine if X explains a good fraction of what you're talking about, you'd want to pursue X as an explanation over chasing explanations that effect the exceptions.
      - SilasX 9y ago
        Again, depends on the claim. If the claim is that "factor X dominates to the point that it can't be ignored", then a persistent counterexample refutes it. If the claim is that "factor X makes it so that the most reliable business model is to do Y", then it wouldn't. My reading of the article is much closer to the former.

SilasX 9y ago

Well, it depends on what the claim is. If you're claiming that overall temperate (or energy content etc) of the atmosphere is on an upward trend because of specific drivers, then one anomalous year doesn't refute that (so long as the rest of the years and samples are in line with the claim/theory). If you're claiming that a certain ubiquitous dynamic exists that forces all providers in one market to do X to stay profitable, then it would refute that claim to point to one provider that is able to consistently, year after year, stay profitable without doing X.

noobermin 9y ago
No one is making the argument with ∀. It would be preposterous to do that in the real world because "for all" quantifiers only work in mathematics. Almost always when someone says "all" in common conversation they either mean some form of "almost always" or "a good majority"; Otherwise, they are thinking too simplistically if they really think it applies to a true "all." Taking exception to "no true Scotsman" arguments of course.
- SilasX 9y ago
  I know they didn't say "for all", but it's functionally equivalent when the claim is that the very nature of the market forces providers into these patterns. That explanation would indeed be refuted by a persistent counter example.
  - noobermin 9y ago
    Yes, I grant that a persistent counter example would point to other factors that could exist. Counter examples usually point to "the world is more complicated" rather than disprove X if X applies to a majority the set it talks about, as I said. I'd imagine if X explains a good fraction of what you're talking about, you'd want to pursue X as an explanation over chasing explanations that effect the exceptions.
    - SilasX 9y ago
      Again, depends on the claim. If the claim is that "factor X dominates to the point that it can't be ignored", then a persistent counterexample refutes it. If the claim is that "factor X makes it so that the most reliable business model is to do Y", then it wouldn't. My reading of the article is much closer to the former.