2 ms·

It stops being a fallacy and starts to become a valid consideration when the search for the thing not found has been thorough enough that it would have been ver

by btilly 5y ago

It stops being a fallacy and starts to become a valid consideration when the search for the thing not found has been thorough enough that it would have been very likely to be found if it had been there. Suppose that I remember leaving $5000 in plain sight on a table. My initial priors are a 90% chance that it is there, a 9% chance that I put it somewhere else by accident, and a 1% that my friend Bob took it. I conduct a search that would have found it with 100% odds if it was on the table, 95% odds if I had put it somewhere else by accident, and a 0% chance if Bob took it. Add in the evidence of the search and all of a sudden the new odds are 0% chance that it is where I thought it was, about a 31% chance that I put it somewhere else, and a 69% chance that Bob took it. This is the first major disease to jump to humans in the last 20 years that we haven't quickly tracked down the path to us. We're good at it. The fact that we didn't succeed this time when more effort was put into the search is something that should give us pause.

austincheney 5y ago
Logic does not cease to be fallacious merely because somebody works hard and becomes frustrated.
- jerf 5y ago
  Many of the logical fallacies are logical fallacies from an Aristotelian perspective of statements that are either True or False, and nothing in between. "True" is a high bar. Many of them dissolve when treated correctly with something like a Bayesian analysis. [1] This is one of them. Absence of evidence is, in fact, evidence of absence, contrary to the frequent assertions otherwise. It just isn't proof. The strength of the evidence is proportional to the amount of the possibility space searched and the quality of that search. It is not fallacious to observe that our normally-successful efforts to find a natural cause failing raises the probabilities of the remaining explanations, and that of those possibilities, "lab origin" has a lot of the remaining probability. It doesn't mean that natural origin is disproved, nor did it mean any of the other possibilities are certain, but it is valid to adjust ones probability estimates. (I've also learned from experience a lot of people will read this as being critical of the Aristotelian perspective or something. I am not. It is a valid logic. There are other valid logics. Many of them are more practically useful than rigid True and False, but Aristotelian logic, like Newtonian physics, is still a very important one to understand as it serves as an important limiting case for many other more complicated logics.) [1]: This is not a position statement on Bayesian vs. frequentist.
  - austincheney 5y ago
    The problem with your reasoning is that you are falsely equating absence of evidence into a qualifier for an opinion or conclusion, which it isn't. The absence of evidence only means that such evidence is missing and is in no way suggestive of anything else. All conclusions drawn from an absence of evidence are otherwise unqualified assumptions. https://en.wikipedia.org/wiki/Tacit_assumption https://en.wikipedia.org/wiki/Tacit_assumption This was the exact reasoning people bet the farm on the housing market prior to late 2008. There was no evidence the housing market ever depreciated, so with that absence of evidence the housing market must therefore not ever depreciate. Contrary to that reasoning the housing market crashed and I got my house as a foreclosure at a heavily depreciated value.

by btilly 5y ago

austincheney 5y ago

Logic does not cease to be fallacious merely because somebody works hard and becomes frustrated.

jerf 5y ago
Many of the logical fallacies are logical fallacies from an Aristotelian perspective of statements that are either True or False, and nothing in between. "True" is a high bar. Many of them dissolve when treated correctly with something like a Bayesian analysis. [1] This is one of them. Absence of evidence is, in fact, evidence of absence, contrary to the frequent assertions otherwise. It just isn't proof. The strength of the evidence is proportional to the amount of the possibility space searched and the quality of that search. It is not fallacious to observe that our normally-successful efforts to find a natural cause failing raises the probabilities of the remaining explanations, and that of those possibilities, "lab origin" has a lot of the remaining probability. It doesn't mean that natural origin is disproved, nor did it mean any of the other possibilities are certain, but it is valid to adjust ones probability estimates. (I've also learned from experience a lot of people will read this as being critical of the Aristotelian perspective or something. I am not. It is a valid logic. There are other valid logics. Many of them are more practically useful than rigid True and False, but Aristotelian logic, like Newtonian physics, is still a very important one to understand as it serves as an important limiting case for many other more complicated logics.) [1]: This is not a position statement on Bayesian vs. frequentist.
- austincheney 5y ago
  The problem with your reasoning is that you are falsely equating absence of evidence into a qualifier for an opinion or conclusion, which it isn't. The absence of evidence only means that such evidence is missing and is in no way suggestive of anything else. All conclusions drawn from an absence of evidence are otherwise unqualified assumptions. https://en.wikipedia.org/wiki/Tacit_assumption https://en.wikipedia.org/wiki/Tacit_assumption This was the exact reasoning people bet the farm on the housing market prior to late 2008. There was no evidence the housing market ever depreciated, so with that absence of evidence the housing market must therefore not ever depreciate. Contrary to that reasoning the housing market crashed and I got my house as a foreclosure at a heavily depreciated value.