From What constitutes non-trivial analysis of available data?

The "obvious maths" fallacy

The discriminating factor is certainly not the level of mathematics, but the level of expertise needed to determine that a particular formula or method is fit to answer.

Suppose an answer applies a simple mathematical operation (to well-sourced numbers). But there is disagreement over whether that operation can or should be used the way it is. Possibly, resolving this disagreement would require an expert opinion, what should be done?

Is the proper response to downvote and comment, as with any other low-quality answer?

Or is it appropriate to also flag to add the “sources needed” notice, indicating that the author should cite a reference showing that their use of that simple math (in the way that they used it) is appropriate?

1 Answer 1


We have a specialized banner for theoretical answers and I think this qualifies.

The central argument of this answer is theoretical in nature. We do not allow answers based uniquely on common sense or pure logic. Answers which are wholly based on a theoretical model are generally downvoted and may be deleted. See FAQ: What are theoretical answers?

We also have one for original research.

This answer is based on original data analysis or non-verifiable data. It is up to the answerer to provide valid, verifiable and potentially replicable evidence. Answers which are wholly based on "original research" are generally downvoted and may be deleted. See FAQ: What constitutes original research?

Normally one of these would apply. Of course these are simply specialized versions of the general "citation needed" banner, which still applies -- they were added because in many cases people did not understand that their answer needed more sources because of their calculations.

In this particular case the OR banner seems to apply (for example, it's not clear that the study cited is representative of the whole US, but the OP assumes so).

  • 4
    But the borderline for whether an answer counts as theoretical would tend to rule out only complicated arguments. School level mathematics that can be followed uncontroversially should be allowed (especially since many they often demolish arguments based on badly presented comparisons or statistics).
    – matt_black
    Jan 16, 2017 at 13:39
  • 1
    @matt_black people commit mistakes thinking the correct math is simple all the time. It's by far the most common mistake: the correct maths are hard, but an clueless user will use the wrong maths which are simple. Since they do no understand their mistake, they will insist their math should stay because its simple. Except, they are wrong, and we will mercilessly tag and delete such answers. We can't really be bound by what you propose.
    – Sklivvz
    Jan 17, 2017 at 0:04
  • Yes people do make stupid arguments, but the question is whether the thread of logic they propose is reasonable. To rule out all arguments based on school-level mathematics is far too strict. I realise we disagree on this, but killing the arguments too early is not the right way to solve it.
    – matt_black
    Jan 17, 2017 at 0:15
  • @matt_black the problem is that the stack exchange model (a group of experts voting on answers) breaks down in these cases. I would not dream of arguing something like this on e.g. biology.se, because there you have actual biologists vetting (and mob downvoting) these kind of answers -- but for the very same reason, they do not allow, e.g. physics questions, even simple ones. On skeptics we do not have the luxury of having a specific science with a majority of experts that can vet answers; we have no limitations by topic, mostly. This means that we need to do vetting elsewhere for the 1/2
    – Sklivvz
    Jan 17, 2017 at 1:33
  • 2/2 community to function. This rule is the implementation of that necessary vetting. We can't be the only community where anybody can make calculations, and not undergo vetting by a community expert in the subject. Therefore, the vetting is done by requiring answers to refer to calculations made in a reputable source.
    – Sklivvz
    Jan 17, 2017 at 1:39
  • Now, I'm not claiming the rule is perfect, but the alternative should be a proposal on how to do this vetting better, not simply "let's not do it".
    – Sklivvz
    Jan 17, 2017 at 1:40
  • I think there is a gap caused by the need to show references to a method. A lot of simple logical deductions are so simple they are not often referenced in the proper literature making totally obvious math that any expert would agree with invalid in an answer. Maybe we need a special flag for such questions to encourage extra attention rather than immediate closure.
    – matt_black
    Jan 17, 2017 at 9:53
  • That's what we already do :) people flag these answers, we add a banner and a few comments and only eventually delete the answer. if there is only basic stuff, no one bothers. the problems arise when the basic stuff is wrong. ultimately we need to be able to nuke this stuff, but it's never immediately. the answer linked by the op is an example
    – Sklivvz
    Jan 17, 2017 at 9:57
  • @matt_black This meta question was motivated by an answer applying school-level mathematics (multiplication) to two numbers that should not have been multiplied. (To predict the number of times an event would occur, rate of the event in a population was multiplied by the size of a subset of the population. A more defensible way would have been to either multiply the rate of the event in the subset by the size of the subset, or the rate of the event in the entire population by the size of the entire population. The author has not fixed it.)
    – ff524
    Jan 17, 2017 at 20:50
  • @ff524 I thought the logical problems with the answer were dealt with reasonably by comments and logical dissection by readers. Thereby minimising the need to rule out the answers as theoretical. I was looking for another example on how long oxygen would last if photosynthesis stopped but that appears to have been deleted (sklivvz, we debated this before, is it still visible to you?)
    – matt_black
    Jan 21, 2017 at 16:08
  • @matt_black this is the deleted answer. I can't readily find any meta discussion about it.
    – Sklivvz
    Jan 21, 2017 at 16:17
  • @Sklivvz I think the debate was held in the comments on my answer. I might have taken it to meta had the OP not deleted the question. My point of disagreement was the rapid flagging of the answer as "original research" despite my contention that the math was transparent and trivial. I think our point of disagreement is what counts as "trivial": i'd argue that anyone with school-level math could follow (or dispute) my argument so it doesn't count as original research. That would be my rule on questions like this.
    – matt_black
    Jan 23, 2017 at 10:55
  • @matt_black, nay: the dispute was that while the math was not the problem, the assumptions behind it were. Specifically, your answer relied on the assumption that what is true for the whole, is true everywhere equally (fallacy of division): in other words that if there is oxygen in the atmosphere, then that oxygen is available to us.
    – Sklivvz
    Jan 23, 2017 at 11:01
  • @Sklivvz That wasn't my memory of the issue (and I thought this was dealt with easily with references to the atmosphere being well mixed).
    – matt_black
    Jan 23, 2017 at 11:03
  • @matt_black fair enough. FWIW, this was my last comment at the time, pointing out what I just said and a couple more issues: "You are assuming that we can use the whole stock in the atmosphere as our oxygen supply; you are also assuming that the amount of oxygen is all that counts, not - for example - its prompt local availability or other local effects; furthermore, you assume that if photosynthesis simply stopped, there would not be any other side effects impacting oxygen levels"
    – Sklivvz
    Jan 23, 2017 at 11:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .