I have a serious concern that good answers are being ruled out by moderators as theoretical even though they are based on a basic understanding of simple physical processes.

My concern is based on how some answers to this question (and some similar previous questions) have been handled. (a separate issue is that the question itself is currently marked as a duplicate despite being about a far more specific claim than the original question and despite looking for and generating answers that the answers to the previous question simply don't address).

Some context

The question asked whether this claim was remotely plausible (given a model that said that oxygen production in the ocean could be drastically reduced):

By 2100, the earth at sea level could have atmospheric oxygen levels comparable to the top of Mount Everest today. And as far as I know, people cannot normally stay on Everest without oxygen masks for more than a few minutes.

I understand this claim as having three components:

  • a model says oxygen production could be drastically reduced (this is (badly) addresses in a previous question leading to the claim that the new question is a duplicate)
  • assuming point 1 is true, drastic effects could be seen by 2100 on atmospheric oxygen levels
  • specifically, the oxygen at sea-level could be as low as it currently is at the top of Everest

The moderator response to the question justifying marking it as a duplicate claimed this:

I don't really understand the difference between this question and the dupe: I understand the other does not ask specifically about the 70% number, but neither does this.

It is worth noting that the previous question didn't address either the timescale or the scale of the problem but the claim in this question specifically mentioned both. The 70% is a round number of the reduction required in oxygen levels to reach the levels on Everest (which I originally assumed was so obvious it didn't need a reference).

I assumed that by providing the direct quote about the specific claim it would be clear why this was different from the original question. The moderator didn't appear to grasp this. I clarified the point but the moderator has not yet reversed this judgement. Perhaps a lack of physical intuition by the moderator inhibits what I assumed would be very obvious. This is relevant for the next part.

The problem with the answers

To me, there is a simple and direct way to address the claim which uses only easily referenceable (googleable or available on Wikipedia) facts. And two answers (not by me) initially took this approach (though I did answer a related but now deleted question using a very similar approach).

Quoting from a now erased answer:

The movement of oxygen to and from the atmosphere is called the oxygen cycle. All of the numbers used below are from this Wikipedia article, and all calculations are provided in WolframAlpha.

Using Wikipedia's numbers, the atmosphere:

  • Stores ~1.4*1018 kg of oxygen.

  • Gains ~3*1014 kg of oxygen per year.

  • About half of which is from the ocean.

  • Loses ~3*1014 kg of oxygen per year.

To me the references to our knowledge about the atmosphere establish that the stock of oxygen in it is about 5,000 times larger than the annual production and consumption rates. There is nothing theoretical about this.

We could also calculate the initial rate of reduction in oxygen levels if there were no more production. At most a complete cessation of oxygen production would lead to a reduction in concentration of 83/5000ths of the current level (based on 83 years to 2100 and using 1/5000 of the stock every year). This calculation seems to me both unproblematic and based on referenceable calculations done by atmospheric scientists (residence time of gases in the atmosphere are calculated this way, see this). But this calculation has been judged to be "theoretical" when used in an answer.

But the simple fact that the oxygen stock is vastly larger than the annual production should establish the implausibility of a claim that stopping production would reduce the concentration by 70% regardless of how we model consumption unless the claim is that a vast new sink for oxygen also occurs (but this isn't the claim).

One objection to this simple approach is that the atmosphere isn't well mixed. Again, poor intuition might lead you to assume this, but a simple google search or looking at the Wikipedia source for the facts about the atmosphere would tell you this isn't the case. If this needs to be justified, then asking for a reference is a better response than damning the answer as theoretical (simple school teaching aids show that the troposphere is well mixed).

So, it seems to me, that using some basic knowledge of atmospheric science (all of it derived directly from reliable sources) we can show that a claim that the atmosphere will lose 70% of its oxygen if we stop producing more oxygen is ridiculously implausible (unless some additional claim is made that enormous new drains of oxygen also happen which is not in this claim).

This is basic physical intuition based on directly referenced facts. It isn't theoretical (and requires no complicated model to be justified). It isn't, therefore theoretical to use these facts to answer the question.

My core concern here is that there are are many questions where simple answers of this sort are a very good way of addressing implausible claims. If we rule out this sort of answer, we are preventing good answers from emerging. The literature may regard many of these as too trivial to address in published work (because of the obvious nature of the answer) thereby leaving ridiculous contentious claims unaddressed.

A counter argument might be that bad math in simple arguments has to be justified or bad answers will slip past the net on this site. But using the "theoretical" argument is a sledgehammer to crack a nut for this type of question. Bad logical steps in these arguments are very easy to spot and address with existing rules. In this case, it seems to me, that an answer where both every fact and every calculation can be justified is still ruled out as theoretical because of the rules designed to cope with something else.

So why are moderators leaping to rule out this sort of answer as "theoretical"?

3 Answers 3


Yes, the policy does seem to be taken too far.

As the poster who provided the answer asked about in this Meta, I'd note that the removed content was basically Wikipedia's numbers showing that there's too much oxygen in the atmosphere to run out for thousands of years. I was pretty thorough in detailing the numbers, but it was still basically just me citing Wikipedia.

That said, I can't deny Sklivvz's point that this did contain some inference. Wikipedia's literal claims were how much oxygen there was in the atmosphere and how fast it was being consumed; getting from those numbers to a how-long-to-run-out required some inferences, like assuming the oxygen won't suddenly disappear tomorrow, which Wikipedia didn't actually claim. In fact, Wikipedia didn't even claim that the world won't blow up tomorrow.

Still, I think that this objection to all inferences goes too far, and can ultimately be used to reject anything. For example, since it's Sklivvz we were trying to convince, here's his top-voted post on SE.Skeptics: his answer to "Has man walked on the Moon?". As we can see, this answer provides a long series of evidence from which he infers that man has probably walked on the moon.

I think that it's valid to reference evidence to support a point. And so, I think that it's valid to point out that there's too much oxygen in Earth's atmosphere to run out any time soon, just as I think it was valid for Sklivvz to post pictures as evidence from which he inferred his own point.

  • Two wrong posts don't make a right. Using an answer from the first couple of months of the site, before the standards had been more firmly established, as a precedent is unreasonable.
    – Oddthinking Mod
    May 15, 2017 at 11:47
  • 4
    @Oddthinking Definitely, I'd agree that two wrongs don't make a right. Just, it's my position that that initial post wasn't wrong in the first place; such evidence-gathering strikes me as consistent with both the spirit of scientific skepticism and the approach used throughout this site.
    – Nat
    May 15, 2017 at 12:05
  • 1
    @Oddthinking Don't get distracted by the precedent. First ask whether the question at issue here even contravenes the theoretical answer rules (I'm not convinced it does since there is no complex chain or reasoning and every fact and calculation is referenced). And only then test whether the bad precedent is appropriate or not. I'm not sure that is true either since the moon landing answers look OK.
    – matt_black
    May 15, 2017 at 14:33
  • More discussion in chat.
    – Sklivvz
    May 18, 2017 at 13:06

Sometimes answers which assume a mathematical model and then perform calculations are permitted:

Length of uncoiled human DNA

It just depends who you are and how the powers that be feel about the situation.

  • Two wrong posts don't make a right. If you have a problem with that answer, please flag it or comment and downvote.
    – Oddthinking Mod
    May 15, 2017 at 11:49
  • 4
    @Oddthinking I think there should be more answers of the form of the DNA answer. I would only comment or downvote if there was an actual error in the answer. I previously posted a comment about a more accurate value of the number of cells, but I deleted my comment because I realized the answerer was already aware of the more accurate value from his question on BiologySE. The answer would be better if it considered whether or not there is any extracellular (cell free) human DNA or non-chromosomal cellular human DNA and gave a reference saying there is none, or included any if such exists.
    – DavePhD
    May 15, 2017 at 12:06
  • @Oddthinking so for example I could comment something like "The presence of double-stranded cell-free DNA (cfDNA) in healthy human plasma has been noted since 1948" journals.plos.org/plosone/article?id=10.1371/… but I wouldn't downvote the answer just for being theoretical.
    – DavePhD
    May 15, 2017 at 12:21
  • I'd have to agree; that answer's got a few mistakes in it, but I don't think that the mere attempt to use mathematical modeling was one of them. Perhaps the largest conceptual mistake was evaluating inequality-based statements using a chain of average values. Another was in assuming that his values were more than ballpark; for example, he used WolframAlpha's estimation for the number of cells in the human body, though this article pegs the number at 37.2% that much.
    – Nat
    May 15, 2017 at 12:44
  • Then, as you note, not all of those cells have complete DNA. Also, the number of cells in the human body varies by person, e.g. much smaller people vs. much larger people. Plus the distance from Earth to Pluto varies, etc.. Overall, it sounds like the claims asked about in that question used conservative estimates to make minimal claims, whereas he assessed the claims using means as equalities, incorrectly arriving at the conclusion that two of the three claims were incorrect. These were mistakes, but I don't object to him attempting to perform those calculations in the first place.
    – Nat
    May 15, 2017 at 12:47
  • The comment is relevant as it proves that this sort of answer can be addressed on this site (the readers are capable of spotting the factual and logical errors in a chain of simple reasoning even though they are not expert biologists). The logic in the DNA question is far longer and more complex then the logic in the atmospheric oxygen question which has been damned as being too theoretical for people to follow.
    – matt_black
    May 15, 2017 at 14:22

Are moderators ruling out good answers based on basic physical principles and facts as theoretical?

No, and the "theoretical answers" FAQ specifically addresses this point (emphasis mine):

Back-of-the-envelope calculations

Answers based on simplified calculations instead of measurements are theoretical. By nature, they implicitly assume a mathematical model, but they generally fail to show that the model is adequate to the circumstances of the question. They also don't investigate their own inaccuracy. They are a form of Original Research.

It isn't enough to add two numbers together, and use that to draw a conclusion. They must also show that these are the right numbers to add, and that adding only those two numbers is the right operation to match reality, and no other factors or complications have been omitted. The popularity of the Martingale Betting System and pyramid schemes are examples where people can be taken in by back-of-the-envelope calculations that don't consider all the issues.

In other words, it's not the simple calculation that makes the answer "theoretical", but the choice of a specific simplified model that makes this unacceptable.

What is the evidence that the model presented is at all valid? None was presented, but it's not so obvious that the model is right. By accepting such a model without question, you are simply assuming that the current oxygen in the air will be available at the surface, that no other secondary effects will significantly influence the availability and so on. Even I can see these problems with these assumptions and I am not a climate scientist.

It's easy to dismiss a claim with a back of the envelope calculation, but we ask of our users to take scientific claims seriously and answer them with scientific precision, not with a superficial dismissal.

Remember that the purpose of the site is not to do climate science, but to give referenced answers on various topics, including climate science. Allowing answers just because they are simplistic, even if apparently correct to our eyes, does not fulfil the purpose of the site: it's the opposite of what we want to build on the site!

  • 1
    So the question arises why does quoting two well-attested measurements fall into this definition. That is the essence of the answer that has been ruled out.
    – matt_black
    May 12, 2017 at 15:52
  • 1
    No Matt, the numbers have absolutely nothing to do with it. The mathematics have nothing to do with it. It's the choice of the formula.
    – Sklivvz
    May 12, 2017 at 16:04
  • 2
    They have everything to do with it and I still totally fail to get your objection. If we were arguing about whether the lifetime of oxygen in the atmosphere were 5000 years or 10000 years we would need to specify some model, but we are not. We are asking what possible mechanism could remove most of the oxygen over a short time given known rates. The claim asserts that stopping production alone is enough: but we know current flow, stock and lifetime. They suggest the claim is impossible as stated. Even without assuming a model.
    – matt_black
    May 12, 2017 at 16:10
  • @matt_black you need to have evidence to support that you don't need to have a model
    – Sklivvz
    May 12, 2017 at 16:16
  • 2
    And you repeat the objection that there might be some secondary effect that means atmospheric oxygen won't be available at the surface: this is specifically referenced by looking at the literature which says the troposphere is well-mixed. And you assert that there might be secondary mechanisms that influence availability: the claim itself doesn't assume any such mechanisms. IF we allow complications like this we will have to accept homeopathy as I can't rule out a possible secondary mechanism that makes it work.
    – matt_black
    May 12, 2017 at 16:17
  • What model do we need to see that the stock of oxygen in the atmosphere is very large compared to the flow: that is a fact not a piece of theory.
    – matt_black
    May 12, 2017 at 16:18
  • There's a lot of energy in the solar system, but my radio is not working, why?
    – Sklivvz
    May 12, 2017 at 16:20
  • Assuming that since there's a lot of something then it's available everywhere is not always true. Why is it true in the case of the atmosphere, but not the sun's energy?
    – Sklivvz
    May 12, 2017 at 16:23
  • 5
    Because the literature of atmospheric science says the atmosphere is well-mixed. I've repeatedly pointed out that this is a well-known and referenced fact of tropospheric science, but you keep raising the hypothetical theoretical model that it might not be true. I'm not the one proposing unjustifiable hypotheses here.
    – matt_black
    May 12, 2017 at 16:25
  • @matt_black if that is true, than that's at least one reference that's missing from the answer, do you agree?
    – Sklivvz
    May 12, 2017 at 16:27
  • 5
    I do agree, but this reference has been added to previous versions of the answer and ignored. Asking for a reference for this fact is a much better response to the answer than ruling out the whole thing as being theoretical.
    – matt_black
    May 12, 2017 at 16:28
  • Let us continue this discussion in chat.
    – Sklivvz
    May 12, 2017 at 16:28

You must log in to answer this question.

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