What questions about the future can we answer, and how should they be answered?

Question

As discussed the question about hypothetical-tags and focusing more on Political Science, it seems that some discussion is warranted for how we can answer questions which are ostensibly about events which have not occurred. Not every hypothetical question can be answered, but many are actively discussed in the literature, and should be accessible on SE.

What questions can we answer? How should we answer those question?

Answer 1

I know you've thought out whether this kind of question should exist, and I applaud you for that, but I disagree with your conclusion. This remains primarily opinion-based. Expert answers are still opinions. We cannot answer until it has already happened. This SE is for facts, and questions like this just invite debate, which is not what we're here for.

Answer 2

By way of example of a good question and answer of this type please see "How likely is a war between the U.S. and China?". This is a question which is important and frequently discussed, and constitutes a question which is quite answerable by politics.SE. One thing to note is, while the question does not state it, the answers that can be given are about "expert opinion," not "What do you rando on the internet think is likely to happen?" The reason that is thus, is framing the question on 'expert opinion' makes it a dull question, and less likely to attract traffic, both from search engines, and from other SE stacks.

As illustrated in the example, I propose several criteria for hypothetical questions, and criteria for a good answer. If accepted, this may need to be added to the help section.

Criteria for Hypothetical Question

1. The hypothetical must have few possible outcomes. Yes or no answers can reasonably predicted, the distribution of voters in 2016 cannot.
2. Any assumptions must be clearly stated. "If ...", "Assuming...", are core.
3. There must be an existing theory wherein the answer can be couched, whether or not the theory is correct. The asker does not need to know the theory, but the answerer does. If the asker does know the theory, she could ask "According to The Party Decides will Trump actually be the Republican Nominee."
4. The answer must rely upon publicly available information. Nothing about decisions in the Oval Office or Kremlin, unless there have been public pronunciations on the topic.
5. There must be comparable past events, preferably several, from which to draw. No questions about "What would happen if the Mahdi returns?" since nothing comparable has ever happened.

Criteria for an Answer

1. NO ORIGINAL RESEARCH. Basically borrowing the rule from Wikipedia. Regardless of how the question is written, all hypotheticals are read as starting out "According to current experts in the relevant field...?" We don't care about you're opinion.
2. Links to reputable sources, especially peer reviewed journals and books, when possible.
3. Clearly state any assumptions, as well as the theories upon which one relies.
4. The answer must rely upon past outcomes as much as possible. In the example, this is less explicit because it is in the cited material, but basically the way the scholars ask and answer the question is by comparing the Sino-American relationship with the U.S.-U.K. or Germany-Russia at different times. Scholarly work always does this.
5. The answers given on SE should be evaluated on how well they reflect experts assesments, not on how well they reflect reality. Basically, if 20 years ago someone had asked if multiple regional wars are likely, and someone had answered with a summary of Huntington and someone else had answered with a summary of Fukuyama, they would both be right as long as they reflected the arguments, even though Huntington's predictions turned out to be more accurate.

Answer 3

TL;DR: This type of question is valid, provided that it asks for historical examples and/or current scholarly thought, not direct speculation about the future.

---

To elaborate on PointlessSpike's answer, which I agree with, consider these questions:

1. Who defines who is an expert? Does a well-reasoned post from Joe R. Blogger count, or is an offhand tweet from someone with a degree in the field more authoritative?
2. What if the experts disagree? You'll either end up with an answer that says "Maybe yes, maybe no", or you'll end up with two competing answers with no criteria to choose between them.
3. Who defines what a comparable event is? WWI and WWII were both major wars that involved all the major powers of the time. But the technologies, strategies, causes, alliances, and outcome of WWI were all different from those of WWII. Which one should we use as a historical example when talking about a hypothetical WWIII?
4. What happens if things change? Questions asking about whether or not something may happen can have their accepted answers invalidated by future developments, either in the form of counter-examples happening (see: Trump) or unanticipated changes to the situation (see: 9/11).

I do like the idea of this type of question, but they have to be framed in such a way that they are not open-ended. Your edit to the question about attacks on sporting events is a great example. It took a speculative question and turned it into a historical one. If such an event were to happen, it may or may not follow the historical trend, but the history itself is incontrovertible. Likewise, asking about the current opinion of experts is valid - if the situation changes, their opinions may change, but the historical opinions were valid when the question was asked.

---

Your example question is a good demonstration of both what not to do, and how it can be fixed.

1. The title is entirely speculative. A better (though less interesting) title would be something like "What do the experts say about a potential war between the US and China?" or something along those lines.
2. The lead-in/background paragraph is a mix of relevant background and overly generalized claims. This is the question, though, so that's OK.
3. The question originally asked about the likelihood of it occurring, which is entirely speculative.

However, Philipp's edit, changes it from speculation to "what do experts say?", which is very answerable. This more closely matches the answer you posted, which is a "here are the expert opinions" answer.

Answer 4

Evaluating "How likely is a war between the U.S. and China?" under the proposed criteria:

1. The hypothetical must have few possible outcomes. Yes or no answers can reasonably predicted, the distribution of voters in 2016 cannot.

This is a question with an infinite number of answers. Examples: 100%; 0%; 50%; and an infinite number of shades between them. There's also many possible reasons why there might be conflict between the two.

2. Any assumptions must be clearly stated. "If ...", "Assuming...", are core.

Well, let's quote something from the question:

In the past, when a country was threatened with losing its international position, this led to a war (e.g Germany fought Russia in WWI and the reverse in WWII).

That sounds like a possible reason for World War I. That doesn't fit my understanding of World War II at all. That could mean that my understanding is flawed. But since this is unsubstantiated, it seems more like an unstated assumption. Worse, the assumption is contestable, meaning that it will draw answers like flies to rotting flesh.

3. There must be an existing theory wherein the answer can be couched, whether or not the theory is correct. The asker does not need to know the theory, but the answerer does. If the asker does know the theory, she could ask "According to The Party Decides will Trump actually be the Republican Nominee."

The question might fit this, but this is a horrible criterion for a question. This means that the question's status should be determined by its answers. So when posted, the question will inherently not be in good shape. It's waiting to be saved by an answer. Evaluating questions should be based on the questions, not the answers.

4. The answer must rely upon publicly available information. Nothing about decisions in the Oval Office or Kremlin, unless there have been public pronunciations on the topic.

The question may fit this, but again it's evaluating the question based on the answer. Better might be: "Questions should not ask for private or classified information to answer." The example question would fit that.

5. There must be comparable past events, preferably several, from which to draw. No questions about "What would happen if the Mahdi returns?" since nothing comparable has ever happened.

I can buy this. Although I would point out that there is an argument that we have no comparable past events since no hot war has ever involved two nuclear powers.

---

So where does that leave us? Your example question fails two of your criteria clearly. It arguably passes one. If I rewrite another it passes that. Another criterion is clearly flawed. I find the difficulty that you are having writing an exemplary question for your point troublesome. If you can't write a great question, how can we expect new users to do so?

How's the question doing? Well, it's already drawn a non-answer from someone challenging a side issue and violating the answer criteria. But it's unflagged. That's not a good sign for your proposal for aggressive policing of answers.

Both answers have drawn significant comments which can be problematic. This isn't a discussion site. If people start discussing, the format breaks down. That's the essence of the "opinion-based" close reason. That opinions will be expressed in ways that draw counter-opinions and generate discussion.

As stands, my opinion is that this is a question that should be closed before it attracts too many other answers that don't fit the criteria.