Pinning keeps the focus on the exact question—especially playstyle
I commented on doppelgreener’s answer for unpinning with
I feel neutral-to-positive about this answer,
and that’s true—it’s a good answer, laying out very reasonable points in favor of unpinning the accepted answer.
But the more I think about it, the more it just doesn’t sit right with me. While I feel pretty good about doppelgreener’s answer, I don’t feel very good about actually taking the step it proposes.
There are other answers that capture some decent reasons to be leery of the change—I quite like Glazius’s about hope, and agree with others’ answers about the idea that there is value to giving the querent themselves a little bit of special privilege. And part of it is just how few situations it’s actually going to change, as Thomas Markov has determined for us. But I don’t feel that these things necessarily capture what I’m concerned about.
It applies to very few cases, and does little harm either way
This seems like a big change, that fundamentally affects how we regard querents, without actually improving the site very much. It applies to a small number of questions. That, I think, should give everyone some pause.
The difference seems primarily philosophical, not practical. It’s not a question of what will make the best site for users, because neither choice is going to substantially affect users’ experiences. But it is a fairly large question of what we think of the people asking questions.
Everyone is the world’s foremost expert on their own opinion
I vehemently disagree with the whole idea of "querent power". The querent most often isn't an expert.
This comment is more what I’m concerned about—it’s certainly true that the querent may well not be an expert in the subject matter, and that they may accept an answer “incorrectly”—there are definitely accepted answers that have irked me. But they are an expert on what it was they wanted—they are in fact the only expert on that. This community has, I think, more than a little bit of an issue when it comes to assuming we know better what people want—we don’t. I mean, look at this discussion, we have maybe 30 people paying attention to this.
We must take the position that the world’s foremost expert on what X wants is X, themselves. To do otherwise—to determine that we, in fact, know better what it is they truly want—is the height of hubris. That’s a vice that this community has been guilty of from time to time, but at least people had the option of promoting the answer that actually engages with their question as they see it.
So yeah, I think we should show some small, minor deference to querents in this—but the green checkmark alone is insufficient even for that. Do you even notice when the accepted answer is a self-answer and something after the first answer is accepted? I certainly don’t. That might as well not exist. Pinning an answer seems like an appropriate level, to me.
Plurality of playstyles, frame challenges, unpinned accepted answers—pick 2
Generally speaking, we prefer frame challenges to also answer the question straight. We do not require that they do so—and even when a frame challenge does, very often it’s a quick and simple answer that serves as much or more to illustrate the point of the challenge than it does to address the question.
This is acceptable because the querent can always choose another answer—and that answer gets pinned to the top, making it “the” answer insofar as we have one. Getting the top slot is what answerers “compete” for, again insofar as it’s a competition. The incentive structure of the site is based around that slot far, far more than it is around reputation. If it weren’t so, no one would object to the pinning of accepted answers in the first place, I’d think.
So the frame challenge policy itself states that they are “risky.” In fact, this isn’t “just” a statement—it is a part of that policy. The reason why frame challenges work the way they do—and are allowed in the form that they are—is because they are risky. This risk—that your answer won’t ever be considered for the top slot because of what it is—is crucial to the functioning of that policy. This is because frame challenges explicitly ignore parts of the question—that is, they decide to effectively answer a different question.
As long as the querent can reject that—and say “this is actually the answer to my question”—then this works out fine. But if the querent cannot “reject” the frame challenge—that is, if the community can “overrule” them and push a frame challenge into the top slot, and say “no, this is the answer to your question,” we have a big problem.
As doppelgreener put it,
FWIW, since writing [the pro-unpinning answer] I'm more in favour of keeping the pin [...] keeping our focus on the querent before our peers is ... probably really important.
Because one of the most prominent (in the sense of noticeable; I don’t have numbers) cases of such a frame challenge is over concerns of playstyle. That is, we have frame challenges that say “well, if you just abandoned that playstyle, this wouldn’t be a problem anymore.” If the playstyle in question is unpopular, that answer can very easily out-score even a good, strong “straight” answer to the question. We have numerous examples like this.
If the frame challenge convinces the querent, then fair enough. Sometimes querents don’t realize there even are other playstyles, or that the problems they’re facing are due to the playstyle they use but haven’t examined—sometimes a frame challenge like that can make someone realize the playstyle they’re using isn’t really suited to their interests. But if it doesn’t, then pushing a frame challenge to the top of the answers effectively invalidates their question. It says “the best answer to this question is to not ask this question.” That is a huge problem. And if answers know that they can achieve that result even without the querent buying in—that is, if they know they’re answering for the community’s preferences even when those explicitly disregard the needs of the querent—then the querent doesn’t get to define what the question is, anymore. Because if “the answer” is explicitly not answering the exact question, but rather answering a different one, and we decide it is nonetheless “the answer,” then what we have decided is that the question is actually different from what the querent said.
The querent decides that, not the community. Like I said, if they are convinced by the frame challenge, fine. But there must be an opportunity for them to not. And the incentive structure of the site—the “prize” that orients answers—has to reflect that. And the checkmark isn’t nearly as good a prize as the top slot is, so if the top slot is up for grabs even without the checkmark, the incentive structure says ignore the checkmark and go for the top slot.
What this all means is that we cannot support a plurality of playstyles, if questions about unpopular playstyles can be overruled by frame challenges pushing popular playstyles that don’t have the same problems. Pinning the accepted answer is one way to accomplish that—it leaves the ultimate determination of what it means to answer the question to the querent. But if we no longer pin the accepted answer, then we simply cannot have frame challenges any longer—at least not about playstyle, we can’t.
And that, I think, is a death knell here—because we can’t enforce that. By definition, we are talking about popular frame challenges—downvotes aren’t going to cut it. We can all agree that we’re supposed to downvote such frame challenges, and it likely won’t matter. Meta doesn’t have remotely that much influence—main site has a whole lot of people voting who are never going to look here. That means a ton of unpopular work for diamond moderators to delete high-rated answers—which is something that really should never happen in the first place. But there’s no other mechanism available to prevent it. And even if we succeed, we’ll still be worse off—because we won’t have the opportunity to even try to suggest that maybe the querent might be better off considering other ways of looking at the game.
We are not in the same position as Stack Overflow
Stack Overflow questions are near-purely objective. When the community disagrees with the querent, it almost-certainly does so because that answer is objectively, often measurably, superior to the answer the querent chose. Compilers and standards specifications create objective, unambiguous right and wrong answers.
Moreover, the vast majority of Stack Overflow interactions consist of someone coming from Google and copying-and-pasting the first thing they see. People go to Stack Overflow to get something working. There’s likely a time-pressure, and most likely, this isn’t fun for them—they have a job to get done.
That is not our site. Our content is usually subjective—and differences of opinion exist. Most times, therefore, if the community disagrees with the querent, it’s over matters of opinion—not actual objective quality—because that’s largely what we deal with. Expert opinion is valuable and useful—if it weren’t, this site wouldn’t serve much purpose—but it’s not the same as fact. On Stack Overflow, it’s very hard to imagine “the community” being substantially wrong about the correct answer to a question. Here, it’s fairly hard to imagine “right” or “wrong” even strictly applying in the first place. As Thomas Markov puts it, nothing here is going to be 100% objective—we don’t have any analogue to “this answer doesn’t even compile.”
And here, these are games. It’s a hobby. Most people are here, I think it’s safe to say, because they are interested in this subject. They enjoy reading about it. They’re much, much more likely to read more of what’s presented, rather than just going with the first thing they see. Especially when the second answer they come to has a higher number than the one before it. I think therefore that the harm of pinning a “wrong” answer—insofar as we even buy that there is such a thing—is vastly less here than it would be there.
So the fact that Stack Overflow did it doesn’t really mean very much here, I think. We’re in a different situation. It doesn’t mean that taking the same step is automatically wrong, but it does mean that it’s not automatically right. And I think there are some reasons here to hesitate.